Clinical Exercise Testing

mill. Whatever the choice of equipment (cycle ergometer Table 3. Useful measurements during CPET in ILD or treadmill) in a particular clinical laboratory, it is crucial to ensure its periodic calibration so that reliable, accurate, Measured variables reproducible and clinically relevant data can be collected Respiratory flow and volume (pneumotachograph) and interpreted. We have shown previously that such mea- Mean-expired and end-tidal PO2 and PCO2 (mass spectrometer/gas sures result in a high reproducibility (coefficient of varia- tion F5%) of both submaximal and peak exercise data in analyzer) patients with ILD (even in those who had not previously Arterial O2 saturation (pulse oximeter) performed cycle ergometer exercise) [17]. It is important to Heart rate (ECG electrodes) take this variability into consideration when conducting Work rate (cycle ergometer) and interpreting exercise tests in patients with ILD. Inspiratory capacity (IC) at rest, during and at end exercise Borg scores (dyspnea and leg fatigue) While exercise testing in a physician-supervised setting is considered relatively safe [9], it is important to consider Derived variables that the risk of serious morbidity or mortality in ILD Oxygen uptake (V˙ O2) patients during exercise is quite low (1 in 10,000) with Carbon dioxide output (V˙ CO2) deaths reported only in patients with known or suspected Respiratory exchange ratio (R = V˙ CO2/ V˙ O2) cardiac disease [85]. Precautionary measures such as a Minute ventilation (V˙ I) normal pre-exercise resting 12-lead ECG (to exclude ar- Ventilatory equivalents for O2 and CO2 (V˙ I/ V˙ O2, V˙ I/ V˙ CO2) rhythmias and/or significant ischemia), an experienced Tidal volume (VT) attendant physician equipped with facilities for cardio- Respiratory frequency (f) pulmonary resuscitation, contribute to the reduction of VD/VT – Dead space ventilation (from arterial blood gas risk significantly [86]. Furthermore, ECG monitoring throughout exercise and after (during the warm down measurements) period) is mandatory. While patients should be encour- End-expiratory (EELV) and end-inspiratory (EILV) lung volumes aged to exercise to their symptom-limited maximum, they Exercise tidal flow-volume loops should be given unambiguous instructions to stop exercise immediately if they develop any chest pain, dizziness or (V˙ I). Other variables such as dead space ventilation (VD/ other symptoms of significant discomfort. The attending VT) and ventilatory cost of O2 uptake and CO2 output (V˙ I/ physician should also stop the exercise test if the patient V˙ O2, V˙ I/V˙ CO2) can also be derived from the above data. develops any of the known and accepted indications for However, it is important to note that arterial blood-gas exercise stoppage [9]. Upon completion of the exercise measurements are required for accurate assessments of test, patients should be monitored for at least 10–15 min VD/VT [87]. Borg scores (dyspnea and leg fatigue) are more (both ECG and SaO2 monitoring) and should be recorded at rest, during exercise and especially at end exer- reassessed by the physician before being allowed to leave cise in our laboratory. Subjects are carefully instructed the laboratory [86]. Informed written consent must be about the usage of Borg scores before exercise commences. obtained from all patients before any exercise testing is Inspiratory capacity (IC) maneuvers at rest, during and at undertaken. end exercise are also recorded when tidal flow volume loops need to be analyzed for the assessment of operating Useful Measurements during CPET lung volumes and ventilatory mechanics during exercise. Table 3 summarizes some of the routinely available Finger-tip pulse oximetry is most commonly used for the and easy to measure variables using current computerized measurement of SaO2 in our laboratory, as it is an effective exercise testing systems. In most cases, respired airflow and a non-invasive measure of arterial O2 saturation [88]. (inspired and expired), O2 and CO2 concentrations, ECG While it is possible that pulse-oximetry may overestimate and SaO2 signals are acquired and digitized both for stor- true SaO2 at high work rates and at low work rates in peo- age and off-line data analysis of derived variables. These ple with a darker complexion [89] and the validity of SaO2 include: (1) O2 uptake (V˙ O2); (2) CO2 output (V˙ CO2); measurements using ear oximetry may be questionable (3) respiratory exchange ratio (R = V˙ CO2/V˙ O2); (4) [89, 90], pulse oximetry remains an effective, simple and end-tidal CO2 tension (PETCO2); (5) heart rate (HR); noninvasive method of measuring SaO2 during exercise (6) arterial O2 saturation (SaO2); (7) tidal volume (VT), [91]. Furthermore, pulse oximetry data have been shown and (8) breathing frequency (f) and minute ventilation to correlate well with simultaneous blood-gas measure- ments especially when trends in SaO2 (not absolute values in a clinically relevant range) are required [88, 91]. However, as shown before [5, 11], invasive techniques Exercise in Interstitial Lung Disease 191

(measurement of arterial blood gases, PAO2 – PaO2, VD/ Table 4. Characteristic responses during CPET in ILD VT) are sometimes necessary in the accurate assessment of both the nature and degree of gas exchange impairment Reduced peak V˙ O2 and W˙ max during exercise in some patients with ILD. Lower maximal heart rate Arterial O2 desaturation during and at end exercise Interpretation of Exercise Test Results Higher V˙ I at submaximal work rates Characteristic responses to exercise testing that are Reduced peak V˙ I demonstrated by patients with ILD are listed in table 4. High f, low VT at submaximal work rates As the interpretation of results of exercise testing in car- Normal peak VT/VC ratio diopulmonary disease is a complex process, the reader is Higher peak V˙ I/FEV1*35 ratio urged to consult additional literature on this topic [6–8, High VD/VT ratios 10, 12]. Briefly, it is important to remember that no set of Unchanged PaCO2 during exercise responses to exercise is invariable in any one disease state (e.g. O2 desaturation may occur in patients with chronic Results of Pulmonary Function Testing and CPET airflow limitation or pulmonary vascular disease). Fur- (Normals vs. ILD) thermore, it must be considered that the results of exercise Table 5 summarizes the results of pulmonary function testing in a patient with ILD may be influenced by a sec- testing and maximal incremental exercise testing in both ond disease process such as ischemic heart disease or oth- ILD patients and healthy young subjects [38, 84]. Where- er connective tissue disorders that may cause muscle ver possible, data have been presented as % predicted val- weakness. These factors as well as others such as poor ues. The ILD patients demonstrate a significant reduction physical fitness and/or malingering may also contribute to in both absolute and relative lung volumes and capacities exercise limitation in patients with ILD. The attending and a significant reduction in DLCO (!50% pred). There physician would therefore have to consider the results of was no evidence of airflow limitation in these patients exercise testing within the context of all other relevant (FEV1/FVC F80%). The data suggest that the exercise clinical information (history, physical examination, chest tests were maximal (note the significant O2 desaturation film, pulmonary function, etc.) for appropriate and effec- at end exercise) and that there is clear evidence of exercise tive interpretation. limitation in these ILD patients (see earlier). Figure 1a–h The guidelines for proper interpretation of cardiopul- illustrates the temporal course of some key variables dur- monary exercise testing in general include: (1) How lim- ing exercise testing in both patients with ILD and healthy ited is the patient? (2) What factors contribute to exercise subjects. It is evident from figure 1a that patients with limitation? (3) What are the abnormal patterns that are ILD stopped exercise at a much lower peak V˙ O2 (56% elicited? (4) Which clinical disorder(s) may result in these pred) and a lower peak heart rate (80% pred). However, patterns? It is also important to ensure that the exercise the heart rate response at a given metabolic rate was sig- test was maximal. The criteria for a maximal exercise test nificantly higher in the ILD patients compared to healthy are discussed elsewhere [6, 12]. Commonly demonstrated subjects. This phenomenon of an increased heart rate patterns of response to maximal exercise testing in ILD response to exercise in patients with significant ILD is patients are summarized in table 4 and include: (1) reduc- possibly secondary to a reduction in stroke volume that tion in peak V˙ O2 and W˙ max; (2) significant O2 desatura- apparently worsens with an increase in disease severity tion during and at end exercise; (3) reduction in peak [26, 43, 94]. heart rate; (4) reduction in peak V˙ I that approaches (or The most commonly observed phenomenon in ILD sometimes exceeds) maximum voluntary ventilation patients during exercise is arterial O2 desaturation and is (MVV) calculated from FEV1 (MVV = FEV1*35); shown in figure 1b. While SaO2 in healthy subjects re- (5) unchanged PaCO2 during exercise; (6) a rapid (↑f) and mains mostly stable throughout and till end exercise, it shallow (↓VT) breathing pattern, and (7) high VD/VT falls significantly (¢ F5%) throughout exercise in pa- ratios. These patterns of response are fairly similar among tients with ILD, reaching its nadir at peak exercise (SaO2 various disease processes that cause ILD [11, 24, 34, 43, has frequently been shown to fall to 85%, or lower). How- 78, 92, 93]. However, notable differences in ventilatory ever, as discussed earlier, while arterial O2 desaturation mechanics and degree of O2 desaturation have been dem- during exercise is typical of ILD patients, it is by no onstrated in patients with ILD and these seem related to means specific to ILD and may occur with exercise in oth- both the underlying pathology and disease severity. 192 Krishnan/Marciniuk

Table 5. Results of pulmonary function Variable Normals ILD testing and CPET (Normals vs. ILD) [38, 84] Age, years 21B2 49B15 n 9 (4 m, 5 f) 7 (6 m, 1 f) TLC, liters VC, liters 5.73B1.09 (99% pred) 4.60B0.25 (70% pred) FEV1, liters 4.79B0.91 (104% pred) 3.22B0.27 (70% pred) DLCO, %pred 3.84B0.69 (97% pred) 2.55B0.24 (74% pred) FEV1/FVC, % 89B9 48B7 Peak V˙ O2, liters/min 86B3 79B2 W˙ max, watts Peak heart rate 2.82B0.88 (98% pred) 1.32B0.05 (56% pred) Peak V˙ I, liters/min 209B68 (99% pred) 106B14 (55% pred) Peak V˙ I, %FEV1*35 186B11 (95% pred) 143B4 (80% pred) Peak VT/VC, % 106B44 SaO2 (end exercise, %) 63B6 76B22 75B25 53B8 57B13 95B2 84B2 er disease states [58, 91]. The relationship between V˙ CO2 The ventilatory response to exercise and the maximal and V˙ O2 is shown in figure 1c. Also shown is the line of predicted values for V˙ I (from MVV = FEV1*35) in both identity which corresponds to an R = 1, levels above groups are shown in figure 1e. While peak V˙ I is reduced which increased R values imply a change to anaerobic significantly in patients with ILD (compared to healthy metabolism. While it is an imprecise technique for the subjects), both groups achieved over 75% of peak pre- dicted V˙ I at end exercise. Furthermore, the increased ven- assessment of metabolic acidosis or the lactate threshold, the V˙ CO2 – V˙ O2 relationship allows a quick and non- tilatory response during submaximal exercise is typical of invasive estimate of the anaerobic threshold [95]. It is also patients with ILD and is further described in figure 1f as important to remember that changes in anaerobic thresh- V˙ I/V˙ CO2. This variable normally falls early in exercise and increases later during heavy exercise with the devel- old may be nonspecific and may be affected by specific opment of metabolic acidosis and hyperventilation. Nor- disease states, therapeutic interventions, including sup- mal subjects have a V˙ I/V˙ CO2 ratio of 25–35 during exer- cise and the elevated V˙ I/V˙ CO2 ratio in the ILD patients plemental O2 breathing [38]. implies that there was an increase in dead space ventila- Figure 1d describes the relationship between V˙ O2 and tion (VD/VT), with little or no increase in alveolar venti- PaCO2 (derived from PETCO2) [9]. PETCO2 represents a noninvasive estimate of PaCO2 and can be measured by lation (stable PaCO2; fig. 1d). most commercially available exercise testing systems. Figures 1f and 1g describe the differences in pattern of While PETCO2 has been shown to correlate well with breathing at various ventilatory levels between the actual PaCO2, its utility is only as an estimate (not a mea- sure) of PaCO2. The data show that normal subjects tend healthy subjects and ILD patients. In both groups, the ini- to regulate their PaCO2 at F40 mm Hg during exercise, tial increase in V˙ I at lower levels of V˙ I is achieved by a until the onset of metabolic acidosis and the compensato- combination of increases in VT and f. At higher V˙ I levels ry increase in ventilation, cause a fall in arterial PCO2. increases in f predominate in contributing to the increase The measurement of PETCO2 (or estimated PaCO2) in in V˙ I as VT gradually becomes asymptotic. The increase ILD patients during exercise has only limited diagnostic in f is consequent to a decrease in both inspiratory and value as it has been shown that PaCO2 remains stable and unchanged (from resting values) during exercise in pa- expiratory durations [10, 52]. As suggested earlier, pa- tients with ILD [10]. However, the stability of the PaCO2 tients with ILD adopt a rapid (↑f) and shallow (↓VT) throughout and at peak exercise implies that there was no breathing pattern at any given level of ventilation com- increase in alveolar ventilation in the patients with ILD pared to healthy subjects. It must noted here that while (compared to normal subjects). the ratio VTmax to VC has been suggested as being differ- ent in patients with ILD, it appears to have little or lim- ited diagnostic value [10], as similar ratios (VT/VC Exercise in Interstitial Lung Disease 193

Fig. 1. Results of exercise testing compared between healthy subjects (dashed lines, open circles) and patients with ILD (solid lines, closed circles) (see table 5 for more details) [38, 84]. V˙ O2 = Oxygen uptake; SaO2 = arterial O2 saturation; V˙ CO2 = CO2 output; PaCO2 = arterial CO2 tension; V˙ I = minute ventilation; VT = tidal volume. 194 Krishnan/Marciniuk

F55%) have been found in healthy subjects and in flow-volume loops) over the range of tidal breathing from patients with ILD, chronic airflow limitation or conges- measured exercise EELV. The term V˙ Ecap had originally tive cardiac failure [58, 96]. been used to describe such a measure of ventilatory capac- ity which is independent of volitional effort and accounts The abnormal respiratory mechanics results in a signif- for changes in dynamic airway function during exercise. icantly reduced maximal voluntary ventilation (MVV) or Thus, if there is no flow limitation (i.e. tidal expiratory ventilatory reserve (V˙ I/MVV) in patients with ILD [7, flow does not meet (or exceed) maximal available airflow 16]. The combination of an increased V˙ I at submaximal throughout expiration), maximal ventilatory capacity work rates and the reduction in MVV results in the (MVC = V˙ Ecap) is an effective measure of the ventilatory increase of the ventilatory demand/capacity (V˙ I/MVV) reserve for a given breathing pattern at any lung volume ratio during exercise. It is not uncommon for V˙ I/MVV to [41, 42]. More details on this technique and its superiority exceed 0.8 in many patients with significant ILD or V˙ I/ over the MVV (FEV1*35) method, in the assessment of MVV to approach 1 in patients with severe ILD during ventilatory reserve in a wide range of subjects during exer- exercise [18], resulting in significant ventilatory con- cise, have been published recently [14]. The results of straints on exercise performance. Traditionally the assess- analyses of mechanical constraints on exercise perfor- ment of the degree of ventilatory constraint has been mance based on exercise flow-volume relationships in based on estimation of the ventilatory reserve or how patients with ILD are now presented in figures 2 and 3. close to MVV (or its estimate) does peak V˙ I get during exercise. Ventilatory reserve is dependent on numerous Exercise Tidal Flow-Volume Relationships (Normals factors [14], chief among which are: (1) the maximal flow- vs. ILD) volume envelope (which is further dependent on age, sex, Figure 2 illustrates flow-volume relationships during body size, muscle function, genetic makeup, aging and exercise in both healthy subjects [84] and in patients with disease); (2) airway function (bronchodilatation or bron- ILD [42]. Data shown are maximal expiratory and inspi- choconstriction) during exercise, and (3) the lung volumes ratory loops and tidal loops at rest and at different exer- at which tidal breathing occurs relative to total lung cise intensities (50 and 100% V˙ O2max in healthy subjects capacity and residual volume (the limits of end-inspirato- during incremental exercise and at 40, 70 and 90% ry and end-expiratory lung volumes). It follows that V˙ O2max in ILD patients performing constant work-rate breathing at low lung volumes (near RV) the shape of the exercise). The data show that with increasing exercise maximal flow volume envelope limits the available venti- intensity, both healthy subjects and patients with ILD are latory reserve which can be further compromised by an able to increase inspiratory and expiratory flows and show increase in chest wall stiffness. Breathing at higher lung no evidence of expiratory flow limitation during exercise. volumes (near TLC), while contributing to a greater venti- While these increases in flows are possible even with a latory reserve, does result in increasing elastic loads on the significant fall in EELV (from resting values) in healthy inspiratory muscles and therefore the work of breathing. subjects, the patients with ILD show no change in EELV Even a visual analysis of the appropriately placed (within from resting values during exercise. This is as a result of the maximal loop) exercise tidal flow-volume loops can the marked reduction in expiratory reserve volume with provide valuable insights into the role of respiratory ILD [1, 39], with increases in exercise VT possible only mechanics and respiratory muscle energetics in their con- from the inspiratory reserve volume (↑EILV). It has been tribution to ventilatory limits on exercise performance. suggested that any possible fall in EELV could have been The estimation of ventilatory reserve based on FEV1 on affected by hypoxemia which has shown to affect resting the other hand neither provides any detailed information lung volume and respiratory mechanics [42, 99]. The on breathing strategy, nor account for the increasing increase in ventilation was thus more dependent on an inspiratory and expiratory flow constraints during exer- increase in breathing frequency (and increased flows) in cise. these patients. While these patients, as a group, do not show evidence of expiratory flow limitation, our study More recently, newer techniques that can quantify [42] showed that significant expiratory flow limitation, mechanical ventilatory constraints during exercise have high EILV/TLC and VT/VC ratios were present in those been developed, although these have not yet been vali- subjects who stopped exercise due to dyspnea. In contrast dated for clinical use [42, 97, 98]. Briefly, these tech- there was no flow limitation in patients who stopped exer- niques calculate a theoretical maximum ventilatory ca- cise due to leg fatigue. Interestingly, the patients with pacity (MVC) which is based on the maximal available inspiratory and expiratory airflows (from the maximal Exercise in Interstitial Lung Disease 195

2 3 196 Krishnan/Marciniuk

the most mechanical constraints (flow limitation and change significantly during exercise (from resting values), ↑EELV) had the least O2 desaturation but higher dyspnea patients with ILD are required to breathe at higher lung scores. These data combined with improved exercise per- volumes (EILV 190%) throughout exercise resulting in a formance with supplemental O2 breathing [38] imply that significant increase in respiratory muscle work and dys- while expiratory flow limitation may occur in some ILD pnea [50], both of which contribute significantly to exer- patients and may contribute to the dyspnea of exercise, it cise limitation in these patients. is arterial hypoxemia and not respiratory mechanics that plays a predominant role in exercise limitation in patients Impact of CPET on Patient Management in ILD with ILD. Ventilatory Mechanics and Lung Volumes during As exertional dyspnea and exercise intolerance are per- Exercise (Normals vs. ILD) haps the most important common clinical presentations Figure 3 summarizes the comparisons between the two of ILD, clinical management should include measures techniques (MVV and MVC) used in the estimation of that alleviate dyspnea as well as improve exercise capaci- ventilatory capacity during exercise in both healthy sub- ty, both of which have a direct and immediate impact on jects and patients with ILD. It is evident that even in the daily lives of these patients. Primarily, CPET provides healthy subjects the ventilatory demand/capacity ratio is the clinician an objective method of assessment of the higher when MVV (FEV1*35) is used as an index of venti- effects of the specific abnormalities caused by the disease, latory reserve, although these differences are small at low as well as assess the effectiveness of treatment of specific exercise intensities. However, in patients with ILD, the disease and its symptoms in these patients. O2 therapy, V˙ I/MVV ratio is significantly greater at all exercise inten- sufficient to prevent arterial O2 desaturation has been sities. It has been shown previously that MVC (estimated shown to improve exercise performance in ILD [15, 38] from flow-volume loops and breathing pattern) is signifi- and remains the mainstay of management of patients with cantly greater than MVV (FEV1*35) in patients with ILD severe ILD. Newer modalities of treatment of exertional [42]. Figure 3 emphasizes the fact that MVV thus signifi- dyspnea that have been tested and include inhaled mor- cantly underestimates potential ventilatory reserve at phine [100] or anesthetic [101] aerosols, have shown no least in patients with ILD during exercise. In contrast, the effects on exertional dyspnea, ventilatory response to V˙ I/MVC data (peak !60%) suggest that a significant exercise or exercise capacity. The effects of inhaled nitric potential for increasing ventilation exists in these patients oxide (NO) on pulmonary hypertension has also been during exercise. Figure 3 also describes differences in studied and it has been shown that while inhaled NO breathing pattern (i.e. tidal volume changes) during exer- (without supplemental O2) causes a significant fall in cise between healthy subjects and patients with ILD. mean pulmonary artery pressure and pulmonary vascular While VT in healthy subjects increases as a result of resistance, PaO2 does not improve [102]. It has also been changes in both EILV and EELV, it remains constant in shown that ILD patients have lower levels of intrinsic NO patients with ILD. Furthermore, as EELV does not and also fail to show any increase in NO production dur- ing exercise [103]. Thus, it is evident that CPET has an Fig. 2. Maximal (long-dashed lines) and tidal flow volume loops at important role to play in clinical decision making, indi- rest (thick solid lines) and during exercise at different intensities in vidual patient management and on the assessment of the healthy subjects (50% VO2max = medium-dashed line and 100% effectiveness of specific therapeutic modalities and physi- V˙ O2max = solid line) and in patients with ILD (40% V˙ O2max = cal rehabilitation in patients with ILD. small-dashed line, 70% V˙ O2max = medium-dashed line and 90% V˙ O2max = solid line) [42, 84]. V˙ O2max = Maximal O2 uptake. Acknowledgments Fig. 3. Comparison of ventilatory constraints and lung volumes dur- ing exercise in healthy subjects (open circles) and patients with ILD The research that was discussed in this report was supported by (closed circles) [42, 84]. V˙ I = Minute ventilation; FEV1 = forced expi- operating grants from the Medical Research Council of Canada, the ratory volume in 1 s; MVC = maximal ventilatory capacity; EILV = Heart and Stroke Foundation of Canada, the Saskatchewan Lung end-inspiratory lung volume; EELV = end-expiratory lung volume; Association and the Saskatchewan Health Research Board. The FRC = functional residual capacity; RV = residual volume; TLC = authors wish to express their appreciation to Mr. R. Clemens who has total lung capacity. ably assisted with the research and with the preparation of the manu- script. Exercise in Interstitial Lung Disease 197

References 1 Pride NB, Macklem PT: Lung Mechanics in 20 Cotes JE, Zejda J, King B: Lung function im- 37 Morrison DA, Stovall JR: Increased exercise Disease; in: Fishman AP, Macklem PT, Mead J pairment as a guide to exercise limitation in capacity in hypoxemic patients after long-term (eds): Handbook of Physiology: The Respirato- work-related lung disorders. Am Rev Respir oxygen therapy. Chest 1992;102:542–550. ry System, Mechanics of Breathing. Baltimore, Dis 1988;137:1089–1093. Williams & Wilkins, 1986, vol 2, pp 659–692. 38 Harris-Eze AO, Sridhar G, Clemens RE, Gal- 21 Javaheri S, Sicilian L: Lung function, breathing lagher CG, Marciniuk DD: Oxygen improves 2 Baydur A: Pulmonary physiology in interstitial pattern, and gas exchange in interstitial lung maximal exercise performance in interstitial lung disease: Recent developments in diagnos- disease. Thorax 1992;47:93–97. lung disease. Am J Respir Crit Care Med 1994; tic and prognostic implications. Curr Opin 150:1616–1622. Pulm Med 1996;2:370–375. 22 Kelley MA, Daniele RP: Exercise testing in interstitial lung disease. Clin Chest Med 1984; 39 Warren CP, Tse KS, Cherniack RM: Mechani- 3 Schwarz MI: Clinical overview of interstitial 5:145–156. cal properties of the lung in extrinsic allergic lung disease; in Manning S (ed): Interstitial alveolitis. Thorax 1978;33:315–321. Lung Disease. St. Louis, Mosby-Year Book, 23 Jernudd-Wilhelmsson Y, Hornblad Y, Heden- 1993, pp 1–22. stierna G: Ventilation-perfusion relationships 40 Hansen JE, Sue DY, Wasserman K: Predicted in interstitial lung disease. Eur J Respir Dis values for clinical exercise testing. Am Rev 4 Hsia CC: Cardiopulmonary limitations to exer- 1986;68:39–49. Respir Dis 1984;129:S49–55. cise in restrictive lung disease. Med Sci Sports Exerc 1999;31:S28–32. 24 Agusti AG, Roca J, Rodriguez-Roisin R, Xau- 41 Gallagher CG: Exercise and chronic obstruc- bet A, Agusti-Vidal A: Different patterns of gas tive pulmonary disease. Med Clin North Am 5 Hansen JE, Wasserman K: Pathophysiology of exchange response to exercise in asbestosis and 1990;74:619–641. activity limitation in patients with interstitial idiopathic pulmonary fibrosis. Eur Respir J lung disease. Chest 1996;109:1566–1576. 1988;1:510–516. 42 Marciniuk DD, Sridhar G, Clemens RE, Zintel TA, Gallagher CG: Lung volumes and expirato- 6 Younes M: Interpretation of clinical exercise 25 Agusti AG, Roca J, Gea J, Wagner PD, Xaubet ry flow limitation during exercise in interstitial testing in respiratory disease. Clin Chest Med A, Rodriguez-Roisin R: Mechanisms of gas- lung disease. J Appl Physiol 1994;77:963–973. 1984;5:189–206. exchange impairment in idiopathic pulmonary fibrosis. Am Rev Respir Dis 1991;143:219– 43 Spiro SG, Dowdeswell IR, Clark TJ: An analy- 7 Wasserman K, Hansen JE, Sue DY, Whipp BJ, 225. sis of submaximal exercise responses in pa- Casaburi R: Principles of Exercise Testing and tients with sarcoidosis and fibrosing alveolitis. Interpretation. Philadelphia, Lea & Febiger, 26 Hughes JM, Lockwood DN, Jones HA, Clark Br J Dis Chest 1981;75:169–180. 1994. RJ: DLCO/Q and diffusion limitation at rest and on exercise in patients with interstitial 44 Renzi G, Milic-Emili J, Grassino AE: The pat- 8 Weisman IM, Zeballos RJ: An integrated ap- fibrosis. Respir Physiol 1991;83:155–166. tern of breathing in diffuse lung fibrosis. Bull proach to the interpretation of cardiopulmo- Eur Physiopathol Respir 1982;18:461–472. nary exercise testing. Clin Chest Med 1994;15: 27 Agusti C, Xaubet A, Agusti AG, Roca J, Rami- 421–445. rez J, Rodriguez-Roisin R: Clinical and func- 45 Leblanc P, Bowie DM, Summers E, Jones NL, tional assessment of patients with idiopathic Killian KJ: Breathlessness and exercise in pa- 9 Jones NL: Clinical Exercise Testing. Philadel- pulmonary fibrosis: Results of a 3-year follow- tients with cardiorespiratory disease. Am Rev phia, Saunders, 1997. up. Eur Respir J 1994;7:643–650. Respir Dis 1986;133:21–25. 10 Marciniuk DD, Gallagher CG: Clinical exer- 28 Wagner PD: Ventilation-perfusion matching 46 Young IH, Daviskas E, Keena VA: Effect of cise testing in interstitial lung disease. Clin during exercise. Chest 1992;101:192S–198S. low dose nebulised morphine on exercise en- Chest Med 1994;15:287–303. durance in patients with chronic lung disease. 29 Crystal RG, Fulmer JD, Roberts WC, Moss Thorax 1989;44:387–390. 11 Markovitz GH, Cooper CB: Exercise and inter- ML, Line BR, Reynolds HY: Idiopathic pul- stitial lung disease. Curr Opin Pulm Med 1998; monary fibrosis: Clinical, histologic, radio- 47 Rampulla C, Baiocchi S, Dacosto E, Ambrosi- 4:272–280. graphic, physiologic, scintigraphic, cytologic, no N: Dyspnea on exercise: Pathophysiologic and biochemical aspects. Ann Intern Med mechanisms. Chest 1992;101:248S–252S. 12 Harris-Eze AO, Marciniuk DD: The role of 1976;85:769–788. exercise testing in interstitial lung disease. Clin 48 Mahler DA, Harver A, Lentine T, Scott JA, Beck Pulm Med 1995;2:213–220. 30 Nordenfelt I, Svensson G: The transfer factor K, Schwartzstein RM: Descriptors of breathless- (diffusing capacity) as a predictor of hypox- ness in cardiorespiratory diseases. Am J Respir 13 Johnson BD, Beck KC, Zeballos RJ, Weisman aemia during exercise in restrictive and chronic Crit Care Med 1996;154:1357–1363. IM: Advances in pulmonary laboratory testing. obstructive pulmonary disease. Clin Physiol Chest 1999;116:1377–1387. 1987;7:423–430. 49 Mahler DA, Harver A, Rosiello R, Dauben- speck JA: Measurement of respiratory sensa- 14 Johnson BD, Weisman IM, Zeballos RJ, Beck 31 Risk C, Epler GR, Gaensler EA: Exercise al- tion in interstitial lung disease: Evaluation of KC: Emerging concepts in the evaluation of veolar-arterial oxygen pressure difference in in- clinical dyspnea ratings and magnitude scaling. ventilatory limitation during exercise: The ex- terstitial lung disease. Chest 1984;85:69–74. Chest 1989;96:767–771. ercise tidal flow-volume loop. Chest 1999;116: 488–503. 32 Kelley MA, Panettieri RA Jr, Krupinski AV: 50 O’Donnell DE, Chau LK, Webb KA: Qualita- Resting single-breath diffusing capacity as a tive aspects of exertional dyspnea in patients 15 Bye PT, Anderson SD, Woolcock AJ, Young screening test for exercise-induced hypoxemia. with interstitial lung disease. J Appl Physiol IH, Alison JA: Bicycle endurance performance Am J Med 1986;80:807–812. 1998;84:2000–2009. of patients with interstitial lung disease breath- ing air and oxygen. Am Rev Respir Dis 1982; 33 Bradvik I, Wollmer P, Blom-Bulow B, Al- 51 De Troyer A, Yernault JC: Inspiratory muscle 126:1005–1012. brechtsson U, Jonson B: Lung mechanics and force in normal subjects and patients with inter- gas exchange during exercise in pulmonary sar- stitial lung disease. Thorax 1980;35:92–100. 16 Burdon JG, Killian KJ, Jones NL: Pattern of coidosis. Chest 1991;99:572–578. breathing during exercise in patients with inter- 52 DiMarco AF, Kelsen SG, Cherniack NS, Gothe stitial lung disease. Thorax 1983;38:778–784. 34 Crausman RS, Jennings CA, Tuder RM, Ack- B: Occlusion pressure and breathing pattern in erson LM, Irvin CG, King TE Jr: Pulmonary patients with interstitial lung disease. Am Rev 17 Marciniuk DD, Watts RE, Gallagher CG: Re- histiocytosis X: Pulmonary function and exer- Respir Dis 1983;127:425–430. producibility of incremental maximal cycle er- cise pathophysiology. Am J Respir Crit Care gometer testing in patients with restrictive lung Med 1996;153:426–435. 53 Fitting JW: Respiratory muscle fatigue limiting disease. Thorax 1993;48:894–898. physical exercise? Eur Respir J 1991;4:103– 35 Hansen JE, Wasserman K: Exercise testing in 108. 18 Marciniuk DD, Watts RE, Gallagher CG: patients with interstitial lung disease. Chest Dead space loading and exercise limitation in 1998;113:1148–1149. 54 Ameredes BT, Clanton TL: Hyperoxia and patients with interstitial lung disease. Chest moderate hypoxia fail to affect inspiratory 1994;105:183–189. 36 Bush A, Busst CM: Cardiovascular function at muscle fatigue in humans. J Appl Physiol 1989; rest and on exercise in patients with cryptogen- 66:894–900. 19 Epler GR, Saber FA, Gaensler EA: Determina- ic fibrosing alveolitis. Thorax 1988;43:276– tion of severe impairment (disability) in inter- 283. 55 Moxham J: Respiratory muscle fatigue: Mech- stitial lung disease. Am Rev Respir Dis 1980; anisms, evaluation and therapy. Br J Anaesth 121:647–659. 1990;65:43–53. 198 Krishnan/Marciniuk

56 NHLBI Workshop Summary: Respiratory at rest and on exercise in patients with idio- 89 Zeballos RJ, Weisman IM: Reliability of non- muscle fatigue. Report of the Respiratory Mus- pathic pulmonary fibrosis. Bull Eur Physiopa- invasive oximetry in black subjects during cle Fatigue Workshop Group. Am Rev Respir thol Respir 1982;18:403–410. exercise and hypoxia. Am Rev Respir Dis Dis 1990;142:474–480. 73 Packe GE, Cayton RM, Edwards CW: Com- 1991;144:1240–1244. parison of right ventricular size at necropsy in 57 Gallagher CG, Hof VI, Younes M: Effect of interstitial pulmonary fibrosis and in chronic 90 Hansen JE, Casaburi R: Validity of ear oxi- inspiratory muscle fatigue on breathing pat- bronchitis and emphysema. Thorax 1980;40: metry in clinical exercise testing. Chest 1987; tern. J Appl Physiol 1985;59:1152–1158. 712. 91:333–337. 74 Widimsky J, Riedel M, Stanek V: Central 58 Gallagher CG, Younes M: Breathing pattern haemodynamics during exercise in patients 91 Escourrou PJ, Delaperche MF, Visseaux A: during and after maximal exercise in patients with restrictive pulmonary disease. Bull Eur Reliability of pulse oximetry during exercise with chronic obstructive lung disease, intersti- Physiopathol Respir 1977;13:369–379. in pulmonary patients. Chest 1990;97:635– tial lung disease, and cardiac disease, and in 75 Weitzenblum E, Ehrhart M, Rasaholinjana- 638. normal subjects. Am Rev Respir Dis 1986;133: hary J, Hirth C: Pulmonary hemodynamics in 581–586. idiopathic pulmonary fibrosis and other in- 92 Miller A, Bhuptani A, Sloane MF, Brown LK, terstitial pulmonary diseases. Respiration Teirstein AS: Cardiorespiratory responses to 59 Harris-Eze AO, Sridhar G, Clemens RE, Zintel 1983;44:118–127. incremental exercise in patients with asbes- TA, Gallagher CG, Marciniuk DD: Role of 76 Agusti AG, Rodriguez-Roisin R: Effect of tos-related pleural thickening and normal or hypoxemia and pulmonary mechanics in exer- pulmonary hypertension on gas exchange. slightly abnormal lung function. Chest 1993; cise limitation in interstitial lung disease. Am J Eur Respir J 1993;6:1371–1377. 103:1045–1050. Respir Crit Care Med 1996;154:994–1001. 77 Follansbee WP, Curtiss EI, Medsger TA Jr, Owens GR, Steen VD, Rodnan GP: Myocar- 93 Karetzky M, McDonough M: Exercise and 60 Roberts WC, McAllister HA Jr, Ferrans VJ: dial function and perfusion in the CREST resting pulmonary function in sarcoidosis. Sarcoidosis of the heart. A clinicopathologic syndrome variant of progressive systemic Sarcoidosis Vasc Diffuse Lung Dis 1996;13: study of 35 necropsy patients (group 1) and sclerosis: Exercise radionuclide evaluation 43–49. review of 78 previously described necropsy pa- and comparison with diffuse scleroderma. tients (group 11). Am J Med 1977;63:86–108. Am J Med 1984;77:489–496. 94 Fischbein A, Luo JC, Solomon SJ, Horowitz 78 Forte S, Carlone S, Vaccaro F, et al: Pulmo- S, Hailoo W, Miller A: Clinical findings 61 Silverman KJ, Hutchins GM, Bulkley BH: Car- nary gas exchange and exercise capacity in among hard metal workers. Br J Ind Med diac sarcoid: A clinicopathologic study of 84 patients with systemic lupus erythematosus. J 1992;49:17–24. unselected patients with systemic sarcoidosis. Rheumatol 1999;26:2591–2594. Circulation 1978;58:1204–1211. 79 Keogh BA, Lakatos E, Price D, Crystal RG: 95 Beaver WL, Wasserman K, Whipp BJ: A new Importance of the lower respiratory tract in method for detecting anaerobic threshold by 62 Fleming HA: Sarcoid heart disease. Br Med J oxygen transfer: Exercise testing in patients gas exchange. J Appl Physiol 1986;60:2020– (Clin Res Ed) 1986;292:1095–1096. with interstitial and destructive lung disease. 2027. Am Rev Respir Dis 1984;129:S76–80. 63 Gibbons WJ, Levy RD, Nava S, et al: Subclini- 80 Mahler DA, Horowitz MB: Clinical evalua- 96 Gowda K, Zintel T, McParland C, Orchard R, cal cardiac dysfunction in sarcoidosis. Chest tion of exertional dyspnea. Clin Chest Med Gallagher CG: Diagnostic value of maximal 1991;100:44–50. 1994;15:259–269. exercise tidal volume. Chest 1990;98:1351– 81 Foster S, Thomas HM 3rd: Pulmonary reha- 1354. 64 Shah NS, Velury S, Mascarenhas D, Spodick bilitation in lung disease other than chronic DH: Electrocardiographic features of re- obstructive pulmonary disease. Am Rev Res- 97 Babb TG, Viggiano R, Hurley B, Staats B, strictive pulmonary disease, and comparison pir Dis 1990;141:601–604. Rodarte JR: Effect of mild-to-moderate air- with those of obstructive pulmonary disease. 82 Sue DY, Wasserman K: Impact of integrative flow limitation on exercise capacity. J Appl Am J Cardiol 1992;70:394–395. cardiopulmonary exercise testing on clinical Physiol 1991;70:223–230. decision making. Chest 1991;99:981–992. 65 Baughman RP, Gerson M, Bosken CH: Right 83 Sue DY: Exercise testing in the evaluation of 98 Johnson BD, Scanlon PD, Beck KC: Regula- and left ventricular function at rest and with impairment and disability. Clin Chest Med tion of ventilatory capacity during exercise in exercise in patients with sarcoidosis. Chest 1994;15:369–387. asthmatics. J Appl Physiol 1995;79:892–901. 1984;85:301–306. 84 Krishnan B, Clemens RE, Laframboise K, Nagpal A, Marciniuk DD: Does the ventilato- 99 Garfinkel F, Fitzgerald RS: The effect of hy- 66 Pinsky MR: Effects of changing intrathoracic ry response limit exercise performance in the peroxia, hypoxia and hypercapnia on FRC pressure on the normal and failing heart; in normal aged? Am J Resp Crit Care Med 1999; and occlusion pressure in human subjects. Lenfant C, Scharf SM, Cassidy SS (eds): Lung 159:A418. Respir Physiol 1978;33:241–250. Biology in Health and Disease, Heart-Lung In- 85 Rochmis P, Blackburn H: Exercise tests: A teractions in Health and Disease. New York, survey of procedures, safety, and litigation 100 Harris-Eze AO, Sridhar G, Clemens RE, Zin- Marcel Dekker, 1989, vol 42, pp 839–878. experience in approximately 170,000 tests. tel TA, Gallagher CG, Marciniuk DD: Low- JAMA 1971;217:1061–1066. dose nebulized morphine does not improve 67 Rodarte JR, Rehder K: Dynamics of Respira- 86 AARC Clinical Practice Guidelin: Exercise exercise in interstitial lung disease. Am J Res- tion; in Fishman AP, Macklem PT, Mead J testing for evaluation of hypoxemia and/or pir Crit Care Med 1995;152:1940–1945. (eds): Handbook of Physiology: The Respirato- desaturation. American Association for Re- ry System, Mechanics of Breathing. Baltimore, spiratory Care. Respir Care 1992;37:907– 101 Winning AJ, Hamilton RD, Guz A: Ventila- Williams & Wilkins, 1986, vol 2, pp 131–144. 912. tion and breathlessness on maximal exercise 87 Lewis DA, Sietsema KE, Casaburi R, Sue in patients with interstitial lung disease after 68 Welch HG: Effects of hypoxia and hyperoxia DY: Inaccuracy of noninvasive estimates of local anaesthetic aerosol inhalation. Clin Sci on human performance. Exerc Sport Sci Rev VD/VT in clinical exercise testing. Chest (Colch) 1988;74:275–281. 1987;15:191–221. 1994;106:1476–1480. 88 Martin D, Powers S, Cicale M, Collop N, 102 Yoshida M, Taguchi O, Gabazza EC, et al: 69 Miyamoto K, Nishimura M, Akiyama Y, Ya- Huang D, Criswell D: Validity of pulse oxi- The effect of low-dose inhalation of nitric mamoto H, Kishi F, Kawakami Y: Augmented mery during exercise in elite endurance ath- oxide in patients with pulmonary fibrosis. heart rate response to hypoxia in patients with letes. J Appl Physiol 1992;72:455–458. Eur Respir J 1997;10:2051–2054. chronic obstructive pulmonary disease. Am Rev Respir Dis 1992;145:1384–1388. 103 Riley MS, Porszasz J, Miranda J, Engelen MP, Brundage B, Wasserman K: Exhaled nit- 70 Davidson D, Stalcup SA, Mellins RB: Systemic ric oxide during exercise in primary pulmo- hemodynamics affecting cardiac output during nary hypertension and pulmonary fibrosis. hypocapnic and hypercapnic hypoxia. J Appl Chest 1997;111:44–50. Physiol 1986;60:1230–1236. Bharath S. Krishnan, MBBS, PhD 71 Enson Y, Thomas HM 3rd, Bosken CH, et al: Faculty of Physical Activity Studies Pulmonary hypertension in interstitial lung University of Regina, 3737 Wascana Parkway disease: Relation of vascular resistance to ab- Regina, Saskatchewan S4S 0A2 (Canada) normal lung structure. Trans Assoc Am Physi- Tel. +1 306 585 4370, Fax +1 306 585 4854 cians 1975;88:248–255. E-Mail [email protected] 72 Hawrylkiewicz I, Izdebska-Makosa Z, Grebska E, Zielinski J: Pulmonary haemodynamics Exercise in Interstitial Lung Disease 199

Weisman IM, Zeballos RJ (eds): Clinical Exercise Testing. Prog Respir Res. Basel, Karger, 2002, vol 32, pp 200–204 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Role of Cardiopulmonary Exercise Testing in Patients with Pulmonary Vascular Disease David M. Systrom Barbara A. Cockrill Charles A. Hales Pulmonary and Critical Care Unit, Medical Services, Massachusetts General Hospital and Harvard Medical School, Boston, Mass., USA Summary Pulmonary Hypertension Making the diagnosis of early and mild pulmonary hyperten- PHTN can be defined as a mean pulmonary artery sion through measurements performed on the patient at rest is pressure (PAP) at rest of greater than 25 mm Hg or more notoriously difficult. Most such patients present with dyspnea than 30 mm Hg with exercise. PHTN can occur due to or fatigue of unknown origin, which are classic indications for increased cardiac output (QT), pulmonary capillary wedge cardiopulmonary exercise testing (CPET). Noninvasive CPET pressure (PCWP) or pulmonary vascular resistance abnormalities suggestive of pulmonary hypertension include (PVR). The PHTN associated with high output states is low V˙ O2max, early anaerobic threshold, inefficient ventilation generally mild and of little clinical relevance unless there and arterial O2 desaturation. For the patient with relatively is a coexisting intrinsic abnormality of the pulmonary cir- severe, established precapillary pulmonary vascular disease, culation. Left-ventricular (LV) systolic dysfunction is fre- CPET can also be used safely to predict survival and to follow quent cause of PHTN and influences exercise capacity natural history or response to treatment. [1]. LV diastolic dysfunction is more likely to be associat- ed with dynamic exercise-induced PHTN. These forms of Introduction pulmonary venous hypertension, deserve mention be- cause they are associated with inefficient ventilation dur- This chapter will discuss the pathophysiology of the ing CPET [2] and may be confused with precapillary dis- exercise limit in pulmonary arterial hypertension (PHTN) ease. The detection and evaluation of diseases manifested and its detection by cardiopulmonary exercise testing by increased PVR is the subject of the present chapter. (CPET). The exercise responses in COPD, interstitial lung disease, congestive heart failure and the use of CPET Secondary Precapillary PHTN in lung transplantation and other thoracic surgical tech- niques are discussed elsewhere. Precapillary PHTN may be primary (PPH) or second- ary to a known disease process, but both forms have vary- ing degrees of pulmonary artery vasoconstriction and remodeling in common. Secondary PHTN may be a coin- cidental feature of a disease that predominately affects the

airways or lung interstitium or it may be largely responsi- Table 1. Secondary pulmonary vasculopa- ble for the patient’s pathologic and clinical presentation. thies For instance, many patients with COPD are more limited by lung mechanics [3] than by mild [4, 5] secondary Sleep disordered breathing [50] PHTN which has little bearing on exercise tolerance. Connective tissue disease [9] When the disease is advanced, however, the pulmonary HIV infection [51] vascular component may compromise aerobic capacity Chronic thromboembolic disease [52] [6] and influence survival [7, 8]. Patients with interstitial Portopulmonary hypertension [53] lung disease tend to present earlier than those with COPD Drugs [54] with clinical and CPET features of pulmonary vascular disease [9, 10]. Other secondary causes of PHTN (table 1) Normal Pulmonary Circulatory Response to not associated with significant obstruction or restriction Exercise are quite similar pathologically and clinically to PPH and for simplicity they will be discussed together. Early diag- The normal pulmonary circulation, even in the elderly, nosis of secondary PHTN is critical as withdrawal of the is a compliant structure able to receive a 5-fold increase in offending drug or treatment of the underlying cause may blood flow with only a minimal pressure rise. Right-sided avoid progressive PHTN with its attendant morbidity pressures do increase during incremental exercise [29, and mortality. 30], but at peak exercise RAP should be in the low teens, mean PAP !30 mm Hg and PCWP !20 mm Hg [29, 30]. Primary Pulmonary Hypertension Changes in PCW and RA pressure are linearly related. For each 1 mm Hg rise of RA pressure, PCW pressure PPH is by definition a disease of uncertain etiology rises 1.4 mm Hg [29]. A markedly increased or decreased characterized by a sustained elevation of PAP without slope of this relation suggests disproportionate left- or demonstrable cause [11, 12] after a comprehensive medi- right-ventricular dysfunction, respectively. cal work-up [13]. The early and correct diagnosis of PPH is also increasingly important given the development of The increase in pressure gradient across the pulmonary increasingly effective medical therapies [14–22] and be- vascular bed during exercise is less than the increase in cause of the long wait at most transplant centers for a cardiac output and pulmonary vascular resistance (PVR) donor lung [23]. therefore falls. This deviation from the classic Ohm-resis- tor pressure-flow relationship is due to both passive and While history, physical exam and laboratory work up active recruitment and distention of the pulmonary capil- may suggest secondary forms of PHTN, the diagnosis of lary bed [29, 30]. Reeves et al. [30] found that in young PPH is much more difficult [13]. Signs and symptoms healthy subjects, the upper limit of a normal PVRmax is related to PHTN are insensitive and nonspecific, as are 56 dyn WsW cm–5 (0.7 Wood units). Granath et al. [31] stud- tests of resting pulmonary function [24] and gas exchange ied 27 healthy older men (age 71 B 6 SD years) and found [25]. Transthoracic Doppler echocardiography has recent- an upper 95% confidence interval for PVRmax of ly evolved as a screening test of choice. Advantages 120 dyn Ws Wcm–5 (1.5 Wood units). include its noninvasive nature, ability to exclude second- ary left ventricular causes and quantification of PHTN as At peak exercise in the normal human, rapid red cell well as RV size and function. Disadvantages, however, transit time through the pulmonary capillary presents a include the fact that the clinician must consider cardiac challenge to oxygen loading. Increased alveolar PO2, im- causes of exercise intolerance in a young patient in proved matching of ventilation and perfusion in the ordering the test and underestimation of PAP in certain upright lung, recruitment of a structurally intact alveolar patients [26]. Perhaps even more important, some symp- capillary surface area and reduction of venous admixture tomatic patients appear to have only mild PHTN at rest prevent arterial O2 desaturation at high cardiac output, or none, but with the stress of exercise (which makes echo- with the exception of the rare elite endurance athlete [32, cardiography technically difficult) dynamic PHTN is elic- 33]. ited [25, 27]. If dynamic PHTN represent early and more treatable disease [5, 28], a screening test done solely at rest Ventilation, which is tightly linked to V˙ CO2 through- may be inappropriately insensitive. out incremental exercise, becomes more efficient if pul- monary circulatory structure and function are normal. A Pulmonary Vascular Disease 201

Table 2. Abnormal pulmonary vascular re- tion, see ref. 6]. CPET done with a right heart catheter in sponse to incremental exercise place will reveal an abnormal rise in right atrial and mean pulmonary artery pressures [38] (table 2) with a relatively PAP 1 30 mm Hg [11, 12, 40] normal PCWP. The combination of increased pressure PVR 1 1.5 Wood units [31] gradient across the pulmonary vascular bed and inade- Q˙ t ! 80% predicted quate cardiac output leads to a blunted fall of PVR. PCWP/RAP slope ! 1.4 [29] RAP 1 14 mm Hg [29] Although right heart catheterization is a critical step in RVEF ! 0.45 [55] the confirmation, grading and treatment of PHTN, its attendant risks [39, 40] suggest the need for an alternative screening test that is noninvasive, safe, reproducible and sensitive. Table 3. CPET characteristics of PHTN* Decreased V˙ O2max Diagnosis of PHTN by CPET Decreased AT Increased V˙ E/V˙ CO2 slope and absolute value at AT The pitfalls in the diagnosis of early and mild PHTN Hyperventilation during submaximal exercise noted above have provided impetus for CPET interpreta- Blunted VD/VT fall tion algorithms based on patterns of change of multiple Arterial O2 desaturation or widened P(A-a)O2 physiologic responses to exercise [41–43] (table 3). Most patients with PPH present with dyspnea or fatigue of * See ref [34, 35, 41] for normal values. unknown origin [11, 12, 34], which are classic indications for CPET [41]. decrease in alveolar dead space occurs as part of pulmo- nary vascular distention and recruitment, especially at the By the time the patient is symptomatic, V˙ O2max is apices of the upright lung. This and normal augmentation usually significantly reduced, though well described cases of VT decrease physiologic VD/VT to ! 0.3 at peak exer- of significant PHTN have been described with surprising cise. The V˙ E/V˙ CO2 falls as a result to ! 37 at the ventilato- well-preserved overall aerobic capacity [39]. A depressed ry threshold [34], though this normal response is depen- maximum O2 uptake is related to decreased maximal car- dent on age and gender [35]. diac output (discussed above) and arterial O2 desatura- tion. Inadequate O2 delivery is associated with a low ven- Abnormal Pulmonary Circulatory Response to tilatory and lactate ‘anaerobic’ thresholds (AT) [34]. Exercise Ventilation in patients with PHTN is excessive [44], Direct measurement of pulmonary hemodynamics at but because the maximum voluntary ventilation is usually rest or as part of CPET (table 2) gives insight into the near-normal, relatively normal breathing reserve at peak pathogenesis of symptoms, can confirm the diagnosis of exercise is most often found [34]. Inefficient ventilation PHTN, quantify disease severity, help with prognostica- during incremental exercise, usually denoted by an in- tion and guide therapy. In PHTN, pulmonary vasocon- creased slope of the linear phase of V˙ E/V˙ CO2 or its abso- striction and structural remodeling limit the stroke vol- lute value at the ventilatory threshold however, is a char- ume response to CPET through increased right heart acteristic of both arterial [34] and venous [2] PHTN. The afterload [34]. When the right ventricle dilates chronically importance of V˙ E/V˙ CO2 is explained by its inverse rela- or dynamically during exercise, ventricular interdepen- tionship to arterial PaCO2 and physiologic dead space to dence decreases LV compliance and diastolic filling. tidal volume ratio: V˙ E/V˙ CO2 = k/(PaCO2 (1 – VD/VT)). Failure of normal pulmonary vascular distention and In patients with PHTN, a first pass radionuclide heart recruitment during exercise in PPH creates underper- scan may show a compromised rise in RVEF and normal fused, ventilated lung units, increasing VD/VT and the LVEF [36], blunted left-ventricular end-diastolic volume ventilatory requirement of exercise. Hyperventilation response, SV and Q˙ tmax [37] !80% predicted [for calcula- during submaximal exercise is also seen in PPH and likely related to hypoxemic stimulation of arterial chemorecep- tors and stimulation of pulmonary circulatory mechano- receptors [45]. Thus, the combined influence of two inva- sively measured variables, VD/VT and PaCO2, makes the 202 Systrom/Cockrill/Hales

noninvasive V˙ E/V˙ CO2 a potentially powerful marker of An increasing number of studies have utilized the 6 the abnormal pulmonary vasculature. min walk or CPET to follow patients with established PPH and its response to therapy [16, 22, 47–49]. Iwase Arterial O2 desaturation and/or exaggerated widening and colleagues [49] recently described the temporal pat- of the P(A-a)O2 with exercise is thought to be a gas tern of recovery of CPET parameters following pulmo- exchange feature that distinguishes pulmonary arterial nary thromboendarterectomy for chronic thromboem- from venous hypertension [41]. In precapillary disease, bolic PHTN. Interestingly, ventilatory efficiency returned abnormally widened P(A-a)O2 is due both to V˙ /Q˙ mis- toward normal during the first postoperative month and matching and a diffusion defect induced by a rapid red correlated with improved PVR. V˙ O2max, likely reflecting cell transit time through a poorly compliant and recruita- whole body effects of chronic disease, took longer to ble pulmonary circulation. Occasionally, a sudden fall in recover [49]. arterial oxygen saturation heralds the opening of a patent foramen ovale and increased right to left shunting due to Conclusions elevation of right atrial pressure [41]. Grading PHTN Severity and Therapeutic Response Cardiopulmonary exercise testing is a useful tool for by CPET the detection of pulmonary vascular disease, especially in the patient with dyspnea or fatigue of unknown In PPH, CPET parameters of aerobic function and gas origin. CPET characteristics of PHTN include abnormal exchange such as V˙ O2max, AT, V˙ E/V˙ CO2 slope and abso- V˙ O2max, early AT, inefficient ventilation and arterial O2 lute value at AT correlate with NYHA classification [34]. desaturation. For the PPH patient with relatively severe, V˙ O2max and AT also correlate with resting pulmonary established disease, CPET can be performed safely [34], hemodynamics [34]. In one recent study which employed and used to stratify survival and to follow natural history micromanometer tipped pulmonary artery catheters dur- or response to treatment. ing CPET, ventilatory equivalents were correlated with pulmonary hemodynamics [38]. V˙ O2max correlates with Acknowledgments survival in chronic thromboembolic pulmonary hyperten- sion [26]. In established PPH, the 6 minute walk also pre- Dr. Systrom is supported by NHLBI 1K24 HLO4022-02. Dr. dicts survival [8, 46]. Hales is supported by NIH HL39150-13 and HL63982-02. References 5 Kessler R, Faller M, Weitzenblum E, Chaouat 10 Orfanos SE, Psevdi E, Stratigis N, Langleben A, Aykut A, Ducolone A, Ehrhart M, Oswald- D, Catravas JD, Kyriakidis M, Moutsopoulos 1 Koelling TM, Kirmse M, Di Salvo TG, Dec Mammosser M: ‘Natural history’ of pulmonary HM, Roussos C, Vlachoyiannopoulos PG: Pul- GW, Zapol WM, Semigran MJ: Inhaled nitric hypertension in a series of 131 patients with monary capillary endothelial dysfunction in oxide improves exercise capacity in patients chronic obstructive lung disease. Am J Respir early systemic sclerosis. Arthritis Rheum 2001; with severe heart failure and right ventricular Crit Care Med 2001;164:219–224. 44:902–911. dysfunction. Am J Cardiol 1998;81:1494– 1497. 6 Oelberg DA, Kacmarek RM, Pappagianopou- 11 Rubin LJ: Primary pulmonary hypertension. N los PP, Ginns LC, Systrom DM: Ventilatory Engl J Med 1997;336:111–117. 2 Kleber FX, Vietzke G, Wernecke KD, Bauer and cardiovascular responses to inspired He- U, Opitz C, Wensel R, Sperfeld A, Glaser S: O2 during exercise in chronic obstructive pul- 12 Gaine SP, Rubin LJ: Primary pulmonary hy- Impairment of ventilatory efficiency in heart monary disease. Am J Respir Crit Care Med pertension. Lancet 1998;352:719–725. failure: prognostic impact. Circulation 2000; 1998;158:1876–1882. 101:2803–2809. 13 McGoon M: The assessment of pulmonary hy- 7 Pierson DJ: Pathophysiology and clinical ef- pertension. Clin Chest Med 2001;22:493–508. 3 Medoff BD, Oelberg DA, Kanarek DJ, Systrom fects of chronic hypoxia. Respir Care 2000;45: DM: Breathing reserve at the lactate threshold 39–51; discussion 51–3. 14 Galie N, Manes A, Branzi A: Medical therapy to differentiate a pulmonary mechanical from of pulmonary hypertension: The prostacyclins. cardiovascular limit to exercise. Chest 1998; 8 Stricker H, Domenighetti G, Popov W, Speich Clin Chest Med 2001;22:529–537. 113:913–918. R, Nicod L, Aubert JD, Soler M: Severe pulmo- nary hypertension: Data from the Swiss Regis- 15 Naeije R, Vachiery J-L: Medical therapy of pul- 4 Stevens D, Sharma K, Szidon P, Rich S, try. Swiss Med Wkly 2001;131:346–350. monary hypertension. Clin Chest Med 2001; McLaughlin V, Kesten S: Severe pulmonary 22:517–527. hypertension associated with COPD. Ann 9 Hsia CC: Cardiopulmonary limitations to exer- Transplant 2000;5:8–12. cise in restrictive lung disease. Med Sci Sports 16 Machherndl S, Kneussl M, Baumgartner H, Exerc 1999;31:S28–32. Schneider B, Petkov V, Schenk P, Lang IM: Long-term treatment of pulmonary hyperten- sion with aerosolized iloprost. Eur Respir J 2001;17:8–13. Pulmonary Vascular Disease 203

17 Naeije R, Vachiery JL: Medical therapy of pul- 31 Granath A, Strandell T: Circulation in healthy 46 Miyamoto S, Nagaya N, Satoh T, Kyotani S, monary hypertension: Conventional therapies. old men, studied by right heart catheterization Sakamaki F, Fujita M, Nakanishi N, Miyatake Clin Chest Med 2001;22:517–527. at rest and during exercise in supine and sitting K: Clinical correlates and prognostic signifi- position. Act Med Scand 1964;176:425–446. cance of six-minute walk test in patients with 18 Nagaya N, Uematsu M, Oya H, Sato N, Saka- primary pulmonary hypertension: Comparison maki F, Kyotani S, Ueno K, Nakanishi N, 32 Hopkins SR, Schoene RB, Henderson WR, with cardiopulmonary exercise testing. Am J Yamagishi M, Miyatake K: Short-term oral ad- Spragg RG, Martin TR, West JB: Intense exer- Respir Crit Care Med 2000;161:487–492. ministration of L-arginine improves hemody- cise impairs the integrity of the pulmonary namics and exercise capacity in patients with blood-gas barrier in elite athletes. Am J Respir 47 Barst RJ, Rubin LJ, McGoon MD, Caldwell precapillary pulmonary hypertension. Am J Crit Care Med 1997;155:1090–1094. EJ, Long WA, Levy PS: Survival in primary Respir Crit Care Med 2001;163:887–891. pulmonary hypertension with long-term con- 33 Dempsey JA, Johnson BD, Saupe KW: Adap- tinuous intravenous prostacyclin. Ann Intern 19 Nauser TD, Stites SW: Diagnosis and treat- tations and limitations in the pulmonary sys- Med 1994;121:409–415. ment of pulmonary hypertension. Am Fam tem during exercise. Chest 1990;97(suppl): Physician 2001;63:1789–1798. 81S–87S. 48 Channick RN, Simonneau G, Sitbon O, Rob- bins IM, Frost A, Tapson VF, Badesch DB, 20 Rubin LJ, Mendoza J, Hood H, McGoon M, 34 Sun XG, Hansen JE, Oudiz RJ, Wasserman K: Roux S, Rainisio M, Bodin F, Rubin LJ: Ef- Barst R, Williams WB, Diehl JH, Crow J, Long Exercise pathophysiology in patients with pri- fects of the dual endothelin-receptor antagonist W: Treatment of primary pulmonary hyperten- mary pulmonary hypertension. Circulation bosentan in patients with pulmonary hyperten- sion with continuous intravenous prostacyclin 2001;104:429–435. sion: A randomised placebo-controlled study. (epoprostenol): Results of a randomized trial. Lancet 2001;358:1119–1123. Ann Intern Med 1990;112:485–491. 35 Habedank D, Reindl I, Vietzke G, Bauer U, Sperfeld A, Glaser S, Wernecke KD, Kleber 49 Iwase T, Nagaya N, Ando M, Satoh T, Sakama- 21 Stricker H, Domenighetti G, Fiori G, Mombel- FX: Ventilatory efficiency and exercise toler- ki F, Kyotani S, Takaki H, Goto Y, Ohkita Y, li G: Sustained improvement of performance ance in 101 healthy volunteers. Eur J Appl Uematsu M, Nakanishi N, Miyatake K: Acute and haemodynamics with long-term aerosol- Physiol Occup Physiol 1998;77:421–426. and chronic effects of surgical thromboendar- ised prostacyclin therapy in severe pulmonary terectomy on exercise capacity and ventilatory hypertension. Schweiz Med Wochenschr 1999; 36 Matthay RA, Arroliga AC, Wiedemann HP, efficiency in patients with chronic thromboem- 129:923–927. Schulman DS, Mahler DA: Right ventricular bolic pulmonary hypertension. Heart 2001;86: function at rest and during exercise in chronic 188–192. 22 Wax D, Garofano R, Barst RJ: Effects of long- obstructive pulmonary disease. Chest 1992; term infusion of prostacyclin on exercise per- 101:255S–262S. 50 Bady E, Achkar A, Pascal S, Orvoen-Frija E, formance in patients with primary pulmonary Laaban JP: Pulmonary arterial hypertension in hypertension. Chest 1999;116:914–920. 37 Maroni JM, Oelberg DA, Pappagianopoulos P, patients with sleep apnoea syndrome. Thorax Boucher CA, Systrom DM: Maximum cardiac 2000;55:934–939. 23 Trulock E: Lung transplantation for primary output during incremental exercise by first- pulmonary hypertension. Clin Chest Med pass radionuclide ventriculography. Chest 51 Mehta NJ, Khan IA, Mehta RN, Sepkowitz 2001;22:583–593. 1998;114:457–461. DA: HIV-Related pulmonary hypertension: Analytic review of 131 cases. Chest 2000;118: 24 Moser K, Auger W, Fedullo P: Chronic throm- 38 Raeside DA, Smith A, Brown A, Patel KR, 1133–1141. boembolic pulmonary hypertension: Clinical Madhok R, Cleland J, Peacock AJ: Pulmonary picture and surgical treatment. Eur Respir J artery pressure measurement during exercise 52 Fedullo P, Auger W, Channick R, Kerr K, 1992;5:334–342. testing in patients with suspected pulmonary Rubin L: Chronic thromboembolic pulmonary hypertension. Eur Respir J 2000;16:282–287. hypertension. Clin Chest Med 2001;22:561– 25 Mohsenifar Z, Tashkin ED, Levy SE, Bjerke 581. RD, Clements PJ, Furst D: Lack of sensitivity 39 Rhodes J, Barst R, Garofano R, Thoele D, Ger- of measurements of VD/VT at rest and during sony W: Hemodynamic correlates of exercise 53 Krowka MJ, McGoon MD: Portopulmonary exercise in detection of hemodynamically sig- function in patients with primary pulmonary hypertension: The next step. Chest 1997;112: nificant pulmonary vascular abnormalities in hypertension. J Am Coll Cardiol 1991;18: 869–870. collagen vascular disease. Am Rev Respir Dis 1738–1744. 1981;123:508–512. 54 Abenhaim L, Moride Y, Brenot F, Rich S, 40 Rich S, Dantzker DR, Ayres SM, Bergofsky Benichou J, Kurz X, Higenbottam T, Oakley C, 26 Lewczuk J, Piszko P, Jagas J, Porada A, Woj- EH, Brundage BH, Detre KM, Fishman AP, Wouters E, Aubier M, Simonneau G, Begaud ciak S, Sobkowicz B, Wrabec K: Prognostic fac- Goldring RM, Groves BM, Koerner SK, et al: B: Appetite-suppressant drugs and the risk of tors in medically treated patients with chronic Primary pulmonary hypertension: A national primary pulmonary hypertension. Internation- pulmonary embolism. Chest 2001;119:818– prospective study. Ann Intern Med 1987;107: al Primary Pulmonary Hypertension Study 823. 216–223. Group. N Engl J Med 1996;335:609–616. 27 Mohsenifar Z, Taskin DP, Wolfe JD, Genovese 41 Wasserman K, Hansen JE, Sue DY, Casaburi 55 Brent BN, Mahler D, Matthay RA, Berger HJ, M: Abnormal responses of wasted ventilation R, Whipp BJ: Principles of Exercise Testing Zaret BL: Noninvasive diagnosis of pulmonary fraction (VD/VT) during exercise in patients and Interpretation. Baltimore, Lippincott, Wil- arterial hypertension in chronic obstructive with pulmonary vascular abnormalities. Respi- liams & Wilkins, 1999, pp 173–174. pulmonary disease: Right ventricular ejection ration 1983;44:44–49. fraction at rest. Am J Cardiol 1984;53:1349– 42 Weisman IM, Zeballos RJ: An integrated ap- 1353. 28 Kullo I, Tajik A, McGoon M: Unexplained proach to the interpretation of cardiopulmo- mild pulmonary hypertension detected by nary exercise testing. Clin Chest Med 1994;15: David M. Systrom, MD Doppler echocardiography: Clinical profile and 421–445. Pulmonary and Critical Care Unit follow-up of 29 patients. Circulation 1998;98: Medical Services I-615. 43 Eschenbacher WL, Mannina A: An algorithm Massachusetts General Hospital and for the interpretation of cardiopulmonary exer- Harvard Medical School 29 Reeves JT, Groves BM, Cymerman A, et al: cise tests. Chest 1990;97:263–267. Fruit Street, Bigelow 952 Operation Everest II: Cardiac filling pressures Boston, MA 02114 (USA) during cycle exercise at sea level. Respir Physi- 44 Theodore J, Robin ED, Morris AJR, Burke Tel. +1 617 726 3734, Fax +1 617 724 6954 ol 1990;80:147–154. CM, Jamieson SW, Van Kessel A, Stinson EB, E-Mail [email protected] Shumway NE: Augmented ventilatory re- 30 Reeves JT, Moon RE, Grover RF, et al: In- sponse to exercise in pulmonary hypertension. creased wedge pressure facilitates decreased Chest 1986;89:39–44. lung vascular resistance during upright exer- cise. Chest 1988;93(suppl):97S–99S. 45 Jones PW, Huszczuk A, Wasserman K: Car- diac output as a controller of ventilation through changes in right ventricular load. J Appl Physiol 1982;53:218–224. 204 Systrom/Cockrill/Hales

Weisman IM, Zeballos RJ (eds): Clinical Exercise Testing. Prog Respir Res. Basel, Karger, 2002, vol 32, pp 205–216 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Asthma and Exercise Ricardo A. Tan Sheldon L. Spector California Allergy and Asthma Medical Group, Inc., Palmdale, Calif., USA Summary and in 6–13% of the general population [3]. The presence of atopy is strongly associated with EIA [4]. Up to 40% of Exercise-induced asthma (EIA) is characterized by transient patients with allergic rhinitis have EIA. In children and airway obstruction after strenuous exertion. Methacholine and young adults, EIA is especially significant as their early histamine challenge testing may be done to confirm suspected exposure to sports can be affected by this problem [5]. For asthma but a negative response does not rule out EIA. Exercise athletic individuals, asthma symptoms induced only by testing on a treadmill or cycle ergometer or hyperventilation exertion can be sufficiently limiting. As more athletes testing is needed to rule out EIA. Prevention is the main objec- with asthma compete and succeed at the highest levels of tive in managing EIA. Nonpharmacologic measures include performance, it is important to be familiar with the nature warming up, covering the mouth and nose in cold weather, and and therapy of exercise-induced asthma so that we can warming down after exercise. Inhaled ß-agonists are the medi- encourage our patients to continue an active life in a safe cations of choice for EIA prophylaxis. Inhaled cromolyn or and healthy manner. nedocromil and antileukotriene agents may also be effective. Education regarding the nature and management of EIA is Pathophysiology important not only for asthmatics but also their families and coaches. With the proper precautions and workout tech- The pathogenesis of exercise-induced asthma is not niques, there is no limit to what persons with asthma can clearly defined. However, there are physiologic events achieve in sports. about which there is general agreement. Inhalation of large volumes of dry, cold air during exercise leads to loss Introduction of heat and water from the bronchial mucosa leading to airway cooling and drying. Numerous studies have docu- Exercise-induced asthma (EIA), also referred to as mented the importance of (1) the amount of ventilation, exercise-induced airway narrowing or exercise-induced and (2) the temperature of inspired air in producing air- bronchospasm (EIB), is defined as acute airway narrow- way obstruction [6, 7]. The larger the amount of ventila- ing occurring after strenuous exertion. Exercise and hy- tion and the colder the air inspired, the greater the severi- perventilation have long been observed to be significant ty of airway narrowing. Warm, humidified air diminishes triggers of bronchial obstruction in chronic asthma. Some the degree of bronchospasm. It is important to note that asthmatics exhibit symptoms only after exercise with no despite these general observations, EIA can still occur other triggers. EIA occurs in 40–90% of asthmatics [1, 2] even when exercise is done in a warm environment [8].

Table1. Clinical features of EIA ing obstruction or poor conditioning rather than true exer- cise-induced bronchoconstriction [15]. Requires 3–8 min of exercise at 80% or more of maximal heart rate Peak symptoms occur at 8–15 min after exercise Clinical Features Spontaneous recovery within 60 min Refractory period of up to 4 h after exercise Exercise-induced asthma is usually preceded by at least Bronchospasm provoked faster by dry and cold air 3–8 min of exercise. Bronchospasm and/or symptoms of chest tightness, cough, wheezing and dyspnea start soon There are two prominent theories about how airway after the end of exercise and peak in about 8–15 min [16]. cooling leads to bronchoconstriction. EIA is believed to There is spontaneous recovery within 60 min after the involve (a) mucosal drying and increased osmolarity stim- exercise ends (table 1). A late phase response has been ulating mast cell degranulation [9] and/or (b) rapid airway described occurring 4–6 h after exercise in some patients rewarming causing vascular congestion, increased perme- [17]. Other studies have failed to demonstrate this re- ability and edema leading to obstruction [10]. A combina- sponse and there is controversy regarding whether a late tion of these two mechanisms is more likely at work in phase response to exercise exists or whether the observed EIA. changes are due to fluctuations in the underlying asthma [18]. Although EIA has traditionally been observed only The first theory proposes that the increase in osmolari- after exercise, newer studies suggest that bronchospasm ty in the airway mucosa produced by airway drying and may occur during exercise, especially during prolonged cooling stimulates the release of inflammatory mediators exertion. from mast cells, including histamine and leukotrienes, that cause bronchoconstriction [9, 11]. The efficacy of the If the person attempts to exercise again after the symp- mast cell stabilizers (cromolyn and nedocromil) and anti- toms subside, he will experience less symptoms the sec- leukotriene agents in preventing EIA support this mecha- ond time. This has been referred to as the refractory peri- nism. od. The duration of this period has been observed by dif- ferent authors to be anywhere from 40 min to up to 3 h. In the other proposed mechanism, airway cooling does Some authors believe ‘refractory period’ is inaccurate not cause bronchoconstriction directly. Rather, it is the since the person is refractory only to exercise but not to rapid rewarming of the airways after exercise that leads to other stimuli such as allergens [15]. Depletion of catechol- bronchoconstriction. Proponents of this mechanism be- amines has been proposed as a mechanism for this phe- lieve that with rapid rewarming, there is a ‘vascular phe- nomenon. When indomethacin, a prostaglandin inhibitor nomenon’ of reactive hyperemia with sudden blood flow is given during an exercise challenge, the refractory period and vascular permeability leading to edema and airway appears to be eliminated implying that prostaglandins obstruction [10]. The greater the difference between the also released during exercise may play a protective role airway temperature during and after exercise, the greater [19]. the severity of obstruction. The observation that inhaling warm air after exercise worsens the bronchoconstriction Diagnosis while cold air lessens it appears to support this theory [12]. In patients with clear-cut symptoms of chest tightness, wheezing, dyspnea or cough after exercise, the history and The observed high prevalence of EIA and bronchial a physical examination while symptomatic can usually hyperresponsiveness (BHR) in elite athletes has led some provide a diagnosis of EIA. In persons with nonspecific investigators to question whether the milder exercise- respiratory symptoms and no history of asthma, diagnos- induced narrowing seen in elite athletes is truly EIA or a tic tests may be necessary. Spirometry should first be done separate condition caused by airway injury from factors to check for airway obstruction as manifested by a de- such as repetitive high-intensity exertion [13]. creased forced expiratory volume in one second (FEV1) and/or peak expiratory flow rate (PEFR). Significant im- EIA generally does not manifest during the exercise provement in pulmonary function after using an inhaled activity but afterwards. In fact, there is bronchodilation ß-agonist confirms the diagnosis of asthma. during exercise and this is most likely due to sympathetic stimulation from catecholamines such as epinephrine [14]. Some persons may complain of symptoms during exercise but studies show that this is often due to preexist- 206 Tan/Spector

Vocal cord dysfunction is an important differential In general, challenge testing should be done when the diagnosis for asthma. It is caused by abnormal adduction subject’s airway responsiveness is as close to baseline as of the vocal cords leading to episodic choking or shortness possible. The subjects’ baseline FEV1 should be at least of breath. A flow-volume loop obtained during spirome- 70% of predicted. Symptomatic asthma or other pulmo- try when a patient has symptoms will show a characteris- nary conditions may be contraindications to the proce- tic flattened inspiratory portion [20]. Direct laryngoscopy dure. Anxiety or anticipation may influence the bronchial can also visualize the characteristic changes in the vocal response so it is important to reassure the patient that cords. there is no ‘correct’ response to the test and that there may be an increase, decrease or no change in lung function. Testing for nonspecific bronchial hyperreactivity by Circadian rhythms may also play a role in the response to bronchial challenge is often used to diagnose asthma if challenge but no specific recommendations on the best spirometry does not demonstrate any airway obstruction time for testing are available at this time. Challenges or reversibility after bronchodilator intake. Spirometry should not be performed after recent allergen exposure, will often be abnormal only during symptomatic periods. exercise, exposure to pollutants and during ongoing or Methacholine is the most commonly used bronchocon- recent viral infections, all of which can increase airway strictor agent used for challenges. A positive methachol- hyperresponsiveness and affect test results. Coffee, choco- ine challenge can confirm the presence of asthma. How- late, cola drinks and smoking should be avoided at least ever, a negative challenge does not rule out EIA, especially 6 h prior to the challenge. Short acting ß-agonist agents, in mild asthmatics with no other triggers. In these pa- short-acting theophylline preparations, ·-adrenergic tients, an exercise challenge may be needed to diagnose agents and cromolyn sodium should be stopped at least EIA [21]. Persons presumed to have EIA but who do not 8 h before testing. Anticholinergic agents, long-acting ß- respond to ß-agonist or cromolyn prophylaxis may also agonists, long-acting theophylline preparations and leuko- need to be reevaluated with an exercise test. Differential triene modifiers should be stopped at least 24 h prior to diagnoses for exercise-associated symptoms include poor the challenge. Antihistamines should be avoided for 48 h conditioning, cardiac disorders and chronic obstructive or longer depending on their half-life. Nedocromil should pulmonary disease. be withheld for 48 h [22]. The American Thoracic Society Guidelines for Methacholine and Exercise Testing do not Methacholine Challenge Testing require withholding of systemic or inhaled steroids prior Bronchoprovocation or bronchial challenge testing to challenges as they do not block the bronchoconstriction with inhaled bronchoconstrictor agents such as metha- induced by methacholine or histamine. Bronchial chal- choline and histamine in a controlled setting and measur- lenges done in patients already on corticosteroids are ing their response by pulmonary function testing is used usually done for research purposes to follow changes in to evaluate nonspecific BHR. This procedure is most BHR. However, systemic and inhaled steroids may de- commonly used to confirm the presence of asthma in crease chronic BHR and may be withheld depending on patients with atypical symptoms such as chronic cough the purpose of the challenge [14, 22]. without wheezing or dyspnea. It is also used to monitor the response of airway inflammation to therapy. Chal- Several standard protocols for bronchial challenges can lenges with exercise, inhaled cold air and inhaled non- be used [23–26]. The ATS Guidelines recommends modi- ionic aerosols can also detect nonspecific BHR. Leuko- fied versions of the (a) two minute tidal breathing proto- trienes, prostaglandins and adenosine are also used as col [26], or (b) the five breath dosing protocol [24]. The bronchoprovocation agents in research studies. latter is more widely used and is described here (table 2). Methacholine and histamine are the two most widely employed bronchoconstrictor agents for challenges. The A baseline spirometry is done prior to the challenge. A responses elicited are usually short-lived and therefore control dose is then administered with 5 breaths of a ideal for testing. Methacholine is generally preferred over saline solution control. A fall of 15% or more in the FEV1 histamine. Histamine has a tendency to produce tachy- following saline inhalation is a contraindication to con- phylaxis with repeated challenges and is associated with tinuing the bronchial challenge. The starting concentra- side effects such as headache and tachycardia. Respon- tion considered safe for methacholine or histamine is siveness to these bronchoconstrictor agents correlates well usually less than 0.1 mg/ml (table 3). The subject is given with asthma severity. 5 breaths of each progressively higher concentration of methacholine or histamine. Spirometry is performed after each set of 5 breaths. The challenge is stopped as soon as Asthma and Exercise 207

Table 2. Sequence of 5 breath dosing protocol for methacholine chal- mal concentration of histamine (10 mg/ml) and metha- lenge [22] choline (25 mg/ml) [24]. 1 Perform baseline spirometry and record baseline vital signs; Exercise Testing FEV1 should be at least 70% of predicted for challenge to pro- Exercise testing can be utilized for detection of nonspe- ceed cific BHR and confirming the diagnosis of asthma but a methacholine challenge is usually easier to perform for 2 Administer 5 breaths of saline control solution; do spirometry at this purpose. A negative methacholine challenge does not 30 and 90 s after the last breath; the best FEV1 after saline inhala- rule out exercise-induced asthma and exercise testing is tion is used as the reference point for PC20FEV1. If FEV1 falls indicated to specifically diagnose bronchoconstriction in- 15% or more from baseline, stop challenge and treat subject with duced by exercise. It is also indicated to determine the nebulized ß-agonist treatment; otherwise, proceed to step 3 severity and response to treatment of known EIA. 3 Administer 5 breaths of most dilute concentration (usually less General Considerations than 0.1 mg/ml); do spirometry 30 and 90 s after last breath; sub- Precautions and required washout of medications are ject may do 2–4 efforts within 3 min generally the same as those for the methacholine chal- lenge. 4 Repeat step 3 with progressively higher concentration of metha- A careful history and physical exam should be done choline; stop challenge when subject FEV1 falls 20% or more, prior to considering exercise testing for suspected asthma. or after highest concentration of methacholine has been given Contraindications to bronchial exercise testing include a (usually 16 mg/ml) history of angina pectoris, proven myocardial infarction or clinically significant valvular disease [29]. An ECG and 5 PC20FEV1 (concentration in mg/ml required to drop subject’s chest X-ray should be done to detect any cardiac abnor- FEV1 by 20% from reference FEV1 after saline control) is com- malities. A stress ECG, and in some situations arterial puted; A PC20FEV1 of 8 mg/ml or less is most widely considered a blood gases may be required, especially in subjects over positive challenge 35 years old with cardiac risk factors such as hyperten- sion, obesity or diabetes [30, 31]. Table 3. Suggested dilution schedule for methacholine using 5- In exercise testing, it is essential that the subject not breath dosimeter protocol [22] exercise for at least 4 h before the challenge. This is impor- tant to avoid the refractory period that can occur for up to Add NaCl Dilution 4 h after exercise in subjects with EIA [32]. ml mg/ml The temperature and humidity of inhaled air are fac- tors that can affect the occurrence and severity of EIA and 100 mg (powdered) 6.25 A 16 these should be controlled during the challenge [33]. To 3 ml of dilution A 9 B4 maximize conditions for induction of EIA, this is best 3 ml of dilution B 9 C1 done by having the subject inhale dry, cool compressed air 3 ml of dilution C 9 D 0.25 that is less than 25°C and with less than 10 mg/l water 3 ml of dilution D 9 E 0.0625 content through a mouthpiece. A less controlled but acceptable alternative is to perform the challenge in an the FEV1 drops by 20% or more. The response to provoca- air-conditioned room with a temperature less than 25° C tion is most commonly measured by the concentration and humidity less than 50% [22]. A noseclip is used to (mg/ml) of methacholine or histamine required to de- prevent nasal breathing, which has been shown to have a crease the subject’s FEV1 by 20% from the baseline saline beneficial effect in EIA by decreasing water loss [34]. control (PC20 FEV1). An alternative measure is the cumu- Mouth breathing alone, which is associated with in- lative dose (in micromoles or breath units) required for a creased water loss, increases the likelihood of EIA. FEV1 fall of 20% (PD20 FEV1). Each breath unit is equiv- alent to one breath of a 1 mg/ml concentration [27]. Exercise Protocol Different protocols have been proposed to standardize A PC20FEV1 of 8 mg/ml or less of methacholine or exercise testing. Although there is no one standardized histamine will be seen in 85–100% of asthmatics and is protocol recommended at this time, the American Tho- generally used as the cut-off value for a positive chal- lenge [23, 28]. Less than 5% of normal patients will have a 20% fall in FEV1 after 5 inhalations of the usual maxi- 208 Tan/Spector

racic Society has formulated guidelines for the different Table 4. General sequence of treadmill exercise challenge (summa- aspects of the exercise challenge [22]. The American rized from American Thoracic Society [22]). Academy of Allergy Bronchopovocation Committee also formulated guidelines in 1979 that are still followed today 1 Baseline spirometry done for preexercise FEV1 [31]. 2 Treadmill speed and grade increased progressively to bring heart A baseline spirometry is performed prior to the chal- rate to 80–90% of maximum age-adjusted predicted value ideally lenge. Various pulmonary function values have been stud- within 2–3 min ied in exercise, including peak expiratory flow rate (PEFR), forced expiratory flow (FEF25–75), forced vital 3 Subject should exercise at target heart rate for at least 4 min. capacity (FVC), forced expiratory volume in 1 s (FEV1) Some authors prefer between 5 and 8 min and specific airway conductance (SGaw) [35]. For exercise testing, FEV1 is the preferred measure of pulmonary func- 4 Testing should be stopped for significant adverse chest symp- tion. The best of three acceptable efforts is used as the toms, severe dyspnea, appearance of new ECG abnormalities, fall baseline value [36]. in blood pressure or any other adverse signs or symptoms The treadmill and cycle ergometer are the most com- 5 After testing stopped, heart rate and ECG should be monitored monly used equipment in exercise protocols with each for at least 3 additional minutes having both advantages and disadvantages. Bronchocon- striction is easier to provoke with the treadmill due to a 6 Spirometry is performed 5, 10, 15, 20 and 30 min postexercise faster increase in minute ventilation [37]. Walking or run- ning on the treadmill may be easier for some subjects hav- 7 A fall of 10% from the pre-exercise FEV1 is generally considered ing difficulty with the coordination needed for cycling. abnormal; a 15% drop is required for diagnosis by some practi- However, the work rate on the treadmill can be harder to tioners accurately measure because of variables such as the sub- ject’s weight [38]. In contrast, the work rate achieved on 8 Continue checking spirometry even if FEV1 fall is diagnostic the cycle ergometer is affected by fewer variables and can before 30 min in order to determine nadir or lowest FEV1 which easily be determined accurately [22]. Cycling is the pre- correlates with EIA severity ferred alternative for subjects who are unable to walk briskly or run due to problems such as weakness or arthri- 9 Serial post-exercise spirometry may be stopped if 2 consecutive tis. Another advantage of the cycle ergometer is the lack of FEV1 values after the nadir show improvement requirement to adjust variables such as speed and incline. Many investigators feel that although treadmill running 10 Administer inhaled ß-agonist treatment as needed to bring FEV1 appears to elicit more bronchoconstriction, the cycle back to within 10% of pre-exercise value ergometer is easier to use and can identify most patients with EIA readily if done properly [39]. The amount of exercise-induced work or stress that will be used in the challenge can be quantified in terms of Treadmill Testing heart rate and/or oxygen consumption (ml/min) if the lat- The treadmill is the most commonly used equipment ter is measured during the challenge. Obviously, factors for performing the exercise challenge (table 4). Electrocar- such as the subject’s age, weight and level of aerobic fit- diographic monitoring with an oscilloscope or video ness will affect these values. In general, however, a work screen for cardiac rhythm and heart rate is usually inte- level that increases the heart rate to more than 90% or grated into most treadmill equipment used for clinical more of the maximum predicted values for age or the oxy- testing. Cardiac rhythm should be monitored continuous- gen consumption to 30–40 ml/kg are considered an ade- ly throughout the test in real time. Care must be taken to quate stimulus for EIA [31]. Other protocols consider a tape down the ECG electrodes to keep them from falling target heart rate of 80% of the maximum predicted ade- off sweaty skin. Areas of the chest where the electrodes are quate [22]. Continuous monitoring of the subject’s heart placed may be shaved to facilitate attachment of elec- rate rather than ventilation is preferred and more practi- trodes. Pulse oximetry may also be measured. Blood pres- cal. The subject’s maximum predicted heart rate is deter- sure should be periodically checked during and after the mined from tables based on age [40] (table 5). The target challenge preferably with automatically inflating arm heart rate for the subject during the challenge is 80–90% cuffs. of the maximum predicted. The target heart rate can be reached in different ways. Protocols have ranged from those that aim to raise the heart rate incrementally over 15–25 min to those that Asthma and Exercise 209

raise it rapidly to the target in 1–2 min [30]. A stepped refractory period while extended exercise at the target approach that increases work in 3 stages is tolerated by heart rate may actually diminish bronchoconstriction [22, most age groups and is commonly followed [31]. Whatev- 41]. However, many authors prefer protocols, including er the protocol used, the ATS recommends that the target the stepped protocol often used in research studies, with a heart rate should be reached in 2–3 min and maintained longer total duration that has a warm-up period of 4 min at a steady state for at least 4 min with the total duration and a steady state period of 5–8 min [30]. This preference of the challenge at 6–8 min. The rationale for this recom- is based on the observation that at least 5 min of exercise mendation is that a long warm-up period can induce a at the target heart rate is needed in many subjects, espe- cially athletic fit ones, to provoke EIA. Table 5. Average age-specific heart rates The speed and incline (grade) of the treadmill are Age Maximum Maximum heart rate/min 80% 75% raised to produce progressive increases in the heart rate. years heart rate/ 95% 90% 85% Table 6 shows a suggested stepwise guide with the speed min and incline expected to induce the desired heart rate at each stage. Nomograms based on studies of previous chal- 5 206 196 185 175 165 155 lenges are helpful in planning the treadmill settings for 10 202 192 182 172 162 152 speed and incline before starting the challenge (fig. 1). The 15 198 188 178 168 158 149 speed is based on subject height while the incline is based 20 194 184 175 165 155 146 on age. 25 191 181 172 162 153 143 30 187 178 168 159 150 140 The subject should have comfortable clothing and run- 35 183 174 165 156 146 137 ning shoes on for the procedure. The subject is instructed 40 180 171 162 153 144 135 to stand on the treadmill and hold on lightly to the rails as 45 176 167 158 150 141 132 the treadmill starts to move. He or she then takes long, 50 172 163 155 146 138 129 slow steps and lets go of the railing when a comfortable 55 169 161 152 144 135 127 balance and rhythm is achieved. The speed and incline 60 165 157 149 140 132 124 are then increased until the subject is running or walking 65 161 153 145 137 129 121 fast. Subject should be instructed not to hold on to the 70 157 149 141 133 126 118 railing while running since this will lessen the work 75 154 146 139 131 123 116 effort. 80 150 143 135 128 120 113 85 146 139 131 124 117 110 The challenge should be stopped if there any cardio- 90 143 136 129 122 114 107 vascular or severe asthma symptoms develop. A physician or technician trained in cardiopulmonary resuscitation Maximum heart rate (y) derived from linear regression equation should administer the challenge with emergency equip- y = 209.2 – 0.74x, where x is age in years. Note that predicted maxi- ment readily available. Indications for terminating the mum heart rate has a standard deviation of approximately challenge include excessive fatigue or dyspnea, a drop in B10 bpm. From Johnson and Buskirk [40]. blood pressure, angina-like symptoms with or without ECG changes, ataxia, premature ventricular beats (more Table 6. Stepped protocol with suggested speed and incline settings [31] Step Duration Target heart rate Treadmill incline min rate I2 50% predicted maximum* 2.5 mph 0% II 2 70% predicted maximum 2/3 target conditions by height** determined by age** III 5–8 90% predicted maximum target conditions by height determined by age * Predicted maximum age-adjusted heart rate from table 7. ** Target exercise of treadmill rate and incline from figure 1. 210 Tan/Spector

than 25% of beats or 10/min) or ventricular tachycardia, widening of the QRS complex, supraventricular tachycar- dia and rate-dependent heart block [29]. After exercise, the heart rate and ECG should be moni- tored for up to 3 min [30]. The subject’s FEV1 is measured in the sitting position at 5, 10, 15, 20 and 30 min postexer- cise. A fall of 10% from the preexercise value is consid- ered abnormal while some investigators require a 15% drop for a diagnosis of EIA [22]. As long as the patient is stable, spirometry measurements should be continued even after the FEV1 has dropped by 15% in order to deter- mine the nadir or lowest point and assess the severity of EIA. If 2 subsequent FEV1 measurements after the nadir show improvement, the measurements may be stopped. A nebulized ß-agonist treatment can be administered if the subject’s FEV1 does not return spontaneously to within 10% of the preexercise value [22]. Cycle Ergometer Fig.1. Treadmill settings for speed and incline. From Eggleston et al. The cycle ergometer is easier for many patients to use, [31]. especially young children and older subjects. It has been generally considered less ‘asthmagenic’ than the treadmill pressed gas mixture containing 5% carbon dioxide can [30] but many investigators now consider it just as effec- also be used [43]. The aim is usually to achieve approxi- tive in bronchial provocation when the target workload is mately 60–70% of maximum voluntary ventilation. Dos- achieved [38]. ing schedules differing in target minute ventilation and Monitoring of vital signs is easier with cycling since the duration have been recommended [44]. A fall of 10% in subject is moving less. A target heart rate or ventilation is the FEV1 is also considered suggestive of asthma while set prior to the challenge in the same manner as with the other authors require a 15% fall for diagnosis [16]. treadmill. A target work rate that will induce the target heart rate or ventilation can then be calculated and used Comparison of Bronchial Provocation Challenges to adjust the ergometer. As with treadmill testing, the ATS Challenges with bronchoconstrictor agents such as me- recommends that the target work intensity as measured thacholine and histamine produce direct constriction of by heart rate or ventilation should be sustained for at least bronchial smooth muscle in susceptible subjects with 4 min [22] while other practitioners prefer to maintain it BHR. On the other hand, exercise and hyperventilation for 5–8 min. challenges induce bronchospasm in asthmatics by a se- quence of inflammatory events. Since other mediators Hyperventilation such as leukotrienes are likely released in EIA, a negative The amount of ventilation is the ultimate determinant methacholine or histamine challenge does not rule out leading to EIA and this can be achieved by either exercise EIA. When utilized for the detection of BHR in a study of or hyperventilation [42]. Eucapnic voluntary hyperventi- lation (EVH) is therefore a valid alternative test for EIA. It requires more equipment but is ideal for patients who are unable to perform physical exercise. The subject breathes through a valve box with a pump set to a specific minute ventilation. A reservoir balloon is attached to the valve box and the patient breathes with enough force to prevent deflation of the balloon. The subject inhales a mixture of cooled room air with a low concentration of carbon dioxide to prevent hypocapnia which has a bron- choconstrictive effect. A simplified system using a com- Asthma and Exercise 211

Table 7. Nonpharmacologic measures for EIA Good conditioning and aerobic fitness is also impor- tant in lessening EIA. Asthmatics should not be discour- Warm up for at least 10 min before actual exercise aged from exercising. Persons who exercise regularly do Cover mouth and nose with scarf or surgical mask during cold not have the rapid and abrupt increases in minute ventila- tion that are more likely to stimulate EIA. weather Exercise in warm, humidified environments if possible Whether or not physical training improves BHR and Warm down or gradually lower the intensity of exercise pulmonary function remains unclear. While some studies report no improvement in EIA severity with training pro- patients with unexplained dyspnea, the exercise challenge grams, there are several studies that do [48–50]. In one was of low clinical utility when compared to the metha- study, a 25% improvement in the severity of EIA was choline challenge [45]. observed with a training program [49]. Most authors believe that a regular regimen of moderate exercise tai- The treadmill and cycle ergometer are probably both lored to a patient’s asthma severity should be prescribed equally effective in provoking EIA when performed prop- for its physical, social and emotional benefits [51]. Walk- erly. Isocapnic ventilation is the alternative if exercise ing or jogging are good initial activities. Swimming is testing cannot be done. More comparative studies need to often recommended as an ideal sport for asthmatics as the be done to determine the comparative usefulness of these warm, humidified air around the pool lowers the occur- different modes in EIA diagnosis. rence of EIA. Exercise challenges using minimum or no equipment Sports and EIA have been studied for screening large populations such as Athletes in general have a 35% incidence of EIA [46]. students and schoolchildren. Free range running can be A survey of US athletes in the 1998 Winter Olympics sur- used but safety concerns limit its use [46]. prisingly showed that 43% had a previous diagnosis of asthma [52]. Early recognition of patients at risk for EIA Therapy is a first step towards diagnosis and preventive treatment. In a study involving 238 male high school varsity football Prevention is the primary mode of therapy for EIA. players, a significant rate of previously undiagnosed EIA There are non-pharmacologic and pharmacologic mea- was observed and associated with risk factors including a sures that can be taken to prevent or lessen the symptoms history of wheezing, residence in a poverty area, race and of EIA. remote history of asthma. This study suggested that a sim- ple screening 1 mile outdoor run may identify individuals Nonpharmacologic Therapy (table 7) at risk [46]. Important factors affecting the intensity of EIA are the The amount of ventilation in exercise is directly relat- person’s underlying baseline bronchial reactivity and lev- ed to the intensity of the activity. It is the increase in ven- el of conditioning as well as the amount of ventilation and tilation produced by exercise and not the kind of exercise air temperature during exercise. The first two factors that is crucial in EIA. This means that any exercise can depend on previous levels of asthma control and general lead to EIA if it is done hard enough or long enough to fitness while the latter two depend on the nature and envi- increase the amount of air being inhaled. In fact, hyper- ronment of the exertion. ventilation maneuvers can be used in some instances An asthmatic’s bronchial reactivity is increased with instead of an exercise challenge to diagnose EIA. The repeated exposure to allergens such as animal dander, intensity of the exercise is important because it is directly pollen, dust mites and mold. Recurrent viral infections proportional to the amount of ventilation. Vigorous activ- can also lead to increased bronchial reactivity. Avoidance ities such as basketball or soccer can cause more severe measures and environmental control to minimize allergen attacks than less vigorous ones like baseball. exposure is essential for good asthma control. Bronchial Cool and dry air worsens airway cooling causing more hyperresponsiveness and the response to exercise can be symptoms. Running in warm days cause less problems diminished by maintenance anti-inflammatory therapy than the same activity during cold days. Figure skaters with medications such as inhaled steroids and antileuko- and hockey players who have asthma are more prone to trienes [47]. EIA in their particular sports because of the cold environ- ment. The incidence of exercise-induced bronchospasm 212 Tan/Spector

in competitive figure skaters has been found in several Table 8. Pharmacologic agents studies to be up to 30–35% [53, 54]. One study suggests that it may be helpful to screen for EIA in aspiring figure First-line skaters so that education and preventive measures can be Inhaled ß-agonists done [54]. Covering the nose and mouth with a scarf or Inhaled cromolyn and nedocromil surgical mask when exercising in cold weather has been found to lessen EIA by warming inhaled air [55]. Swim- Additional agents ming has been recommended to many asthmatics because Antileukotriene agents the warm, humidified air around the swimming pool Calcium channel blockers allows more active exertion without triggering asthma. Oral ß-agonists ·-Agonists As mentioned earlier, a refractory period occurs up to Theophylline several hours after recovery from EIA. Athletes with asth- Inhaled anticholinergic agents ma can take advantage of this by always warming up prior to vigorous exertion. This warm-up period induces a (e.g. ipatropium bromide) refractory period during which there are less asthma symptoms with the actual exercise [56]. A warm-up peri- Newer agents od of 10 min is usually adequate [57]. Runners can warm Inhaled heparin up by running repeated short sprints [58]. Gradual in- Inhaled furosemide crease in intensity of exercise as opposed to rapid in- creases is also beneficial [11]. Warming down or slowly Table 9. Asthma drugs in the Olympics lessening the intensity of exercise instead of stopping [from The United States Anti-Doping Agen- abruptly can also make EIA less severe. This may make cy (USADA) Guide to Prohibited Classes of airway rewarming and the resultant vascular dilation and Substances and Prohibited Methods of Dop- edema more gradual and less intense [10]. ing. Updated December 2000] Education on the nature of EIA and how to control it Allowed with or without medications needs to be given to asthmat- Albuterol* ics as well as family members and coaches. Today, there is Terbutaline* no reason for asthma to be a restricting factor in sports. Salmeterol* Many well-known athletes have asthma, including Dennis Salmeterol/ipratropium* Rodman and Jackie Joyner-Kersee. Nancy Hogshead, the Cromolyn/nedocromil Olympic swimmer gives the athlete’s perspective on asth- Aminophylline/theophylline ma in her book, Asthma and Exercise [59]. Ipratropium Antileukotrienes Pharmacologic Therapy (table 8) All inhaled and nasal corticosteroids Inhaled ß-agonists are the medications of choice for prophylaxis of EIA. If given 15 min to an hour before Banned exercise, short-acting ß-agonists such as albuterol (Vento- All inhaled ß-agonists not listed above linTM) and terbutaline (BrethaireTM) can prevent symp- All oral and injectable ß-agonists toms. Long-acting inhaled ß-agonists such as salmeterol All systemic corticosteroids (SereventTM) have a slower onset of action, act for 12 h at a time and should be taken at least 1–2 h before exercise * Written notification of asthma and/or [60]. Formoterol (ForadilTM), a new rapid-acting ß-ago- exercise-induced asthma by a respiratory or nist with a long duration of action (up to 12 h) may be team physician is necessary and must be pro- ideal for use just before exercise. Albuterol, terbutaline, vided to USADA and the Relevant Medical salmeterol and salmeterol/ipratropium (CombiventTM) Authority prior to competition. are the only inhaled ß-agonists allowed by the US and International Olympic Committees (table 9). The Nation- The next most commonly used prophylactic agents for al Collegiate Athletic Association (NCAA), on the other EIA are the inhaled mast cell stabilizers cromolyn and hand, allows all inhaled ß-agonists [21]. nedocromil. When used 10–20 min before exercise, they have been found to block or attenuate EIA. Cromolyn and nedocromil are approved for use by the US and Interna- tional Olympic committees. When compared with in- Asthma and Exercise 213

haled ß-agonists, cromolyn is not as effective in prevent- mechanisms for bronchoconstriction include (a) mucosal ing EIA [61]. drying and increased osmolarity stimulating mast cell degranulation, and (b) rapid airway rewarming after exer- The antileukotrienes are currently recommended cise causing vascular congestion, increased permeability mainly for chronic asthma therapy but have both been and edema leading to obstruction. EIA symptoms start shown to diminish bronchospasm in exercise challenge after exercise, peak 8–15 min postexercise and sponta- studies [62]. The two types of antileukotrienes are leuko- neously resolve in about 60 min. A refractory period lasts triene receptor antagonists, e.g. zafirlukast (AccolateTM) up to 3 h after recovery, during which repeat exercise or montelukast (SingulairTM), and 5-lipoxygenase inhibi- causes less bronchospasm. The amount of ventilation and tors (e.g. zileuton (ZyfloTM). In one study, standard single the temperature of inspired air are important factors in doses of montelukast, zafirlukast and zileuton were as determining the severity of EIA. Greater ventilation and effective as salmeterol in attenuating EIA. Zileuton had a cold,dry air increase the risk for EIA. shorter duration of action (up to 4 h) while montelukast, zafirlukast and salmeterol were protective for up to 8 h Methacholine and histamine challenge testing may be [63]. When compared in the long-term treatment of EIA done to detect nonspecific BHR to confirm suspected over 8 weeks, montelukast and salmeterol were both asthma but a negative response does not rule out EIA. found to be effective but salmeterol lost much of its pro- Exercise or hyperventilation testing is needed to rule out tective effect after 4 and 8 weeks of use [64]. EIA. The treadmill and cycle ergometer are the preferred modes of exercise testing. Both have advantages and dis- Other agents which may be added include inhaled anti- advantages but are appropriate diagnostic tools for EIA cholinergic agents, theophylline, calcium channel block- diagnosis when performed properly. Hyperventilation ers, antihistamines, ·-agonists and oral ß-agonists [65]. testing is an alternative for those unable to do exercise These agents have shown varying protection against EIA testing. in different studies and may not be as effective if used alone. Prevention is the main objective in managing EIA. Nonpharmacologic measures include warming up prior to The widely used anticoagulant heparin,when given in vigorous exertion, covering the mouth and nose in cold the inhaled form, has been shown in at least one study to weather, exercising in warm, humidified environments if be more effective than cromolyn in preventing EIA. Its possible and warming down after exercise. Aerobic fitness mechanism of action is not clear but is thought to be due and good control of baseline bronchial reactivity also help to inhibition of mast cell release [66]. Another widely used diminish the effects of EIA. drug, the diuretic furosemide, when given in the inhaled form, has been shown in several studies to inhibit EIA in Inhaled ß-agonists are the medications of choice for children and adults [67, 68]. Further studies in larger EIA prophylaxis. Inhaled cromolyn or nedocromil may groups are needed to confirm the effectiveness in prevent- also be used. Antileukotriene agents have been shown to ing EIA. diminish EIA after single doses and with chronic usage without loss of protection. If inhaled ß-agonists or cromo- Pharmacologic treatment after the symptoms of EIA lyn are not adequate for prophylaxis, additional agents have started is identical to treatment of asthma symptoms can be considered including antileukotrienes, anticholin- from other triggers. Appropriate anti-inflammatory treat- ergic agents [such as ipatropium bromide (AtroventTM)], ment of underlying chronic persistent asthma with in- theophylline, calcium channel blockers, a-agonists, anti- haled steroids and other medications will decrease the histamines and oral ß-agonists. Newer agents under study incidence of exercise-induced bronchospasm. include inhaled heparin and inhaled furosemide. Conclusions Education regarding the nature and management of EIA is important not only for asthmatics but also their Exercise-induced asthma (EIA) is characterized by families and coaches. With the proper precautions and transient airway obstruction occurring after strenuous workout techniques, there is no limit to what persons with exertion. A fall of 15% or more in the FEV1 after exercise asthma can achieve in sports. is diagnostic while a 10% drop is already considered abnormal. Inhalation of large volumes of dry, cold air dur- ing exercise leads to loss of heat and water from the bron- chial mucosa and airway cooling and drying. Proposed 214 Tan/Spector

References 1 Bardagi S, Agudo A, Gonzalez CA, Romero 19 O’Byrne PM, Jones JL: Effect of indomethacin 36 American Thoracic Society: Standardization of PV: Prevalence of exercise-induced narrowing on exercise-induced bronchoconstriction and spirometry: 1994 update. Am J Resp Crit Care in schoolchildren from a Mediterranean town. refractoriness after exercise. Am Rev Respir Med 1995;152:1107–1136. Am Rev Resp Dis 1993;147:1112–1115. Dis 1986;134:69–72. 37 Anderson SD, Connolly NM, Godfrey S: Com- 2 Karjalainen J: Exercise response in 404 young 20 McFadden ER, Zawadski DK: Vocal cord dys- parison of bronchoconstriction induced by cy- men with asthma: No evidence for a late asth- function masquerading as exercise induced cling and running. Thorax 1971;26:396–401. matic reaction. Thorax 1991;46:100–104. asthma: A physiologic cause for choking during athletic activities. Am J Resp Crit Care Med 38 Souhrada JF, Souhrada M: Provocative chal- 3 Milgrom H, Taussig LM: Keeping children 1996;153:942–947. lenge by exercise or hyperventilation; in Spec- with exercise-induced asthma active. Pediat- tor SL (ed): Provocation Testing in Clinical rics 1999;104:38. 21 Nastasi KJ, Heinly TL, Blaiss MS: Exercise- Practice. New York, Marcel Dekker, 1995, pp induced asthma and the athlete. J Asthma 425–449. 4 Helenius IJ, Tikkanen HO, Haahtela T: Occur- 1995;32:249–257. ence of exercise induced bronchospasm in elite 39 McFadden ER: Exercise-induced airway nar- runners: Dependence on atopy and exposure to 22 American Thoracic Society: Guidelines for me- rowing; in Middleton E (ed): Allergy: Principles cold air and pollen. Br J Sports Med 1998;32: thacholine and exercise challenge testing 1999. and Practice, ed 5. Mosby, New York, 1998, pp 125–129. Am J Respir Crit Care Med 2000;141:309– 953–962. 329. 5 Afrasiabi R, Spector SL: Exercise-induced asth- 40 Johnson WR, Buskirk ER (eds): Science and ma: It needn’t sideline your patients. Phys 23 Cockcroft DW, Killian DN, Mellon JJ, Har- Medicine of Exercise and Sports, ed 2. New Sportsmed 1991;19:49–62. greave FE: Bronchial reactivity to inhaled his- York, Harper & Row, 1974, p 125. tamine: A method and clinical survey. Clin 6 Strauss RH, McFadden ER Jr, Ingram RH Jr, Allergy 1977;7:235. 41 Silverman M, Anderson SD: Standardization Jaeger JJ: Enhancement of exercise-induced of exercise tests in asthmatic children. Arch Dis asthma by cold air. N Engl J Med 1977;297: 24 Chai H, Farr RS, Froehlich LA, Mathison DA, Child 1972;47:882–889. 743–747. McLean JA, Rosenthal RR, Sheffer AL, Spec- tor SL, Townley RG: Standardization of bron- 42 Zeballos RJ, Shturman-Ellstein R, McNally JF 7 Deal EC Jr, McFadden ER Jr, Ingram RH Jr, chial inhalation challenge procedures. J Allergy Jr, Hirsch JE, Souhrada JF: The role of hyper- Strauss RH, Jaeger JJ: Role of respiratory heat Clin Immunol 1975;56:323–327. ventilation in exercise-induced bronchocon- exchange in production of exercise-induced striction. Am Rev Resp Dis 1978;118:877– asthma. J Appl Physiol 1979;46:467–475. 25 Yan K, Salome CM, Woolcock AJ: Rapid 884. method for measurement of bronchial respon- 8 Anderson SD, Schoeffel RE, Black JL, Da- siveness. Thorax 1983;38:760–765. 43 Phillips YY, Jaeger JJ, Laube BL, Rosenthal viskas E: Airway cooling as the stimulus to RR: Eucapnic voluntary hyperventilation of exercise-induced asthma: A reevaluation. Eur J 26 Juniper EF, Cockcroft DW, Hargreave FE: Ti- compressed gas mixture. Am Rev Resp Dis Respir Dis 1985;67:20–30. dal breathing method. In Histamine and Me- 1985;131:31–35. thachioline Inhalation Tests: Laboratory Pro- 9 Sheppard D, Eschenbacher WL: Respiratory cedure and Standardization, ed. 2. Lund, Astra 44 Argyros GJ, Roach JM, Hurwitz KM, Eliasson water loss as a stimulus to exercise-induced Draco AB, Lund, 1994. AH, Phillips YY: Eucapnic voluntary hyper- bronchoconstriction asthma. J Allergy Clin im- ventilation as a bronchoprovocation technique: munol 1984;73:640–642. 27 Fish JE: Bronchial challenge testing; in Middle- Development of a standardized dosing sched- ton E Jr, et al (eds): Allergy: Principles and ule in asthmatics. Chest 1996;109:1520–1524. 10 McFadden ER Jr: Exercise-induced airway ob- Practice, ed. 4. St Louis, Mosby, 1993, pp 613– struction. Clin Chest Med 1995;16:671–682. 627. 45 Zeballos RJ, Weisman IM, Connery SM, Brad- ley JP: Standard treadmill vs incremental cycle 11 Spector SL: Update on exercise-induced asth- 28 Hopp RJ, Bewtra AK, Nair NM et al: Specifici- ergometry in the evaluation of airway hyper- ma. Ann Allergy 1993;71:571–577. ty and sensitivity of methacholine inhalation reactivity in unexplained dyspnea. Am J Resp challenge in normal and asthmatic children. J Crit Care Med 1999;159:A419. 12 McFadden ER Jr, Lenner KA, Strohl KP: Post- Allergy Clin Immunol 1984;74:154. exertional airway rewarming and thermally- 46 Kukafka DS, Lang DM, Porer S, Rogers J, Cic- induced asthma. J Clin Invest 1986;78:18–25. 29 Trautlein JJ: Risk factors in exercise testing. J colella D, Polansky M, D’Alonzo GE Jr: Exer- Allergy Clin Immunol 1979;64:625–626. cise-induced bronchospasm in high school ath- 13 Anderson SD, Holzer K: Exercise-induced letes via a free running test: Incidence and epi- asthma: Is it the right diagnosis in elite ath- 30 Cropp GJA: The exercise provocation test: demiology. Chest 1998;114:1613–1622. letes? J Allergy Clin Immunol 2000;106:419– Standardization of procedures and evaluation 428. of response. JACI 1979;64:627–633. 47 Freezer NJ, Croasdell H, Doull IJ, Holgate ST: Effect of regular inhaled beclomethasone on 14 Spector SL: Allergen inhalation challenges; in 31 Eggleston PA, Rosenthal RR, Anderson SD, exercise and metacholine airway responses in Spector SL (ed): Provocative Testing in Clini- Anderson R, et al: Guidelines for the methodol- schoolchildren with recurrent wheeze. Eur cal Practice. New York, Marcel Dekker, 1995, ogy of exercise challenge testing of asthmatics. Resp J 1995;8:1488–1493. pp 325–368. JACI 1979;64:642–645. 48 Svenonius E, Kautto R, Arborelius M: Im- 15 McFadden ER Jr, Gilbert IA: Exercise-induced 32 Edmunds AT, Tooley M, Godfrey S: The re- provement after training of children with exer- asthma. N Engl J Med 1994;330:1362–1367. fractory period after exercise-induced asthma: cise-induced asthma. Acta Paediatr Scand Its duration and relation to severity of exercise. 1983;72:23–30. 16 Godfrey S: Bronchial challenge by exercise or Am Rev Resp Dis 1978;117:247–254. hyperventilation; in Spector SL (ed): Provoca- 49 Henriksen JM, Nielsen TT: Effect of physical tive Challenge Procedures: Background and 33 Anderson SD, Schoeffel RE, Follet R, et al: training on exercise-induced bronchoconstric- Methodology. Mount Kisco, Futura, 1989, pp Sensitivity to heat and water loss at rest and tion. Acta Paediatr Scand 1983;72:31–36. 365–394. during exercise in asthmatic patients. Eur J Resp Dis 1982;63:459–471. 50 Haas F, Pasierski S, Levine N, et al: Effect of 17 Boulet LP, Legris C, Turcotte H, et al: Preva- aerobic training on forced expiratory airflow in lence and characteristics of late asthmatic re- 34 Shturman-Elistein R, Zeballos RJ, Buckley JM, exercising asthmatic humans. J Appl Physiol sponse to exercise. J Allergy Clin Immunol et al: The beneficial effect of nasal breathing on 1987;63:1230–1235. 1987;80:655. exercise-induced bronchoconstriction. Am Rev Resp Dis 1978;118:65–73. 51 Satta A: Exercise training in asthma. J Sports 18 Peroni DG, Boner AL: Exercise-induced asth- Med Phys Fitness 2000;40:277–283. ma: Is there space for late-phase reactions? Eur 35 Cropp GJA: Relative sensitivity of different Resp J 1996;9:1335–1338. pulmonary function tests in the evaluation of exercise-induced asthma. Pediatrics 1975;56 (suppl):860. Asthma and Exercise 215

52 Weiler JM, Ryan EJ: Asthma in United States 60 Schaanning J, Vilsik J, Henriksen AH, Bratten 65 Spector SL: Treatments of exercise-induced Olympic athletes who participated in the 1998 G: Efficacy and duration of salmeterol powder asthma other than B-agonists; in Weiler J (ed): Olympic Winter Games. J Allergy Clin Immu- inhalation in protecting against exercise-in- Allergic and Respiratory Diseases in Sports nol 2000;106:267–271. duced bronchoconstriction. Ann Allergy Asth- Medicine. New York, Marcel Dekker, 1997. ma Immunol 1996;76:57–60. 53 Provost-Craig MA, Arbour KS, Sestili DC, 66 Garrigo J, Danta I, Ahmed T: Time course of Chabalko JJ, Ekinci E: The incidence of exer- 61 Rohr AS, Siegel SC, Katz RM, et al: A compari- the protective effect of inhaled heparin on exer- cise-induced bronchospasm in competitive fig- son of inhaled albuterol and cromolyn in the cise-induced asthma. Am J Resp Crit Care ure skaters. J Asthma 1996;33:67–71. prophylaxis of exercise-induced broncho- Med 1996;153:1702–1707. spasm. Ann Allergy 1987;59:107–109. 54 Mannix ET, Farber MO, Palange P, Galassetti 67 Novembre E, Frongia G, Lombardi E, Rest M, P, Manfredi R: Exercise-induced asthma in fig- 62 Makker HK, Lau LC, Thomson HW, Binks et al: The preventive effects and duration of ure skaters. Chest 1996;109:312–315. SM, Holgate ST: The protective effect of in- action of two doses of inhaled furosemide on haled LTD4 receptor antagonist ICI 204219 exercise-induced asthma in children. J Allergy 55 Schacter EN, Lach E, Lee M: The protective against exercise-induced asthma. Am Rev Res- Clin Immunol 1995;96:906–909. effect of a cold weather mask on exercise- pir Dis 1993;147:1413–1418. induced asthma. Ann Allergy 1981;46:12–16. 68 Munyard P, Chung KF, Bush A: Inhaled furo- 63 Coreno A, Skowronski M, Kotaru C, McFad- semide and exercise-induced bronchoconstric- 56 Reiff DB, Nozhat BC, Neil BP, Philip WI: The den ER: Comparative effects of along-acting tion in children with asthma. Thorax 1995;50: effect of prolonged submaximal warm-up exer- beta-2-agonists, leukotriene receptor antago- 677–679. cise on exercise-induced asthma. Am Rev Res- nists, and a 5-lipoxygenase inhibitor on EIA. pir Dis 1989;139:479–484. JACI 2000;106:500–506. Ricardo A. Tan, MD California Allergy and Asthma 57 Wright LA, Martin RJ: Nocturnal asthma and 64 Villaran C, O’Neill SJ, Helbling A, van Noord Medical Group, Inc. exercise-induced bronchospasm. Postgrad Med JA: Montelukast versus salmeterol in patients 11645 Wilshire Blvd., Suite 1090 1995;97:83–90. with asthma and exercise-induced broncho- Palmdale, CA 90025 (USA) constriction. Montelukast/Salmeterol Exercise Tel. +1 310 966 9022, Fax +1 310 966 9042 58 Schnall RP, Landau LI: Protective effects of Study Group. J Allergy Clin Immunol 1999; E-Mail [email protected] repeated short sprints in exercise-induced asth- 104:547–553. ma. Thorax 1980;35:828–832. 59 Hogshead N, Couzens GS: Asthma and Exer- cise. New York, Henry Holt, 1990. 216 Tan/Spector

Weisman IM, Zeballos RJ (eds): Clinical Exercise Testing. Prog Respir Res. Basel, Karger, 2002, vol 32, pp 217–230 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Evaluation of Impairment and Disability: The Role of Cardiopulmonary Exercise Testing Darryl Y. Sue UCLA School of Medicine, Los Angeles, Calif., and Medical-Respiratory Intensive Care Unit and Associate Chair, Department of Medicine, Harbor-UCLA Medical Center, Torrance, Calif., USA Summary heart transplantation, and many clinicians depend on exercise testing to determine postoperative risk in pa- Cardiopulmonary exercise testing has a strong physiological tients with marginal cardiopulmonary function. Finally, a basis and well-documented clinical rationale for impairment primary role for exercise testing remains in the evaluation evaluation. Both determination and quantitation of impairment of patients with unexplained dyspnea or exercise intoler- are enhanced by exercise testing, notably because of the inade- ance. In such patients, common (ischemic heart disease, quacy of resting pulmonary function tests to predict exercise ventilatory limitation), unusual disorders (mitochondrial capacity, and both over- and underestimation of work capacity myopathies), and combined disorders may be suggested have been found. Furthermore, exercise arterial blood gases or confirmed. are very sensitive tests for subtle lung disease. Exercise testing can be especially useful in those with co-existing disorders or The determination and quantification of impairment, unsuspected cardiovascular disorders. Exercise testing is not not withstanding disability, can be complex, difficult, or indicated in those who lack complaints of exertional dyspnea contested, and objective measurements of functional ca- or fatigue, and those with severe respiratory impairment (ATS pacity and its loss should be valuable adjuncts to other criteria for impairment by FEV1 or DLCO) also do not require clinical information. Although it would appear that com- exercise testing. Cardiopulmonary exercise testing can provide plaints about functional or work capacity could be ideally objective determination of exercise capacity, increased sensi- evaluated during actual exercise, integrative cardiopul- tivity for pulmonary gas exchange abnormalities, and the ability monary exercise testing has been characterized as expen- to identify unsuspected or unanticipated non-pulmonary sive, unavailable, and difficult to perform [1–2]. Nev- causes of impairment. ertheless, there are considerable data supporting the use of cardiopulmonary exercise testing in impairment evalu- Introduction ation, much of which demonstrate that estimation of exer- cise capacity from clinical and physiological measure- Cardiopulmonary exercise testing has become well es- ments at rest is poor at best. tablished as a clinical tool for evaluating patients in wide variety of clinical situations. For severe congestive heart Not all assessment of impairment relates to occupa- failure, measurement of peak oxygen uptake (V˙ O2) is the tional exposures, and not all determinations of disability best discriminant for selection of patients to undergo are questions about the ability to perform a job [1]. How- ever, the focus of this discussion is on determining impairment in patients with occupational disease. A stan- dardized approach to all patients or claimants is impracti- cal, even those with potential occupational diseases from

Table1. Cardiopulmonary exercise testing: clinical questions compared to impairment/disability Clinical question Impairment/disability question Physiological variables Is maximum exercise capacity reduced What is the maximum exercise capacity? Peak V˙ O2 compared to normal and compared to normal? expressed as ml/kg/min What is the maximum sustainable level of V˙ O2 at the lactic acidosis threshold; work Does impaired O2 flow limit exercise exercise? rate at the lactic acidosis threshold capacity at submaximal work rates? How much reduction in ventilatory capacity Breathing reserve Does reduced ventilatory capacity limit is seen at maximum exercise? maximum exercise? Arterial blood gases (PaO2), alveolar-arterial Is there evidence of subtle pulmonary gas PO2 difference, dead space/tidal volume How severe are pulmonary gas exchange exchange abnormalities that would indicate ratio abnormalities during exercise? the presence of unsuspected lung disease? Heart rate and heart rate reserve, blood Is exercise limited by cardiovascular factors, Are there coexisting conditions that limit lactate during recovery, breathing reserve effort, or musculoskeletal problems? exercise outside of the lungs? Estimate or measurement of task specific Can the subject perform exercise at a level requirements (V˙ O2) and duration comparable to that of a specific job or occupation? exposure to the same injurious substance. For example, about all systems involved with exercise, is intrinsically physicians sometimes seek to link a toxin or injurious sub- quantitative, and provides measurements of both de- stance to a disease in a causal sense. In others, the injury is mand and capacity. established already but the extent of impairment or loss of capacity is in question. A variety of strategies, ranging The integrative cardiopulmonary exercise test has two from epidemiological to medical evaluation of an individ- major components. There is an exercise stress given to the ual patient, have been used. Because the lungs are often patient in the laboratory (usually treadmill or cycle er- the most heavily involved in a variety of occupations, gometer) that is designed to reproduce the symptoms, if interstitial lung disease, occupational asthma, chronic ob- any, noted by the patient during exertion. Second, res- structive lung disease, lung cancer, and pleuropulmonary pired gas exchange and other data are collected during disorders are among the most common evaluated. How- exercise that allow an assessment of work capacity and, ever, the age of these subjects (many men) together with importantly, the integrative function of the organ systems other factors simultaneously put them at risk for ischemic necessary to perform exercise. The questions asked in a heart disease and congestive heart failure. The question typical clinical situation are designed to identify or con- that we should ask, therefore, is whether pulmonary func- firm dysfunction of each of the various organ systems (ta- tion tests, including spirometry, lung volumes, and diffus- ble 1). ing capacity for carbon monoxide, along with radiograph- ic imaging, are sufficient to estimate exercise capacity and When the integrative cardiopulmonary exercise test is its reciprocal, impairment. Even at first glance, the an- employed for evaluation of occupational disease, differ- swer should be no. That is, while pulmonary function tests ent questions may be appropriate (table 1). In a patient at rest go a long way in quantifying the capacity of the with known exposure to asbestos who has interstitial lung respiratory system, they say nothing about the demand of disease (asbestosis), reduced vital capacity, and an abnor- exercise. Furthermore, none of these tests provide infor- mal chest roentgenogram, the severity of the impairment mation about the cardiovascular system or the musculo- rather than the presence of asbestosis may be in question. skeletal capacity of the subject, or help determine either On occasion, this individual may be completely asymp- peak capacity or sustainable work rate. In theory, cardio- tomatic and have no complaints about exercise intoler- pulmonary exercise testing would be ideal for the purpose ance, yet be a claimant in an individual or group action of impairment evaluation, largely becomes it overcomes related to occupational exposure. Often, a patient or clai- the limitations of resting studies, integrates information mant with complaints of exercise intolerance may have co-existing cardiopulmonary disease, obesity, disorders caused by smoking, poor conditioning, or other factors 218 Sue

unrelated to the occupational exposure. Finally, a com- ity was defined as any restriction or lack (resulting from parison of work capacity measured during the exercise impairment) of ability to perform an activity within the test to the estimated rate of work being performed on a range considered normal for a human being. Thus, dis- given job may help in determining disability as opposed ability here is more similar to a global form of impair- to impairment. ment. The WHO reserved the term handicap to represent the total effect of disability on the subject’s life and work. Defining Impairment and Disability Harber and Fedoruk [7] point out a fundamental dif- This discussion cannot review in detail the definitions ference that physicians must keep in mind. The usual of impairment, disability, and fitness for work, and there evaluation of a patient seeks to determine if the patient’s are several comprehensive sources of this information [1– function is abnormal; that is, sufficiently different (lower) 5]. However, it is important to distinguish impairment than a normal population. Thus, a vital capacity that is and disability. Impairment, as used in the United States, below the 95% confidence limit for a sample of the same is considered to be decreased functional capacity on a gender, height, and age would be interpreted as abnormal. medical basis. Impairment lends itself to being measur- For impairment evaluation, however, this is overly sim- able, often in an objective manner. For example, the plistic because it focuses on the amount of function lost, American Thoracic Society statement described impair- but does not address the level of function remaining. ment as ‘...purely a medical definition. Most impairments result from a functional abnormality, which may or may Because cardiopulmonary exercise testing provides ob- not be stable at the time the evaluation is made, and may jective measurement of physiologic function, this form of be temporary or permanent’ [3]. The American Medical testing clearly relates to assessing impairment rather than Association (AMA) definition of impairment is similar: ‘a disability. Thus, physicians using exercise testing can loss, loss of use, or derangement of any body part, organ make a statement about maximum or sustained exercise sytem, or organ function’ [2]. Disability is a more global capacity, and whether and how much exercise capacity is statement about the impact of identified impairment on limited. Furthermore, if findings warrant, information the subject. Disability assessment requires social, eco- about abnormal physiologic function may be found, such nomic, environmental, and other data. Disability was as abnormal lung gas exchange. On the other hand, exer- defined by the American Thoracic Society as ‘a term that cise testing helps in disability evaluation only so much as indicates the total effect of impairment on a patient’s life. it provides information about impairment. It is affected by such diverse factors as age, gender, educa- tion, economic and social environment, and energy re- Physiological Basis for Cardiopulmonary Exercise quirements of the occupation’ [3]. The AMA defines dis- Testing ability as ‘an alteration of an individual’s capacity to meet personal, social, or occupational demands or statutory or The features of cardiopulmonary exercise testing that regulatory requirements because of an impairment’ [2]. lend itself particularly well to the assessment of impair- Physicians are tasked with identifying and measuring ment include (1) objective measurements of work perfor- impairment but, although opinions are sought from physi- mance, albeit usually with a non-work related task; cians about a patient’s disability, disability decisions are (2) measurements of physiological function of several or- most often made through an administrative or other non- gan systems, and (3) evidence that decision-making in the medical system. An increasingly important concept is impairment process is improved. accommodation with respect to disability. The level and type of disability may change as the job requirement and Simple measurements of exercise capacity such as a environment change to accommodate an individual. At timed walk, or treadmill or cycle exercise with electrocar- times, conflicts develop because some physicians are diogram monitoring, do not require much equipment and asked to act in the patient’s best interest, while others may be useful in determining suitability for major sur- represent the employer. World Health Organization defi- gery, risk stratification after myocardial infarction, or nitions were different [6], for example. Impairment was screening for ischemic heart disease in certain popula- defined as any loss or abnormality of psychological, physi- tions. However, impairment evaluation is often more sub- ological, or anatomical structure or function while disabil- tle and requires more sophisticated methods. This is because claimants and subjects may have little or no impairment, few symptoms, and may not have a single disorder or abnormal organ system. Evaluation of Impairment and Disability 219

Integrative Cardiopulmonary Exercise Testing In assessing work capacity and cause of limitation, Measurement of gas exchange during exercise, includ- peak V˙ O2 must not be designated as purely a pulmonary ing oxygen uptake (V˙ O2), carbon dioxide output (V˙ CO2), variable. For example, the quotient of V˙ O2/heart rate is heart rate, end-tidal PO2 and PCO2, and minute ventila- known as the oxygen-pulse (O2-pulse) and, when mea- tion (V˙ E) distinguishes integrative cardiopulmonary exer- sured at peak V˙ O2, gives a non-invasive estimate of maxi- cise testing from other kinds of exercise testing. Usually, mum cardiac stroke volume [10]. Similar, we can regard pulse oximetry is used and, on occasion, arterial blood V˙ O2 (and peak V˙ O2) as equal to the product of cardiac gases are very useful. Computer-controlled data collection output (Q) and the arterial-mixed venous O2 content dif- using ‘breath-by-breath’ techniques are highly suitable ference (C[a-v]O2). Any limitation in the heart rate re- during rapidly changing exercise stress, and commercial sponse (chronotropic incompetence, beta-blockade) or systems that incorporate gas analyzers, flow meters, and stroke volume response (cardiomyopathy, valvular heart displays are readily available. Extensive discussion of disease) will cause a decrease in cardiac output. If, at peak interpretation of exercise gas exchange variables is pro- exercise, arterial O2 content is abnormally low, as seen in vided in other references [8–11]. arterial hypoxemia, anemia, or carbon monoxide expo- Exercise on a treadmill or cycle ergometer is performed sure, or mixed venous O2 content is abnormally high, because the leg muscles are able to generate a much larger peak V˙ O2 can be abnormally low. work rate than the arms, and it is the physical work rate that creates the metabolic stress that tests the capacity of Submaximal Exercise the heart, lungs, and circulation. The patient or subject Maximal exercise studies are essential in determining performs a measured exercise stress while gas exchange impairment because the exercise capacity, loss of capaci- measurements are made. Most often, the work rate being ty, and remaining capacity of the individual can be deter- performed by the subject is increased incrementally every mined. But, submaximal measurements have a role in minute by increasing the resistance to pedaling a cycle assessing work capacity, impairment, and disability as ergometer or by increasing the slope or speed of the tread- well. The increase in V˙ O2 (¢V˙ O2) per increase in work mill. The exercise protocol is planned in advance for the rate (¢WR) during an exercise test in which work rate is subject to reach a point of voluntary termination of exer- increased linearly with time is related to the proportion of cise in 8–12 min. It is emphasized that this kind of testing aerobic vs. anaerobic metabolism in the exercising mus- depends on having the subject or patient exercise as much cles. A low ¢V˙ O2/¢WR ratio means a greater proportion and as long as possible. The symptoms, if any, that stop of anaerobic metabolism because of decreased oxygen exercise should be identified and recorded. availability to the exercising muscles [12–14]. This may be seen in those with heart or circulatory disorders, Exercise V˙O2 and Peak V˙O2 including pulmonary hypertension. The question of whether a patient has normal exercise capacity (and therefore can perform work at this rate) is Lactic Acidosis Threshold and Sustained Exercise best answered by whether or not the peak V˙ O2 during Capacity exercise is normal when compared to values found in nor- The work rate at which there is a sustained appearance mal subjects of the same gender, size, and age. If peak V˙ O2 of blood lactate is important in assessing sustainable work is normal in a given subject, this indicates that the func- capacity. During an incremental exercise test in a normal tional capacities of the several organ systems needed to subject, lactic acid and lactate do not appear in the blood transport O2 are normal. These include the respiratory until the work rate exceeds some particular value for a system (lung gas transfer, pulmonary circulation, and ven- subject doing that particular kind of work (e.g. leg cycling tilatory capacity), heart (heart rate and stroke volume), exercise). This lactate threshold, expressed as V˙ O2, liters/ circulation (blood vessels and hemoglobin concentration), min, is well below the peak V˙ O2 in normals. Below this and ability of exercising muscles to take up and consume work rate, blood lactate is not different than at rest (!1– oxygen (sufficient muscle mass, normal oxidative metabo- 2 mEq/l); above this work rate, lactate is above this range lism, no musculoskeletal disorders). In contrast, the sub- and, if exercise at this level is continued, the lactate con- ject with a decreased peak V˙ O2 is impaired by definition, centrations continues to rise. Lactate is the end product of and this finding triggers a search for the cause of decreased anaerobic glycolysis, a metabolic pathway that produces exercise capacity, with the goal of isolating the involved energy from glucose or glycogen but requires no O2. There organ system and the mechanism of abnormality. is considerable evidence that the appearance of lactate 220 Sue

during exercise indicates that an increasing proportion of serve are high; in this situation, such things as peripheral energy production is produced from anaerobic metabo- vascular disease, myocardial ischemia, musculoskeletal lism in addition to energy produced from aerobic path- disease, and lack of effort should be considered. A ventila- ways, indicating that oxygen delivery to the exercising tory limitation to exercise is also suggested when compari- muscles is inadequate to meet fully the needs of oxidative son to the maximum resting flow-volume loop for that metabolism [15–19]. Others have explained the appear- patient is made [23, 24]. Both inspiratory and expiratory ance of lactate by other mechanisms, including decreased flows and tidal volume are superimposed onto the flow- removal or metabolism of lactate or changes in the pro- volume loop; if they approach or reach the outer envelope, portion of muscle fiber types participating in contraction. flow or ventilatory limitation is documented. Regardless of the mechanism of physiologic mechanism of lactate appearance, the presence of increased blood lac- Pulmonary Gas Exchange and Efficiency of tate identifies a work rate above which sustained exercise Ventilation cannot be performed. Below that work rate, a subject Pulmonary gas exchange efficiency and ventilation- should be able to continue working at that rate indefinite- perfusion mismatching are best assessed using arterial ly; above that work rate, the level of exercise cannot be blood gases with calculation of alveolar-arterial PO2 dif- sustained, and subjects are limited by progressive dys- ference, arterial-end tidal PCO2 difference, and dead pnea or fatigue. space/tidal volume ratio. Arterial blood gases during exer- cise are among the most sensitive indicators of lung dis- Clinically, patients with abnormally low capacity for ease [25, 26], and these are likely to be earliest findings in oxygen delivery because of disorders of the lungs, heart, patients with subtle or questionable interstitial lung dis- pulmonary or systemic circulation, or anemia during ease or pulmonary vascular disease. Non-invasive mea- exercise develop lactate appearance at relatively low work surements of ventilatory efficiency may be useful in those rates, supporting the relationship between decreased oxy- without arterial blood gases or as a screening tool. The gen delivery and likelihood of anaerobic metabolism. An ratio of minute ventilation to CO2 output (V˙ E/V˙ CO2) in estimate of the V˙ O2 at which lactate appears in the blood patients with congestive heart failure and pulmonary can be made non-invasively by the finding of increasing hypertension is an independent marker of severity of dis- amounts of CO2 (V˙ CO2) in the expired gas relative to the ease as well as correlating with other variables. Because amount of oxygen taken up (V˙ O2). The additional CO2 is V˙ E (BTPS)/V˙ CO2 (STPD) is inversely proportional to formed from reaction of lactic acid with bicarbonate, pro- alveolar PCO2 and dead space/tidal volume ratio, a high ducing carbonic acid that dissociates into CO2 and H2O. ratio must mean either hyperventilation or high dead In practice, this point can be found by either finding the space/tidal volume ratio. point at which V˙ CO2 increases more rapidly than V˙ O2 during an incremental exercise test, or by identifying the Physiologic Basis for Exercise Testing in point at which the additional CO2 stimulates ventilation Impairment Evaluation (increase in V˙ E and V˙ E/V˙ O2 relative to V˙ E/V˙ CO2) [20–22]. The V˙ O2 where this occurs has been termed the anaerobic Measurements that can be and usually are made during threshold or lactic acidosis threshold to distinguish this integrative cardiopulmonary exercise testing are shown in from the point at which blood lactate begins to accumu- table 1, with each variable linked to specific questions late. about the function or dysfunction of an organ system. The potential features of integrative cardiopulmonary exercise Heart Rate Reserve and Breathing Reserve testing that lend themselves particularly well to evalua- In a subject with decreased peak V˙ O2, the question of tion of impairment include: (1) objective measurement of cardiac vs. respiratory limitation is frequently an issue. maximum work capacity; (2) identification of the level of Heart rate reserve is the difference between predicted and sustainable work rate; (3) clarification of cardiac com- actual maximum heart rate. A large heart rate reserve sug- pared to ventilatory limitation, and (4) high sensitivity for gests that, at the time the subject stopped exercise, the finding pulmonary gas exchange abnormalities compared limiting factor to further exercise was some organ system to measurements made at rest. other than the heart. A large breathing reserve (maximum voluntary ventilation – maximum exercise V˙ E) supports a In 1986, the American Thoracic Society statement on limiting factor other than the ventilatory capacity. Not evaluation of impairment and disability secondary to infrequently, both heart rate reserve and breathing re- Evaluation of Impairment and Disability 221

Table 2. American Thoracic Society (1986) recommendations for nary function (FEV1) in patients with lung disease is that respiratory impairment FEV1 can only estimate ventilatory capacity but not ven- tilatory requirement. Ventilatory requirement is depen- FVC FEV1 FEV1/FVC DLCO dent on multiple variables as can be seen in the following formula relating minute ventilation (V˙ E) to V˙ O2: Normal 1 80 1 80 1 75 1 80 Mild 60–79 60–79 60–74 60–80 V˙ E = V˙ O2 ! R ! 863/(PaCO2 ! [1 – VD/VT]) Moderate 50–59 40–59 40–59 40–59 Severe where R is the gas exchange ratio (equal to the respiratory ! 50 ! 40 ! 40 ! 40 quotient in the steady-state), PaCO2 is arterial partial pressure of CO2, VD/VT is the dead-space/tidal volume Values are percentages of predicted value for each variable. ratio, and 863 adjusts for BTPS and STPD volumes and FVC = Forced vital capacity; FEV1 = forced expiratory volume in 1 s; barometric pressure. If, on the cycle ergometer, there is a DLCO = single-breath diffusing capacity for carbon monoxide. For predictable relationship between V˙ O2 and work rate impairment, any FVC, FEV1, FEV1/FVC, or DLCO in the mild, (about 10 ml/min/W plus resting V˙ O2 and V˙ O2 of un- moderate, or severe impairment range defines the level of respiratory loaded cycling) [10], then work rate can be substituted in impairment. this equation. However, rather than providing a predic- tion formula for exercise V˙ E, this formula points out the respiratory disorders recommended a systematic evalua- pitfalls in trying to estimate how much ventilation an tion process for the determination of impairment [3]. The individual needs at a given work rate. Ventilatory require- statement was ‘concerned primarily with impairments ment is inversely proportional to the efficiency of gas related to reduced lung function,’ and presented a rating exchange (1 – VD/VT) and the ventilatory set point system for impairment from lung disease based on forced (PCO2), and directly proportional to the gas exchange vital capacity (FVC), forced expiratory volume in 1 s ratio and metabolic rate. The ventilatory set point is (FEV1), FEV1/FVC, and single-breath diffusing capacity determined from a variety of factors, including genetic or for carbon monoxide (DLCO) as shown in table 2. Thus, familial chemosensitivity, metabolic acidosis or alkalosis, it was implied that there was a strong relationship be- and degree of hypoxemia during exercise. tween measurements made at rest (FEV1 and DLCO) and measurements made during exercise (peak V˙ O2 and work Ventilatory capacity (maximum exercise V˙ E) is often capacity). The authors concluded that the majority of sub- predicted from FEV1 [27, 28] or estimated from the maxi- jects undergoing an evaluation for impairment would not mum voluntary ventilation. The pattern of breathing (rate require exercise testing, and resting pulmonary function and tidal volume) varies between those with obstructive, provides sufficient information for accurate categoriza- interstitial restrictive, pleural restrictive, and other lung, tion of patients in terms of impairment. heart, and chest wall disorders [24, 29]. Furthermore, inspiratory muscle fatigue is enhanced during exercise by It is important to review the basis for this conclusion hyperinflation, high airway resistance, and stiff lungs or and perhaps to consider if this recommendation is ideal chest wall; and fatigue during exercise cannot be antici- for the population to be examined. First, can resting pul- pated from the FEV1 determination. monary function adequately predict peak V˙ O2 and work capacity, especially in those with moderate decrease in Figure 1 incorporates estimates of ventilatory capacity lung function? Second, does an exercise test add to the superimposed on a range of ventilatory requirements (or accuracy of evaluation, notably in subjects for whom an demands) and demonstrates the range of peak V˙ O2 (or exercise test might not otherwise be considered indi- work rate) that might be found for two hypothetical sub- cated? jects having the same FEV1. In this analysis, FEV1 is used to predict maximum exercise minute ventilation (V˙ E) by Ventilatory Requirement during Exercise and using two estimates of maximum minute ventilation from Ventilatory Capacity FEV1 (7, 14). Panel a shows the two different maximum If exercise capacity is limited by shortness of breath, in V˙ E values that would be predicted using 35 or 40 ! FEV1 most cases there is insufficient ventilatory capacity to = maximum V˙ E. Next, in panel b, the dead space/tidal meet the ventilatory requirement for the work being per- volume ratio (VD/VT) at maximum exercise ranges from formed. The primary basis for potentially inaccurate pre- about 0.12 to 0.30 at maximum exercise in normal sub- diction of maximum work capacity from resting pulmo- jects and may exceed 0.35 in those with ventilation-perfu- sion mismatching from lung disease. In this analysis, large 222 Sue

Fig. 1. Potential physiological basis for inac- curacy of predicting exercise capacity (maxi- mum oxygen uptake, peak V˙ O2) from lung function at rest (FEV1) using a range of val- ues to illustrate maximum differences. Max- imum minute ventilation (V˙ E) during exer- cise is assumed to be between 35 and 40 times FEV1, so for a given FEV1 (star), a shows the range of maximum V˙ E (1a and 2a). In b, for a range of dead space/tidal volume ratios (VD/VT), V˙ E can span a considerable range of alveolar ventilation (V˙ A): points 1b and 2b. For a given alveolar ventilation, a greater CO2 output (V˙ CO2) can be elimi- nated at a higher than at a lower PCO2 (1c and 2c). For different respiratory gas ex- change ratios (R), a range of V˙ O2 (1d–2d) for the different V˙ CO2 is possible. Therefore, the same FEV1 value in 2 different patients can be associated with a wide range of peak V˙ O2 or work rates. This strongly suggests that maximum exercise capacity will be poorly predicted from FEV1 alone. From Sue DY: Exercise testing in the evaluation of impairment and disability. Clin Chest Med 1994;15:369–387, with permission. and small VD/VT values were chosen to illustrate the Table 3. When is resting pulmonary function likely to predict exer- range of alveolar ventilation (V˙ E ! [1 – VD/VT]) that cise capacity? may be seen. Considerable potential differences in V˙ A can be found. Although V˙ A is usually considered to deter- Only when subjects or patients are ventilatory-limited during mine PCO2 at a given V˙ CO2, in this example, the combi- exercise nation of V˙ A and PCO2 values (high and low) shows the potential range of V˙ CO2, (panel c). Finally, V˙ O2 can be Metabolic rate increases in proportion to exercise work rate expressed as the quotient of V˙ CO2 and R, the respiratory Ventilatory requirement at a given metabolic rate is predictable gas exchange ratio; the potential range of V˙ O2 (and work rate) at maximum exercise can be determined (panel d). (V˙ E/V˙ O2 or V˙ E/V˙ CO2) Therefore, the range of gas exchange disturbances in Ventilatory response (CO2 response) is predictable patients with lung disease has the potential of markedly Accurate measurement of pulmonary function at rest dissociating any consistent relationship between FEV1 Absence of cardiac disease (CHF, myocardial ischemia) during and work capacity. In summary, this analysis shows that even if the ventilatory capacity of the respiratory system exercise for ventilation is successfully predicted from resting FEV1, the ventilatory requirement for a given level of function to predict ideally exercise capacity, the require- work cannot be predicted from the FEV1. Finally, al- ments outlined in table 3 must be met. though maximum exercise V˙ E is related to FEV1 in most subjects, there are likely differences in this relationship Can Exercise Performance Be Predicted Accurately between those with normal lung function, those with from Resting Pulmonary Function? obstruction, patients with restrictive lung disease, and Despite the physiological argument that anticipates obese subjects [9, 10]. In summary, for resting pulmonary inaccuracy in predicting exercise capacity from resting pulmonary function, clinical trials would be desirable to Evaluation of Impairment and Disability 223

answer the question: Can resting lung function reliably 39 patients were classified in this category. Of these 39, peak V˙ O2 in 23 (58.9%) was higher than 15 ml/kg/min; in estimate exercise performance? 14 (35.9%), peak V˙ O2 exceeded 60% of predicted. On the other hand, of the 39 who did not meet criteria for severe Cotes et al. [30] tested the American Thoracic Society impairment, 8 (20.5%) had a peak V˙ O2 !15 ml/kg/min and 13 (33.3%) had a peak V˙ O2 !60% predicted. When statement hypothesis [3] that exercise limitation could be their data are reported as likelihood ratios (LR), resting pulmonary function tests had an LR for peak V˙ O2 predicted accurately from degree of lung function reduc- !15 ml/kg/min of only 2.0 (sensitivity 41%, specificity tion, by using peak V˙ O2 as the dependent variable in a 79.5%), the LR for peak V˙ O2 !60% predicted was only stepwise multiple regression analysis to identify which 1.92 (sensitivity 64.1%, specificity 66.6%). There was a variables significantly affected peak V˙ O2, and how much of the variance (r2) each of these variables could explain. poor correlation of exercise capacity with FEV1, FVC, lung volumes, and arterial blood gases at rest, with r2 rang- A group of 157 men with a variety of suspected or known ing from 0.05 to 0.28. The authors concluded that resting occupational lung disorders were studied, among whom pulmonary function tests are not predictive of exercise there were diagnoses of coal workers’ pneumoconiosis, sil- performance in COPD, and that cardiopulmonary exer- icosis, asbestosis, occupational asthma, and allergic al- cise testing is needed for accurate determination of im- veolitis. All subjects had (1) FEV1, FVC, or DLCO !80% of predicted, or FEV1/FVC !75%; (2) limitation of exer- pairment. cise by dyspnea rather than chest pain, fatigue, or other reasons; (3) maximum V˙ E within 2 SD of the maximum Other studies have had varied results in different lung V˙ E predicted from FEV1. The latter criterion was meant to confirm that lack of breathing reserve was the essential disorders. In obstructive lung disease, Vyas et al. [32] cause of exercise termination; this would make it more found that in 14 patients with a mean FEV1 of 29% of predicted, the peak V˙ O2 could be predicted satisfactorily. likely that pulmonary function would predict exercise The percentage of variance explained for peak V˙ O2 was 78–82% using FEV1 and VC. Pineda et al. [33] found that capacity. Results were expressed as fraction of total vari- FEV1 accounted for 56% of the variance in peak V˙ O2. ance of peak V˙ O2 (dependent variable) accounted for (r2) Jones et al. [34] found a correlation between work rate by independent variables taken singly and stepwise. For and severity of airway obstruction (r = 0.621) but this single variables, variances were for FVC 15%, for FEV1 20%, for FEV1/FVC 8.4%, and for DLCO 25%. The com- explained just 37% of the variance in work rate. Foglio et bination of FEV1 (or FVC and FEV1/FVC) and DLCO could account for 29% of peak V˙ O2 variance and 12–14% al. [35] examined the determinants of exercise perfor- of peak V˙ O2 expressed as % predicted. However, when FEV1 (or FVC and FEV1/FVC) was combined with exer- mance in 105 patients with chronic airway obstruction. In cise V˙ E (measured or extrapolated to the value at V˙ O2 = 1 liter/min), these variables could account for 54% of the the model they used, age was the major factor but residual variance of peak V˙ O2 and 44% of the variance in peak V˙ O2, % predicted. The authors concluded that loss of volume and dyspnea scores explained the largest amount exercise capacity was not accurately predicted from rest- of the variance. Because the covariates together only ing lung function indices alone, but the predictability of peak V˙ O2 was enhanced by adding the result of V˙ E during accounted for 26–34% of the variance, they concluded submaximal exercise (V˙ E at V˙ O2 = 1 liter/min). They com- mented that the assumption that a helpful and consistent that FEV1 was an inconsistent predictor, and suggested the importance of performance evaluation in these pa- relationship between resting pulmonary function and ex- tients. ercise capacity existed was based on a limited number of Carlson et al. [36] reported that exercise tolerance in reports. COPD patients was difficult to predict from measure- Using the ATS guidelines, Ortega et al. [31] also sought ments of lung function. In a group of 119 patients with a to determine the precision by which resting pulmonary mean FEV1 = 1.41 B 0.64 liters, stepwise linear regres- function predicted exercise capacity, and especially how sion yielded a prediction formula for peak V˙ O2 that had an r = 0.90 using DLCO, MVV, peak exercise VD/VT, impairment and disability might be quantified. 78 stable and resting V˙ E. This formula, although it could account for 81% of the variance, had two important limitations. men with chronic obstructive pulmonary disease (mean This formula used exercise VD/VT, thereby including a FEV1 45%) were studied. Using ATS Guidelines to define severely impaired (FEV1 !40% predicted, FVC !50% variable measured during exercise to predict exercise predicted, FEV1/FVC !40%, or DLCO !40% predicted), capacity. Using only resting pulmonary function, the mul- tiple regression had an r2 = 74%. In addition, the 95% confidence limit for a given individual patient averaged 224 Sue

about 36% of the mean peak V˙ O2 (0.233 liters/min with described by Keogh et al. [41] had arterial PaO2 180 mm mean peak V˙ O2 = 1.25 B 0.50 liters/min). Hg at rest, but decreased PaO2 and increased P(A-a)O2 during exercise. These authors concluded that exercise- Exertional dyspnea was the most frequent reason for induced hypoxemia was best demonstrated during an exercise test because it was not predictable from resting stopping exercise in patients with interstitial lung disease PaO2. Hansen and Wasserman found that DLCO pre- dicted exercise limitation better than either lung volumes as reported by Rampulla et al. [37], being found in 62%. or FEV1 [39] in 42 patients with interstitial lung disease. These patients were severely limited with a mean peak Asbestosis and other occupational lung diseases may V˙ O2 of 15.1 ml/kg/min. The severity of dyspnea corre- behave differently than other interstitial lung diseases lated with peak V˙ O2 but dyspnea in individual subjects such as idiopathic pulmonary fibrosis. One group [42] was not well predicted from resting lung function. How- studied 9 age-matched patients with each disorder with comparable peak V˙ O2 (pulmonary fibrosis 1.09 B 0.27 ever, in 20 men with pulmonary fibrosis reported by Spiro liters/min; asbestosis 1.07 B 0.37 liters/min). Resting et al. [38], the mean peak V˙ O2 of 1.4 liters/min correlated lung function was similar. Among the pulmonary fibrosis well with resting FEV1, VC, and TLC. The report of Han- group, arterial PO2 fell more, P(A-a)O2 was larger, and sen and Wasserman [39] also compared resting FEV1 % VD/VT was greater during exercise than those with asbes- predicted, TLC % predicted, DLCO % predicted, and tosis. In contrast, asbestosis patients and those with cryp- resting PaO2 with peak V˙ O2 % predicted in 42 patients togenic fibrosing alveolitis matched for comparable de- with interstitial lung disease and no other limiting factors. grees of exercise-induced desaturation and peak V˙ O2 (1.3–1.5 liters/min) had similar resting lung function, Predicted values were used to standardize patients of dif- DLCO, and severity of chest X-ray abnormality scores [43]. Dyspnea scores correlated with exercise perfor- ferent gender, age, and size. The correlation coefficients mance in 153 workers exposed to silica dust and 62 (r) and % variance explained (r2) were, for FEV1 r = 0.498, patients with silicosis in a study by Wang et al. [44], but a 24.8%, for TLC 0.574, 32.9%, DLCO 0.764, 58.3%, and considerable proportion (30–56%) of each group reported more severe dyspnea by questionnaire than was verified resting PaO2 0.570, 32.5%. The authors concluded that during exercise testing. They concluded that objective ventilatory mechanics did not limit exercise capacity in physiological measures like exercise testing may be of val- ue in dyspneic silica exposed subjects. Another confound- the majority of patients with interstitial lung disease. In ing factor after asbestos exposure may be the influence of pleural thickening or plaques. In a study of 90 subjects several studies, Marciniuk et al. [23, 40] provided more with asbestos exposure for at least 1 year and more than 20 years since first exposure, asbestosis was defined as evidence that mechanical ventilatory limitation did not ILO profusion of 1/0 or greater [45]. Among those without parenchymal involvement, diffuse pleural thickening was determine maximal exercise performance in interstitial associated with lower maximal work capacity, and more severe increase in P(A-a)O2 and VD/VT at maximum lung disease, but arterial hypoxemia might contribute exercise. The authors hypothesized that subtle asbestos- induced parenchymal disease was present. greatly. First, patients with interstitial lung disease did Impairment from Nonventilatory Limitation during not usually reach the maximal flow-volume envelope at Exercise One of the major potential reasons why resting pulmo- the end of exercise, suggesting that ventilatory reserve was nary function may not be highly predictive of exercise performance is nonventilatory limitation during exercise. present [23]. Second, supplemental oxygen breathing sig- The best correlation of pulmonary function and exercise capacity will be seen in those who are limited by dyspnea nificantly increased exercise performance in 7 patients from lung disease rather than fatigue, chest pain, muscu- loskeletal disorders, or exercise-induced asthma or heart with interstitial disease [40]. The latter group had a mean peak V˙ O2 of only 56 B 13% predicted, but oxygen breath- ing led to a significant increase in peak minute ventila- tion, work rate, exercise duration, and peak V˙ O2 (1.25 B 0.21 liters/min vs. 1.39 B 0.26 liters/min). Correlation with peak V˙ O2 has not been the only rela- tionship sought. Non-invasive measurements have been used to identify patients with abnormal exercise arterial blood gases, but sensitivity is poor. For example, in 276 current or former shipyard workers, the predictive ability of DLCO to find abnormal arterial blood gases was exam- ined [25]. Abnormal exercise blood gases were defined as P(A-a)O2 1 35 mm Hg, arterial-end tidal PCO2 difference (P[a-ET]CO2) 1 0 mm Hg, or VD/VT 10.30 at maximum exercise. A DLCO !80% predicted had only 22% sensi- tivity although 96% specificity (likelihood ratio = 5.5); the 95% confidence limit for normal DLCO was even less sensitive. 53 patients with idiopathic pulmonary fibrosis Evaluation of Impairment and Disability 225

failure. Rampulla et al. [37] found that of 66 consecutive Cardiopulmonary Exercise Testing in the patients with chronic lung disease (COPD and interstitial Assessment of Impairment and Disability lung disease), only 42% stopped exercise due to dyspnea while others stopped because of fatigue (41%), cardiac Cardiopulmonary exercise testing in impairment eval- limitation (12%), and other reasons. In our study of exer- uation has been called ‘difficult to perform, more expen- cise testing in impairment evaluation [46], 138 had abnor- sive, and sometimes more invasive than conventional mally low exercise capacity, yet only 18% of these were tests.’ [2]. However, a number of studies emphasize the limited by obstructive or restrictive lung disease, while usefulness of cardiopulmonary exercise testing for deter- 69% had a cardiovascular cause of exercise limitation. mination of impairment and disability specifically from These subjects’ exercise capacities would have been poor- occupational exposure [30, 42, 43, 46, 48, 50–53]. An ly estimated from FEV1, VC, or DLCO. Of the 42 patients appropriate perspective, it can be argued from these stud- with interstitial lung disease reported by Hansen and ies, is that exercise testing complements clinical evalua- Wasserman [39], peak V˙ O2 correlated much better with tion, improves the accuracy of resting pulmonary func- measurements of circulatory status, such as the lactic aci- tion, and supplements roentgenographic studies. Exercise dosis threshold, the oxygen-pulse extrapolated to pre- testing adds to diagnostic accuracy, both quantitatively dicted heart rate, and the ratio of V˙ O2 to work rate, than (measurement of work capacity, peak V˙ O2, and sustained with ventilatory or gas exchange variables. Because the work capacity) and qualitatively (identification of the severity of circulatory dysfunction correlated with the cause of exercise limitation). severity of lung impairment (loss of lung volume, abnor- mal gas transfer index, and dead space/tidal volume Exercise Testing and Decision Making on Impairment ratio), the pulmonary circulation was implicated in these We [46] had the opportunity to perform exercise stud- patients rather than the heart or systemic circulation. ies and obtain clinical information on 348 current or Neder et al. [47] recently reported predictive values for former shipyard workers, all men, who complained of exercise testing in normal subjects, using a randomly exercise limitation but whose FEV1 was at least 40% pre- selected sample of 120 sedentary individuals from a pool dicted and FEV1/VC more than 40%, and had no obvious of more than 8,000 subjects. Although the intent of the cardiac dysfunction. In this retrospective analysis, we study was to establish another set of normal predictive compared the consultant’s conclusion about impairment values, their data provide evidence for potential success and cause of impairment based on chest radiographs, pul- of predicting cardiovascular and ventilatory variables monary function tests, resting electrocardiogram, and from resting data. For example, peak V˙ O2 was fairly well clinical information with the conclusion reached with the predicted from age, size, activity score, and leisure time aid of the cardiopulmonary exercise test data. Normal score, with coefficients explaining 70–90% of the vari- work capacity (no impairment) was defined as being with- ance. But ventilatory variables were much less predicta- in the 95% confidence limit for normal peak V˙ O2 [10, 54]. ble. Predicted maximum V˙ E when gender, age, and size Interpretation of exercise tests was standardized for car- were included as variables had r2 values of only 0.379 diovascular limitation, ventilatory limitation, and noncir- (men) and 0.536 (women). There were similar r2 values culatory, nonrespiratory disorders, including poor effort for V˙ E/MVV, respiratory frequency, and tidal volume, as and musculoskeletal problems [46]. well as for heart rate and lactic acidosis threshold. Non-exercise evaluations led to the conclusion that 148 subjects would have normal work capacity, but 46 of In summary, there is a limited physiological basis for these (31%) had a peak V˙ O2 below the 95% confidence predicting exercise capacity from resting pulmonary func- limit. On the other hand, 66 men were predicted to have tion in patients with lung disease, and non-ventilatory low work capacity, but of these, only 43 (67%) were cor- limitation is even harder to predict. Evidence supporting rectly categorized. Among those whose exercise capacity this concept comes from a number of studies demonstrat- could not be estimated, 60% had normal peak V˙ O2 and ing the poor predictive value of such variables as FVC, 37% were abnormally low. Thus, the sensitivity of non- FEV1, and DLCO. These results show, not surprisingly, exercise data was 31% for impairment, with a specificity that accurate determination of exercise capacity is best of 49%. The non-exercise test information had a positive performed by actually performing an exercise test. predictive value for impairment of 65% for this popula- tion; the corresponding negative predictive value was 69%. It should be noted that impairment in this study was 226 Sue

defined as an abnormal peak V˙ O2, the usual clinical defi- A report of a working group of the European Society nition, rather than in terms of functional capacity or exer- for Clinical Respiratory Physiology [55] decided that cise reserve. Resting pulmonary function tests had low equating the degree of respiratory impairment from pul- sensitivity for predicting abnormal exercise arterial blood monary function tests with the degree of reduced exercise gases. For example, single-breath diffusion capacity for capacity was not ideal. They noted also that the World carbon monoxide (abnormal !80% predicted), had a sen- Health Organization definition of respiratory disability sitivity of 45% and specificity of 91% for abnormal exer- required assessment of both losses of lung function from cise alveolar-arterial PO2 difference (135 mm Hg). For pulmonary function tests and information about exercise predicting abnormal dead space/tidal volume ratio (10.30 performance [6]. Except for those with severe respiratory at maximum exercise), diffusing capacity for carbon mon- impairment (resting pulmonary function), the working oxide had 18% sensitivity and 94% specificity. group recommended assessing exercise capacity from symptom-limited exercise testing, with measurement or The cause of exercise limitation should have a major estimation of peak V˙ O2. They proposed a scale of percent- effect on the ability to predict impairment because only age ‘respiratory disability’ with a score ranging from 0 to pulmonary function (and clinical features suggesting se- 100%, referring to the estimated loss of function as deter- vere limitation) should be particularly helpful. Further- mined from exercise testing. The peak V˙ O2 at the lower more, as in other studies [37, 39], nonventilatory causes of limit of the 95% CI was chosen as the point of 0% disabil- limitation are likely to be common in this patient popula- ity (using 1.64 SD below the mean predicted value). For tion. Only 25 of 138 with impairment were limited by 100% disability, the group chose a peak V˙ O2 of 0.5 liters/ obstructive or restrictive lung disease, while 69% (95/138) min or about twice the resting V˙ O2 for normal subjects (a had a cardiovascular cause of exercise limitation. These very severely limited exercise capacity corresponding to data are similar to those of Agostoni et al. [48] who found only a few walking steps at a time). They presented data that 37% of 120 former asbestos workers had cardiac limi- from 157 men with respiratory impairment (using ATS tation rather than ventilatory limitation. pulmonary function criteria) and found that 28% had no impairment from exercise testing by their calculation; In a further retrospective analysis of these data, we 38% had 1–19% respiratory disability; 11% had severe matched smoking and never-smoking shipyard workers disability (60–79% respiratory disability); and 1% had by age and duration of asbestos exposure [52]. Smokers 100% respiratory disability. had significantly lower VC, FEV1, FEV1/VC, and DLCO than nonsmokers, and smokers more frequently had low- Cotes et al. [56] subsequently compared this method er peak V˙ O2 and O2 pulse and more often had an abnor- with an established empirical rating for total cardiorespi- mally increased maximum exercise P(A-a)O2. Specifical- ratory disability derived from clinical information, chest ly, 45% (33/73) of smokers had a peak V˙ O2 less than 80% radiograph, and FEV1 and FVC. Sixty-two former coal of predicted compared to only 20.5% (15/73) of nonsmok- miners, asbestos workers, or others with potential occupa- ers. Smokers were more likely to have impairment caused tional lung disorders were studied. Disability scores (%) by heart disease (49%), poor effort (21%) or musculoskel- were calculated using measured peak V˙ O2 and peak V˙ O2 etal causes (12%) than by airway obstruction (6%). estimated from age, V˙ O2 at 1 liter/min, and FEV1 [55]. The two scores correlated well with each other (r = 0.72), Exercise Testing and Quantitation of Impairment and both correlated reasonably with the empirical rating The 1986 ATS Statement [3] concluded that degree of (r = 0.51). However, addition of radiographic evidence of impairment could be classified from resting lung function progressive massive fibrosis, FEV1, and clinical grade of as normal, mildly impaired, moderately impaired, or breathlessness further reduced variance (r2 = 0.49) be- severely impaired (table 2) in those with lung disease. If tween the new rating system and the empirical disability exercise testing were added, then useful stratification score. Furthermore, in subjects with a discrepancy be- could be achieved on the basis of peak V˙ O2, expressed as tween the estimated and measured peak V˙ O2, informa- ml/kg/min. It was estimated that office work required tion from the exercise test helped explained the differ- about 5–7 ml/kg/min, moderate labor about 15 ml/kg/ ence. min, and strenuous labor 20–30 ml/kg/min. For manual labor at a comfortable working place, a work rate at Recently, Neder et al. [57] proposed expressing disabil- approximately 40% of peak V˙ O2 was chosen. Interesting- ity using what they termed loss of aerobic capacity (V˙ O2, ly, this is similar to the sustainable work rate level in nor- % predicted) rather than remaining capacity (peak V˙ O2 mal subjects as estimated by the lactic acidosis threshold. ml/kg/min), in contrast with some of the ATS recommen- Evaluation of Impairment and Disability 227

Table 4. Relative value of resting and exercise testing for impairment evaluation Resting studies Cardiopulmonary exercise Is ventilatory capacity during exercise reduced? ++ +++ Is maximum work capacity reduced? ++ (if ventilatory-limited) ++++ What is the physiological basis for reduced work capacity? + +++ Are there subtle pulmonary gas exchange abnormalities present? + ++++ Does cardiac function limit exercise? ++ (if severely reduced) +++ Table 5. Suggested strategy for using cardiopulmonary exercise test- ject may have only a small decrease in peak V˙ O2, % pre- ing in impairment evaluation dicted, but have severe impairment by peak V˙ O2, ml/kg/ min (e.g. !15 ml/kg/min). The authors reviewed data Subject or patient Cardiopulmonary exercise test suggested from 75 subjects with silica exposure. Only 19 (25.3%) had a peak V˙ O2 125 ml/kg/min (normal by ATS criteria) Asymptomatic, no disproportionate complaints about but 53.3% of the same 75 subjects had a peak V˙ O2 170% exertional dyspnea or work-related exercise intolerance of predicted (defined as normal). Out of 56 subjects with exercise intolerance impairment defined by the ATS exercise criteria (peak concern about subtle pulmonary gas V˙ O2 !25 ml/kg/min), 21 had no impairment using the Mild respiratory exchange abnormalities that may relate peak V˙ O2, % predicted. As hypothesized, older subjects impairment to an occupational exposure and those who were overweight were more likely to be dis- Mild-to-moderate cordant. The authors concluded that in evaluation of respiratory impairment above reasons impairment, comparison to predicted peak V˙ O2 was more useful than using the recommendation of peak V˙ O2, ml/ Moderate respiratory symptoms inconsistent with degree of kg/min. impairment impairment defined by resting pulmonary function testing Recommendations for Exercise Testing in Impairment Severe respiratory and Disability Evaluation impairment above reasons Although there are no absolute indications for the Any addition of exercise testing to medical history, physical resting tests may have underestimated examination, resting pulmonary function tests, and chest the degree of impairment because of roentgenogram for assessment of respiratory impairment, pulmonary vascular involvement, certain guidelines appear to be justifiable. These are sum- unsuspected coexisting disease, or marized in tables 4 and 5. Exercise testing is not indicated inability to estimate ventilatory in those who lack complaints of exercise dyspnea, fatigue, requirement for exercise or intolerance. Those with clearly severe respiratory im- pairment, such as those who meet the ATS criteria for usually none, but may be useful in some impairment by FEV1 or DLCO, also do not require exer- subjects in whom impairment is cise testing. overestimated Cardiopulmonary exercise testing is indicated when accurate measurement of work capacity is desired or accurate determination of exercise when symptoms or exercise intolerance are inconsistent capacity is desired with clinical findings and resting tests. There is strong evi- dence that evaluation and quantitation of impairment is need for apportioning cause of enhanced by use of exercise testing, notably because of the impairment lack of success of resting pulmonary function tests to pre- dict exercise capacity, and we and others have found that dations. They reasoned that considerable loss of capacity both over- and underestimation of work capacity is found could be sustained in a young, non-obese individual with a high predicted normal peak V˙ O2, but even then that individual may have only mild or no impairment by peak V˙ O2, ml/kg/min. On the other hand, an older, obese sub- 228 Sue

in subjects being assessed for impairment. In addition, cardiovascular or pulmonary vascular disorders. In sum- exercise arterial blood gases, especially with calculation of mary, cardiopulmonary exercise testing provides objec- alveolar-arterial PO2 difference and dead space/tidal vol- tive determination of exercise capacity, increased sensi- ume ratio, are very sensitive tests of subtle lung disease. In tivity for pulmonary gas exchange abnormalities, and the some patients, exercise testing is useful for diagnosis as ability to identify unsuspected or unanticipated non-pul- well as measurement of impairment. This is especially monary causes of impairment. true in those with co-existing disorders or unsuspected References 1 Sood A, Beckett WS: Determination of disabil- 14 Koike A, Hiroe M, Adachi H, Yajima T, Itoh 27 Campbell SC: A comparison of the maximum ity for patients with advanced lung diseases. H, Takamoto T, Taniguchi K, Marumo F: Car- voluntary ventilation with forced expiratory Clin Chest Med 1997;18:471–482. diac output-O2 uptake relation during incre- volume in one second: An assessment of sub- mental exercise in patients with previous myo- ject cooperation. J Occup Med 1982;24:531– 2 American Medical Association; in Cocchiarella cardial infarction. Circulation 1992;85:1713– 533. L, Andersson GBJ (eds): Guides to the Evalua- 1719. tion of Permanent Impairment, ed 5. Chicago, 28 Gandevia B, Hugh-Jones P: Terminology for AMA Press, 2001. 15 Beaver WL, Wasserman K: Muscle RQ and measurements of ventilatory capacity. Thorax lactate accumulation from analysis of the 1957;1:290–293. 3 American Thoracic Society: Evaluation of im- V˙ CO2-V˙ O2 relationship during exercise. Clin J pairment/disability secondary to respiratory Sports Med 1991;1:27. 29 Marciniuk DD, Gallagher CG: Clinical exer- disorders. Am Rev Respir Dis 1986;133:1205– cise testing in interstitial lung disease. Clin 1209. 16 Koike A, Weiler-Ravell D, McKenzie DK, Chest Med 1994;15:287–304. Zanconato S, Wasserman K: Evidence that the 4 Cotes JE, Reed JW, Elliott C: Breathing and metabolic acidosis threshold is the anaerobic 30 Cotes JE, Zejda J, King B: Lung function im- exercise requirements of the work place; in threshold. J Appl Physiol 1990;68:2521–2526. pairment as a guide to exercise limitation in Whipp BJ, Wasserman K (eds): Exercise: Pul- work-related lung disorders. Am Rev Respir monary Physiology and Pathophysiology. New 17 Wasserman K: The anaerobic threshold to Dis 1988;137:1089–1093. York, Marcel Dekker, 1991, p 495. evaluate exercise performance. Am Rev Respir Dis 1984;129(suppl):S35–S40. 31 Ortega F, Montemayor T, Sanchez A, Cabello 5 Hansen JE, Wasserman K: Disability evalua- F, Castillo J: Role of cardiopulmonary exercise tion; in Murray JF, Nadel JA (eds): Textbook 18 Wasserman K, Beaver WL, Whipp BJ: Gas testing and the criteria used to determine dis- of Respiratory Medicine, ed 1. Philadelphia, exchange theory and the lactic acidosis (anaer- ability in patients with severe chronic obstruc- Saunders, 1988, p 699. obic) threshold. Circulation 1990;81(suppl II): tive pulmonary disease. Am J Respir Crit Care 14–30. Med 1994;150:747–751. 6 World Health Organization: International Classification of Impairments, Disabilities, 19 Wasserman K, Koike A: Is the anaerobic 32 Vyas MN, Banister EW, Morton JW, Grzy- and Handicaps. Geneva, World Health Organi- threshold truly anaerobic? Chest 1992;101 bowski S: Response to exercise in patients with zation, 1980. (suppl):211S–218S. chronic airway obstruction. II. Effects of breathing 40% oxygen. Am Rev Respir Dis. 7 Harber P, Fedoruk MJ: Work placement and 20 Beaver WL, Wasserman K, Whipp BJ: A new 1971;103:401–412. worker fitness: Implications of the Americans method for detecting the anaerobic threshold with disabilities act for pulmonary medicine. by gas exchange. J Appl Physiol 1986;60:2020– 33 Pineda H, Haas F, Axe K, Haas A: Accuracy of Chest 1994;105:1564–1571. 2027. pulmonary function tests in predicting exercise tolerance in chronic obstructive pulmonary 8 Sue DY, Wasserman K: Impact of integrative 21 Dickstein K, Barvik S, Aarsland T, Snapinn S, disease. Chest 1984;86:564–567. cardiopulmonary exercise testing on clinical Karlsson J: A comparison of methodologies in decision making. Chest 1991;99:981–992. detection of the anaerobic threshold. Circula- 34 Jones NL, Jones G, Edwards RHT: Exercise tion 1990;81(suppl II):38–46. tolerance in chronic airway obstruction. Am 9 Wasserman K, Whipp BJ: Exercise physiology Rev Respir Dis 1971;103:477–491. in health and disease. Am Rev Respir Dis 112: 22 Sue DY, Wasserman K, Moricca RB, Casaburi 219–249. R: Metabolic acidosis during exercise in 35 Foglio K, Carone M, Pagani M, Bianchi L, chronic obstructive pulmonary disease. Chest Jones PW, Ambrosino N: Physiological and 10 Wasserman K, Hansen JE, Sue DY, Casaburi 1988;94:931–938. symptom determinants of exercise perfor- R, Whipp BJ: Principles of Exercise Testing mance in patients with chronic airway obstruc- and Interpretation, ed 3. Baltimore, Lippincott 23 Marciniuk DD, Sridhar G, Clemens RE, Zintel tion. Respir Med 2000;94:256–263. Williams & Wilkins, 1999. TA, Gallagher CG: Lung volumes and expira- tory flow limitation during exercise in intersti- 36 Carlson DJ, Ries AL, Kaplan RM: Prediction 11 Weber KT, Janicki JS (eds): Cardiopulmonary tial lung disease. J Appl Physiol 1994;77:963– of maximum exercise tolerance in patients with Exercise Testing: Physiologic Principles and 973. COPD. Chest 1991;100:307–311. Clinical Applications. Philadelphia, Saunders, 1986. 24 Gallagher CG: Exercise limitation and clinical 37 Rampulla C, Baiocchi S, Dacosto E, Ambrosi- exercise testing in chronic obstructive pulmo- no N: Dyspnea on exercise. Pathophysiologic 12 Hansen JE, Sue DY, Oren A, Wasserman K: nary disease. Clin Chest Med 1994;15:305– mechanisms. Chest 1992;101(suppl):248S– Relation of oxygen uptake to work rate in nor- 326. 252S. mal men and men with circulatory disorders. Am J Cardiol 1987;59:669–674. 25 Sue DY, Oren A, Hansen JE, Wasserman K: 38 Spiro SG, Dowdeswell IRG, Clark TJH: An Single-breath diffusing capacity for carbon analysis of submaximal exercise responses in 13 Hansen JE, Casaburi R, Cooper DM, Wasser- monoxide as a predictor of exercise gas ex- patients with sarcoidosis and fibrosing alveoli- man K: Oxygen uptake as related to work rate change. New Engl J Med 1987;316:1301– tis. Br J Dis Chest 1981;75:169–180. increment during cycle ergometer exercise. Eur 1306. J Appl Physiol 1988;57:140–145. 39 Hansen JE, Wasserman K: Pathophysiology of 26 Wagner PD: Ventilation-perfusion matching activity limitation in patients with interstitial during exercise. Chest 101;(suppl):192S–198S. lung disease. Chest 1996;109:1566–1576. Evaluation of Impairment and Disability 229

40 Harris-Eze AO, Sridhar G, Clemson RE, Zintel 46 Oren A, Sue DY, Hansen JE, Torrance DJ, 53 Wiedemann HP, Gee JBL, Balmes JR, Loke J: TA, Gallagher CG, Marciniuk DD: Role of Wasserman K: The role of exercise testing in Exercise testing in occupational lung disease. hypoxemia and pulmonary mechanics in exer- impairment evaluation. Am Rev Respir Dis Clin Chest Med 1984;5:157–171. cise limitation in interstitial lung disease. Am J 1987;135:230–235. Respir Crit Care Med 1996;154:994–1001. 54 Hansen JE, Sue DY, Wasserman K: Predicted 47 Neder JA, Nery LE, Castelo A, Andreoni S, values for clinical exercise testing. Am Rev 41 Keogh BA, Lakatos E, Price D, Crystal RG: Lerario MC, Sachs A, Silva AC, Whipp BJ: Pre- Respir Dis 1984;129:S49–S55. Importance of the lower respiratory tract in diction of metabolic and cardiopulmonary re- oxygen transfer: Exercise testing in patients sponses to maximum cycle ergometry: A ran- 55 Cotes JE: Rating respiratory disability: A re- with interstitial and destructive lung disease. domised study. Eur Respir J 1999;14:1304– port on behalf of a working group of the Euro- Am Rev Respir Dis 1984;129(2 Pt 2):S76–80. 1313. pean society for clinical respiratory physiology. Eur Respir J 1990;3:1074–1077. 42 Agusti AGN, Roca J, Rodriguez-Roisin R, 48 Agostoni P, Smith DD, Schoene RB, Robert- Xaubet A, Agusti-Vidal A: Different patterns son HT, Butler J: Evaluation of breathlessness 56 Cotes JE, Chinn DJ, Reed JW, Hutchinson JE: of gas exchange response to exercise in asbestos in asbestos workers. Results of exercise testing. Experience of a standardised method for as- and idiopathic pulmonary fibrosis. Eur Respir Am Rev Respir Dis 1987;135:812–816. sessing respiratory disability. Eur Respir J J 1988;1:510–516. 1994;7:1074–1076. 49 Howard J, Mohsenifar Z, Brown HV, Koerner 43 Markos J, Musk AW, Finucane KE: Functional SK: Role of exercise testing in assessing func- 57 Neder JA, Nery LE, Bagatin E, Lucas SR, similarities of asbestosis and cryptogenic fi- tional respiratory impairment due to asbestos Ancao MS, Sue DY: Differences between re- brosing alveolitis. Thorax 1988;43:708–714. exposure. J Occup Med 1982;24:685–689. maining ability and loss of capacity in maxi- mum aerobic impairment. Brazil J Med Biol 44 Wang XR, Araki S, Yano E, Wang MZ, Wang 50 Pearle J: Exercise performance and functional Research 1998;31:639–646. ZM: Dyspnea and exercise testing in workers impairment in asbestos-exposed workers. exposed to silica. Indust Health 1995;33:163– Chest 1981;80:701–705. Darryl Y. Sue, MD 171. Department of Medicine 51 Risk C, Epler GR, Gaensler EA: Exercise al- Harbor-UCLA Medical Center 45 Shih JF, Wilson JS, Broderick A, Watt JL, Gal- veolar-arterial oxygen pressure difference in in- Box 400, 1000 W. Carson St. vin JR, Merchant JA, Schwartz DA: Asbestos- terstitial lung disease. Chest 1984;85:69–74. Torrance, CA 90509-2910 (USA) induced pleural fibrosis and impaired exercise Tel. +1 310 222 2401, Fax +1 310 320 9688 physiology. Chest 1994;105:1370–1376. 52 Sue DY, Oren A, Hansen JE, Wasserman K: E-Mail [email protected] Lung function and exercise performance in smoking and nonsmoking asbestos-exposed workers. Am Rev Respir Dis 1985;132:612– 618. 230 Sue

Weisman IM, Zeballos RJ (eds): Clinical Exercise Testing. Prog Respir Res. Basel, Karger, 2002, vol 32, pp 231–241 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Role of Cardiopulmonary Exercise Testing in the Preoperative Evaluation for Lung Resection A.H. Diacon C.T. Bolliger Lung Unit, Department of Internal Medicine, University of Stellenbosch, Tygerberg, Cape Town, South Africa Summary planned extent of resection. Typically, exercise capacity after lobectomy is only slightly reduced, or not at all. Pneumonecto- To date, the majority of lung resection candidates in indus- my leads to a loss of approximately 20% in exercise capacity. In trialized nations suffer from lung cancer. Most of them are this chapter, we outline the principles and current recommen- former or current smokers and therefore at risk for chronic dations in the assessment of the lung resection candidate, with obstructive lung disease (COPD) and coronary artery disease special emphasis on the role of CPET. (CAD), conditions associated with increased surgical mortality and morbidity. On the other hand, lung cancer mortality Introduction approaches 100% when not treated surgically, and postopera- tive complications can, therefore, almost always be accepted Some decades ago, the majority of patients undergoing unless they are fatal. Although the assessment before lung lung surgery were suffering from the sequelae of tubercu- resection has been studied over many years, no single or com- losis. With the advent of effective antituberculous drugs bined parameter has been shown to sufficiently predict the sur- and the dramatic increase in the incidence of lung cancer gical outcome. In recent years, cardiopulmonary exercise test- worldwide, the majority of today’s lung resection candi- ing (CPET) is increasingly used to assess the surgical risk. dates suffer from lung cancer, for whom surgery is the Moreover, CPET in combination with split function studies is only curative form of treatment. Fortunately, advances in used to predict postoperative exercise capacity. Many candi- surgical techniques and postoperative care have made dates for lung resection, however, can be operated without lung resections accessible for older patients and those with complex testing like CPET and split function studies, which are limited cardiopulmonary reserves. Though a limited sur- not universally available and costly. Patients with normal or gical mortality may be acceptable in a disease with close only slightly impaired pulmonary function and no cardiovascu- to 100% mortality in itself, appropriate patient selection lar risk factors tolerate pulmonary resections up to a pneu- and risk management have consequently become a core monectomy without any problems, and exercise testing is not issue in lung cancer surgery. indicated in these cases. Candidates for lung resection with abnormal cardiopulmonary function should be further evaluat- The ideal parameter for the evaluation of the surgical ed with CPET and split function studies according to a prede- risk in lung resection would be technically simple, inex- fined protocol, and the individual risk assessed in respect to the

measurement of transfer factor of the lung for carbon monoxide (TLCO), patients at risk for complications can be identified and further investigated with CPET and, if necessary, split function studies for the estimation of pre- dicted postoperative (ppo) values. This evaluation pro- cess can be incorporated into a comprehensive algorithm in order to facilitate decision making and avoid unneces- sary testing. In this chapter, we discuss the single steps of the func- tional evaluation of the lung resection candidate with emphasis on the role of CPET. In addition, we present a validated practical algorithm for the functional assess- ment of the lung resection candidate. Fig. 1. Management of lung cancer. The circles represent the three Evaluation of Lung Resection Candidates entities that determine the management of a patient with lung cancer NSCLC = Non-small cell lung cancer; SCSL = small cell lung cancer; General Approach to the Lung Resection Candidate FEV1 = forced expiratory volume in the 1st second; TLCO = transfer Most lung resection candidates today are lung cancer factor for lung carbon monoxide; VO2max = maximal oxygen con- patients. In this condition, the initial three steps are sumption. With permission from Bolliger [72]. assessment of tumor type, tumor stage and operability (fig. 1). Some definitions of the terms used in this chapter pensive, highly reproducible and universally available. are briefly summarised below: Many a parameter has been considered for the selection of Tumor type refers mainly to the discrimination of patients for lung surgery. The use of vital capacity for the small-cell lung cancer (SCLC) versus non-small-cell lung prediction of the surgical risk was reported as early as in cancer (NSCLC). Only the latter is usually treated with 1955 [1]. Subsequently, various factors derived from his- lung resection. In SCLC, the mainstay of therapy is che- tory and symptoms, lung function, pulmonary hemody- mo- and radiotherapy. namics and radiological imaging have been proposed. Tumor staging of NSCLC according to the TNM sys- While invasive hemodynamic parameters are currently tem takes the extent of the tumor into account. In lung considered obsolete for classical lung resection surgery, cancer patients, resectability is compulsory for curative none of the other single parameters has shown a satisfac- surgery. Resectability means the possibility to surgically tory predictive value for the outcome after lung resec- remove the tumor completely. Preoperative surgical plan- tion. ning is normally based on radiological (chest X-ray, chest CT, positron emission tomography) and endoscopic Cardiopulmonary exercise testing (CPET), in contrast, (transbronchial needle aspiration, mediastinoscopy) stag- comprehends many specific physiologic factors in one ing, whereby local tumor extent and mediastinal lymph test. CPET evaluates directly the cardiopulmonary re- node involvement are assessed. The extent of planned serves and imitates the stress of undergoing surgery to a resection, usually lobectomy or pneumonectomy, sub- certain extent. In recent years, CPET was extensively stantially contributes to the final decision, whether to pro- studied and generally accepted for the prediction of peri- ceed to surgical resection or to evaluate alternative treat- operative risk and the long term postoperative outcome ment methods. In candidates with severely impaired lung after lung resection. However, most lung resection candi- function, the possibility of tissue sparing procedures such dates with normal or only slightly impaired lung function as atypical sublobar or sleeve resection must be weighed and good general condition can be resected safely without against the increased risk of local tumor recurrence. CPET, which is not universally available and costly. This article focuses on the functional reserves for Therefore, a stepwise approach to the functional assess- which the term operability is used. Certain authors make a ment of the lung resection candidate is practical and var- further distinction between ‘operability’ and ‘functional ious combinations of CPET with other parameters are resectability’. The first addresses the ability of the patient recommended. Beginning with simple clinical parame- to tolerate the stress of the operation and the early post- ters, basic pulmonary function testing (PFT), including 232 Diacon/Bolliger

operative phase (30 days). The latter looks at the perma- Table 1. Generally accepted risk factors for nent loss of cardiopulmonary function. However, the postoperative complications term ‘functional resectability’ has not been used widely and tends to lead to confusion. In practise, it is simpler Higher planned extent of resection and more generally accepted to use the term ‘operability’ Poor exercise performance for both these aspects and to reserve the term ‘resectabili- Poor predicted postoperative FEV1 ty’ for the surgical aspect concerned with the question Poor predicted postoperative TLCO whether the tumor can be totally removed based on ana- Age 1 70 years tomical information. pCO2 1 45 mm Hg Although the bulk of the literature describes predomi- ever, the cardiac risk of lung resection candidates has not nantly malignant disease, the principles outlined below received much attention in recent years, despite major are generally also applicable to non cancer cases. The advances in treatment options with coronary catheter extent of resection, however, is often less in benign dis- interventions (PTCA) or coronary artery bypass grafting ease, because radical excision respecting the rules of can- (CABG). Recent American guidelines for the assessment cer surgery is not necessary. of the cardiovascular risk in non-cardiac surgery combine various clinical risk predictors with the presumed risk of Most lung resection candidates are former or current the planned surgical procedure. Intrathoracic surgery has, smokers and at increased risk for ischemic heart disease according to these guidelines, an inherent moderate risk and chronic obstructive lung disease. In general, it is of of 1–5% for fatal or nonfatal cardiac complications [8]. utmost importance, that the patient is investigated and Therefore, every lung resection candidate should be eval- treated in the best possible condition. Patients with uated for potential cardiac disease and treated according- chronic obstructive lung disease should receive optimized ly before surgery is performed. antiobstructive therapy before undergoing surgery [2]. Smokers should be advised to quit and be offered counsel- Pulmonary Complications ing for this, preferably at least 8 weeks before surgery [3]. Postoperative pulmonary complications generally in- Repeat assessment of functional measurements may be clude atelectasis, pneumonia, pulmonary embolism, pro- necessary after improving medical therapy (i.e. in a pul- longed mechanical ventilation, CO2 retention, and death monary rehabilitation program). Due to the often rapid [4, 9, 10]. Pulmonary complications can frequently occur progress of disease in lung cancer patients the period or after major nonpulmonary surgery and contribute to mor- ‘window for functional improvement’ is usually limited to tality and length of hospital stay [11]. Poor exercise capac- 6–8 weeks. Insufficient treatment both reduces the pa- ity is a recognised predictor for pulmonary complications tient’s chances to be eligible for lung resection and in elective general surgery [12, 13]. Exercise testing, how- increases the risk of perioperative complications. ever, is not routinely used to assess the pulmonary risk in non lung resection surgery. Operative Morbidity and Mortality The risk of pulmonary complications is inherently The perioperative risk refers to the immediate or short higher in lung resection, because some functional lung tis- term risk leading to perioperative morbidity and mortali- sue is almost always resected. Contrary to non-pulmonary ty. Overall perioperative mortality rates for pneumonec- surgery, many risk factors have been studied for lung tomies ranges from 5% up to 17% and for lobectomies parenchymal resection. Generally accepted risk factors from !1% to 5% [4, 5]. An overall mortality rate of !5% are listed in table 1. Impaired lung mechanics, gas ex- is considered to be good and !2% excellent. Because lung change and exercise capacity are clearly associated with cancer is a disease with a mortality approaching 100% if poor surgical outcome, and the risk of pulmonary compli- not treated surgically, a certain degree of non-fatal postop- cations is higher with greater extent of resection [14–17]. erative complications can almost always be accepted. Ma- Older patients and patients with CO2 retention or im- jor complication categories in lung resection are cardiac paired TLCO, however, can undergo parenchymal resec- and pulmonary. tion with acceptable risk when assessed with exercise test- ing and should therefore not be excluded from surgery Cardiac Complications based on these factors alone [18–20]. An abnormal electrocardiogram (ECG) has been asso- ciated with postoperative myocardial infarction, arrhyth- mias and heart failure since the early 1960s [6, 7]. How- Preoperative Exercise Testing in Lung 233 Resection Candidates

Assessment of Operability value. However, prediction of postoperative functional Pulmonary Function Testing parameters is not a simple arithmetic matter. It must be Since the first report about the usefulness of vital taken into account that the disease process is normally capacity for predicting surgical outcome in 1955, many unevenly distributed and that the contribution of the lung static and dynamic lung function variables have been pro- tissue to be resected to overall lung function is variable. posed over time [1, 21]. However, only FEV1 has stood Furthermore, considerable compensation for lost lung pa- the test of time as the best single predictor of postopera- renchyma can occur over time. tive pulmonary complications [5, 22–25]. Very few perio- perative complications occurred in a large series of pa- Several methods to calculate ppo function have been tients selected based on FEV1 12 l for pneumonectomy, proposed. To date, technetium-99 macroaggragate perfu- FEV1 1 1 l for lobectomy and FEV1 10.6 l for segmentec- sion scans (split function studies) are widely in use for this tomy [5]. Regrettably, most reports on FEV1 focus on purpose. This technique allows quantification of the indi- absolute values rather that % of predicted (% pred). The vidual contribution of the parenchyma to be resected to latter approach has the advantage of taking age, sex and overall function, is easy to perform and generally avail- height into consideration, a notion that will probably gain able. The formula to calculate a ppo value based on perfu- importance with the current massive increase in female sion studies is, as proposed by Olsen: value ppo = preoper- lung cancer patients. ative value ! (1 – contribution of the parenchyma to be TLCO was suggested early as a risk predictor, and pro- resected) [34]. posed values were from !50% pred to !60% pred as pro- hibitive for pneumonectomy or major pulmonary resec- Many authors have confirmed the accuracy of the esti- tions [26, 27]. In a recent study, the correlation of lower mated functional loss in FEV1 and other lung function TLCO% pred with a higher rate of pulmonary complica- parameters after resection [35]. With the advent of the tions after lung resection was confirmed [28]. TLCO as an ppo concept, FEV1 ppo has received increasing interest. isolated parameter, however, did not gain much impor- Though not a consistent finding, various studies had bet- tance until the emergence of split function studies (see ter results for prediction of perioperative complications below) [4]. with FEV1 ppo than with baseline FEV1 [25, 36–38]. However, there is no consensus about ‘how much’ FEV1 Prediction of Postoperative Function with Split ppo is required as a minimum for safe resection, and val- Function Studies ues of 700–1,000 ml or 30% pred have been proposed [25, Lung resection candidates with normal lung function 36, 37]. tolerate resections up to an entire lung quite well and can, with certain restrictions, lead a normal postoperative life With the same formula as described by Olsen, TLCO in their private as well as in their professional environ- ppo can be calculated. A minimal value of TLCO ppo of ment. However, the majority of lung resection candidates 40% pred has been suggested. One report showed that are patients with cancer caused by smoking, which also both FEV1 ppo and TLCO ppo in % pred alone and in leads to COPD. Such patients have impaired functional combination were good indicators for postoperative com- reserves and are therefore at risk for permanent disability plications [4]. In another study, the product of FEV1 ppo after extensive resection of lung tissue. It has been shown, ! TLCO ppo in % pred was highly predictive of compli- that resections of not more than one lobe usually lead to a cations [39]. In contrast, baseline lung function was not mild early postoperative functional impairment with little predictive in both reports. permanent deficit in PFT (^10%) and preserved exercise capacity [29–31]. Pneumonectomies lead to a permanent Based on the literature of several decades, it is fair to loss of lung function of roughly one third and a moderate conclude that no single parameter has been established as drop in exercise capacity (B20%) [31–33]. the superior predictor of the perioperative risk. Presently, Based on the extent of planned resection and the pre- the most convincing concept is to use a combination of operative functional reserves, a forecast of the remaining functional tests. In general, the approach to use percent- postoperative proportion of functional parameters can be age of predicted rather than absolute values is probably undertaken. One might be tempted, for example for a superior to arbitrary absolute values, because differences lobectomy, to simply deduct a fifth of the preoperative in gender, age and body mass are taken into account. function to calculate the predicted postoperative (ppo) Cardiopulmonary Exercise Testing Cardiopulmonary exercise testing assesses both the cardiovascular and pulmonary reserves and, additionally, simulates the stress of a thoracic operation to a certain 234 Diacon/Bolliger

Table 2. Studies with minimal achievement Author Type of exercise Patients Conclusions Van Nostrand, 1968 [73] stair climbing 213 ^1 flight: high mortality Berggren, 1984 [74] cycle ergometry 82 670 years old 1 83 W during 6 min: low risk Bagg, 1984 [75] 12 min walking 22 no prediction of respiratory complications Olsen, 1991 [76] stair climbing 54 ! 3 flights: high complication rate Bolton, 1992 [77] stair climbing 70 63 flights: accept. risk for lobectomy 65 flights: accept. risk for pneumonectomy Holden, 1992 [48] 6 min walking 16 high risk 1 1,000 feet: acceptable risk Pollock, 1993 [41] stair climbing 16 high risk 1 44 steps: acceptable risk Pate, 1996 [37] stair climbing 31 men with COPD 1 4.6 flights correspond to (standardised) VO2max 1 20 ml/kg/min: low risk stair climbing 12 high risk 63 flights: acceptable risk extent. CPET has gained increasing importance since a tation, which is probably the reason why they have not report in 1982, when maximum exercise capacity was become very popular [42–45]. An algorithm for preopera- found to be superior to resting lung mechanics (FEV1, tive functional evaluation based on a submaximal test has FVC) in predicting operative mortality in a small number been proposed by a Japanese group [46]. of patients [40]. Various exercise protocols have been pro- posed, which can roughly be divided into three distinct Maximal Tests. The bulk of the literature has been groups by the type of exercise intensity demanded from about incremental maximal or symptom-limited tests us- the patient: minimal achievement, submaximal, and ing a ramp protocol. In such a test setting, the patient is maximal or symptom-limited tests. motivated to exercise until exhaustion or until an exercise induced symptom such as dyspnea or leg fatigue make it Minimal Achievement Tests. The principle of a mini- impossible to continue. The highest oxygen consumption mal achievement test is to declare a subject fit for a cer- achieved is called VO2max or peakVO2. Maximal proto- tain extent of resection, if a predefined minimal achieve- cols are easy to standardise, non-invasive, and VO2max is ment is accomplished. Tests such as stair climbing or reproducible [47]. Although VO2max has been shown to walking over 6 or 12 min have the advantage to be simple predict postoperative outcome in many reports, different and cheap. Different protocols have been proposed (ta- cut-off values for safe resections make general recommen- ble 2). Most of these tests are poorly standardised, how- dations somewhat difficult (table 3). Furthermore, com- ever, and allow identification of low risk patients rather parison of reports is complicated by incongruent defini- than exact risk stratification in compromised patients. tions of end points, differences in testing protocols, small The lack of adequate cardiac monitoring is an additional size or special selection of samples and inconsistent disadvantage, and the exact type of limitation can not be reporting in absolute values or percent of predicted [23, reliably detected. However, it was shown in patients with 28, 48, 49]. Nevertheless, a preoperative VO2max of COPD, that a standardised symptom-limited maximal 120 ml/kg/min is generally accepted to date as safe for stair climb test helps estimation of VO2max [41]. In the any resection up to pneumonectomy, and a value of absence of sophisticated equipment, such a test is certain- !10 ml/kg/min as predictive for a very high complication ly acceptable to select patients fit for resection. In border- rate, irrespective of the extent of resection. In analogy to line patients, however, more elaborate studying is neces- pulmonary function tests, VO2max values should be ex- sary. pressed in % predicted, which also takes age and sex into consideration. Submaximal Tests. To avoid the physical stress of a test to exhaustion, submaximal tests have been proposed. The postoperative loss of exercise capacity is usually In a submaximal test, the subject is exercised up to a pre- overestimated when looking at PFT values alone. After defined level, and the observed values are used to calcu- lobectomy, there is no long term loss in VO2max, whereas late the operative risk based on previous experience. after pneumonectomy a loss of 20–23% is observed [31, However, most of these tests involve invasive instrumen- 32, 50]. In recent years, the predicted postoperative Preoperative Exercise Testing in Lung 235 Resection Candidates

Table 3. Studies with preoperative exercise testing Author, year Patients Mortality Findings/remarks Recommendations Eugene, 1982 [40] 19 16% VO2max ! 1,000 ml: 75% mortality VO2max ! 1,000 ml: high risk Smith, 1984 [9] 22 0 VO2max ! 15 ml/kg/min: VO2max 1 20 ml/kg/min: low risk 100% complication rate VO2max ! 15 ml/kg/min: high risk Bechard, 1987 [10] 50 4% VO2max ! 10 ml/kg/min: 29% mortality VO2max ! 10 ml/kg/min: high risk Morice, 1992 [18] 37 (high risk) 0/8 pts offered surgery to patients with VO2max 1 15 ml/kg/min: acceptable risk VO2max 1 15 ml/kg/min Dales, 1993 [49] 117 !1% risk higher if: VO2max ! 1,250 ml use VO2max 1 1,250 ml and FEV1 ! 60% pred ventilatory reserve 1 25 l ventilatory reserve ! 25 l Epstein, 1993 [65] 42 2% VO2max ! 500 ml/m2/min: use risk index higher complication rate Bolliger, 1995 [51] 25 (high risk) 12% calculates ppo VO2max ppoVO2max ! 10 ml/kg/min: high risk Bolliger, 1995 [53] 80 4% VO2max ! 60% pred: VO2max ! 60% pred: high risk Pate, 1996 [37] 12 (high risk) 8.3% higher complicaton rate VO2max ! 60% pred: prohibitive 1 1 lobe 97 9.3% VO2max ! 43% pred: prohibitive any resection Richter Larsen, low complication rate despite high risk 1997 [78] 65 (high risk) 6.2% VO2max 1 10 ml/kg/min: acceptable risk complications higher in patients Ribas, 1998 [79] with VO2max ! 50% predicted VO2max ! 12 ml/kg/min: high risk Wyser, 1999 [52] 137 1.5% complications higher if paO2 use FEV1ppo, TLCO ppo, O2 desaturation decreases during exercise VO2max not predictive Brutsche, 2000 [17] 125 1.6% prospective evaluation VO2max 1 75% pred: low risk Wang, 2000 [80] 57 4% VO2max in 68 high-risk patients VO2max ppo 1 35% pred: low risk VO2max ! 60% pred: VO2max ! 60% pred: high risk higher complication rate VO2max 1 90% pred: low risk VO2max ! 15 ml/kg/min: use exercise-induced increase in TLCO, higher complication rate FEV1% pred, TLCO % pred, VO2max/kg VO2max = Maximal oxygen uptake; ppo = predicted postoperative. Adapted from Reilly [81]. VO2max (VO2max ppo) has been proposed to assess the Clinical Application surgical risk. Interestingly, the same simple formula used to predict FEV1 ppo and TLCO ppo was found useful for Despite the availability of modern tests like split func- this prediction [51]. However, VO2max ppo and its pro- tion studies and CPET, most patients without cardiac dis- posed value of 10 ml/kg/min as prohibitive for any resec- ease can undergo resections up to pneumonectomy based tion still warrant confirmation in a large number of on simple PFT. To avoid unnecessary cost, it is important patients [52]. An additional advantage of calculating to take a stepwise approach to the evaluation of the lung VO2max ppo is the estimation of a patient’s postoperative resection candidate. An ideal tool for this purpose is an professional working capacity. algorithm, including all necessary preoperative variables (PFT, ppo function, and CPET). The algorithm should In conclusion, symptom-limited maximal exercise test- result in a clear-cut recommendation, whether lung resec- ing has the advantage of good standardisation and repro- tion is possible and if yes, to what extent. ducibility. The equipment is affordable and testing is not invasive, and the value of VO2max in predicting perioper- CPET is now recognised widely as the test of choice for ative complications is clearly established. patients with abnormal basic lung function values. It is 236 Diacon/Bolliger

Fig. 2. Proposed algorithm for the assess- ment of operability of lung resection candi- dates. Patients undergo successive steps from top to bottom, until they qualify for varying extents of resection or are deemed inoperable. The ‘safety loop’ for patients with cardiac problems is indicated in the upper left-hand corner. The dashed line leading from exercise testing back to the car- diac workup is for patients with a negative cardiac history and a normal ECG, who show symptoms or signs of ischemia during exercise. * Consider eligibility for combined tumor resection and lung volume reduction surgery in carefully selected patients. TI = Thallium; TC = technetium; VO2max = maximal oxygen consumption during exer- cise; ppo = predicted postoperative; FEV1 = forced expiratory volume in the 1st second; TLCO = transfer factor for lung carbon mon- oxide. With permission from Bolliger and Perruchoud [35]. not clear, however, what extent of lung function impair- complication rates from 4 to 1.5% and from 20 to 11%, ment can still be tolerated for lung resection without addi- respectively, while the proportion of patients deemed tional CPET. inoperable remained unchanged (fig. 2). The algorithm presented in this chapter was proposed Patients who have a negative cardiac history and a nor- by Bolliger et al. [53] after evaluation of a consecutive mal ECG can undergo lung resection without further series of 80 lung resection candidates with multiple assessment of the cardiac risk. Any patient with active or regression analysis. The best predictor for postoperative suspected cardiac disease should first undergo a thorough complications was VO2max % pred, which was slightly cardiac work-up and, if necessary, even coronary bypass better than VO2max absolute. Of 9 patients with VO2max surgery in case of ischemic heart disease [54]. Only of !60% pred, 8 had complications, including all 3 who patients whose cardiac condition is amenable to treat- died. In a high risk subgroup of 25 patients who under- ment can undergo further investigation for pulmonary went split function studies, VO2max % pred and VO2max resection. ppo were predictive, while FEV1 ppo and TLCO ppo were not significantly different in patients with or without Since resection of the diseased lung up to pneumonec- complications [51]. The initially suggested algorithm was tomy seldom results in a functional loss of 150% [55], and slightly amended after prospective evaluation in a follow- postoperative values of 140% for FEV1 and TLCO are up series of 137 consecutive patients [52]. Adherence to safe [4, 39], the algorithm allows resections up to a pneu- this algorithm resulted in a reduction of mortality and monectomy without any further tests, if FEV1 and TLCO are both 180% predicted. If either FEV1 or TLCO is Preoperative Exercise Testing in Lung 237 Resection Candidates

!80% pred, exercise testing with the measurement of again after treatment. To our knowledge, no published VO2max is performed. Rarely, exercise testing will pick data on the issue of operability after neoadjuvant treat- up ischemic heart disease in patients with a negative car- ment are available to date, so that each case should be diac history and a normal ECG. This will also lead to a reassessed for operability as well as for resectability before cardiac work-up (interrupted line in the algorithm). If surgery. VO2max is 175% pred or 120 ml/kg/min, patients qualify for resection up to a pneumonectomy; if it is !40% pred Controversial Issues or 10 ml/kg/min they are inoperable. Patients Unable to Exercise All patients with VO2max values in between undergo A limitation of exercise testing is that not all patients split-function studies with the aid of a pulmonary perfu- can exercise with standardised equipment such as cycle sion scan to determine their predicted postoperative func- ergometers and treadmills. This group is very heteroge- tion (ppo-function). Firstly, FEV1 ppo and TLCO ppo are neous. Some patients are able to perform an exercise test analyzed; if the values for both parameters are !40% with alternative equipment, i.e. a rowing ergometer. pred, patients are deemed inoperable. If either one is Other patients will have to be evaluated with PFT and 140% pred, then VO2max ppo becomes the decisive fac- split function studies alone. However, patients unable to tor. With a VO2max ppo of either !35% pred or !10 ml/ exercise because of orthopedic problems, vascular dis- kg/min patients are also deemed inoperable; whereas ease or chronic neurologic disorders generally have a patients with a VO2max ppo 135% pred and 110 ml/kg/ clearly decreased level of fitness and eligibility for major min are operable up to the extent which was used for the surgery must be determined with caution. In one report, prediction of postoperative function. inability to perform cycle ergometry was found to be a independent predictor of bad outcome after lung resec- Another algorithm focused on high risk patients with tion [62]. FEV1 or TLCO of !60% pred has most recently been pro- posed by Weisman, aiming to reduce cost for additional Quantitative CT Scanning testing without jeopardizing the overall outcome [56]. A new approach to the prediction of postoperative Naturally, different approaches to the use of CPET are function is quantitative computed tomography. One possible, and local availability of testing facilities or cost group found excellent correlation for FEV1 and FVC in a factors may influence the choice of a specific method. group of 34 patients [63]. In a most recent study, the use- fulness of quantitative CT scanning comparing various In many institutions, an interdisciplinary discussion of techniques for estimating ppo function was confirmed, patients proposed for lung resection, involving pulmono- although perfusion studies remained the most accurate logist, radiologist, thoracic surgeon and (radio)-oncolog- [64]. Because lung cancer patients regularly undergo CT ist, has proven very useful. Based on cancer staging, scanning for preoperative staging, this technique has the resectability and operability estimates, a comprehensive potential to replace perfusion scans in the evaluation of management plan should be established. The operating the majority of patients relegating additional perfusion surgeon must be informed about the extent of resection scans for patients with moderate-severe impairment of functionally possible, and in selected cases atypical resec- cardiopulmonary reserves only. tion techniques must be considered to optimise risk man- agement. Cardiopulmonary Risk Index Epstein et al. [65] proposed a cardiopulmonary risk Establishment of functional operability is a dynamic index (CPRI) for the preoperative risk-assessment of lung process and investigations must sometimes be repeated to resection candidates. The CPRI assigns points for various keep up with the current state of each patient. Especially cardiac and pulmonary risk factors. The CPRI has been with preoperative (neoadjuvant) treatment protocols for compared to VO2max and both have shown to be predic- NSCLC, several weeks can go by between initial assess- tive for perioperative complications. Other authors, how- ment and surgery. The data on surgical outcome after ever, could not confirm the predictive value of the CPRI neoadjuvant treatment are conflicting, and a negative [66]. Furthermore, some parameters used are rather sub- impact on the operative risk seems possible, particularly jective. Although some prediction of risk seems possible, after combined chemo-radiotherapy with irradation doses of 145 Gy [57–61]. Based on our experience, preoperative chemotherapy can lead to functional deterioration and inoperability, but the converse can also be true, for exam- ple if a previously obstructed bronchus becomes patent 238 Diacon/Bolliger

the value of the CPRI for the selection of pulmonary spective studies including lung cancer patients with ana- resection candidates is not clearly defined. tomically resectable tumors previously deemed inopera- ble due to limited cardiopulmonary reserves. New Horizons with Lung Volume Reduction Surgery? Lung volume reduction surgery (LVRS) is performed Conclusion in selected extremely symptomatic emphysematous pa- tients in order to improve lung function, quality of life CPET has gained an important role in the preoperative and exercise capacity [67]. During LVRS, the most dys- assessment of lung resection candidates. While PFTs and functional parts of one or both lungs are removed mainly cardiac studies adequately assess specific risk compo- by means of wedge resection. Because emphysema and nents, CPET allows simultaneous measurement of pul- lung cancer share cigarette smoking as a common risk fac- monary and cardiovascular parameters in a single test. tor, coincidence can occur. Therefore, the combination of With additional pulmonary perfusion scans, accurate pre- cancer surgery with LVRS, which improves overall lung dicted postoperative values for FEV1, TLCO and, most function postoperatively, might become a promising new recently, VO2max can be calculated. In combination with treatment option in selected lung cancer patients pre- the planned extent of resection, the individual risk of the viously deemed inoperable. However, most tumors re- pulmonary resection candidate can be assessed. sected to date in combination with LVRS have been detected incidentally during work-up for LVRS. There- The current trend seems to favor split-function studies fore, these tumors are in a very early stage and prognosis and exercise testing with the measurement of VO2max. after surgical resection is favorable. Early reports are The majority of reports found VO2max a good indepen- promising, but data for long term survival are not avail- dent risk predictor. Maximal exercise tests are fairly easy able to date [68, 69]. In lung cancer surgery, it is well to standardise, and, in our opinion, VO2max represents known that resections of less than a lobe are associated the single best parameter to evaluate both the pulmonary with a smaller chance of curative resection even in T1, as well as the cardiovascular reserves. Excessive testing N0, stage I NSCLC [70, 71]. Whether the risk of less radi- can be avoided by adherence to an integrated algorithm, cal surgery outweighs the benefit from LVRS is currently providing rational decision-making based on established not known. This problem warrants investigation in pro- risk factors. References 1 Gaensler EA, Cugell DW, Lindgren I: The role 7 Mittman C: Assessment of operative risk in 11 Lawrence VA, Dhanda R, Hilsenbeck SG, Page of pulmonary insufficiency in mortality and thoracic surgery. Am Rev Respir Dis 1961;84: CP: Risk of pulmonary complications after el- invalidism following surgery for pulmonary tu- 197–207. ective abdominal surgery. Chest 1996;110: berculosis. J Thorac Surg 1955;29:163. 744–750. 8 Eagle KA, Brundage BH, Chaitman BR, Ewy 2 Smetana GW: Preoperative pulmonary evalua- GA, Fleisher LA, Hertzer NR, Leppo JA, Ryan 12 Williams-Russo P, Charlson ME, MacKenzie tion. N Engl J Med 1999;340:937–944. T, Schlant RC, Spencer WH 3rd, Spittell JA Jr, CR, Gold JP, Shires GT: Predicting postopera- Twiss RD, Ritchie JL, Cheitlin MD, Gardner tive pulmonary complications. Is it a real prob- 3 Warner MA, Offord KP, Warner ME, Lennon TJ, Garson A Jr, Lewis RP, Gibbons RJ, lem? Arch Intern Med 1992;152:1209–1213. RL, Conover MA, Jansson-Schumacher U: O’Rourke RA, Ryan TJ: Guidelines for peri- Role of preoperative cessation of smoking and operative cardiovascular evaluation for non- 13 Older P, Hall A, Hader R: Cardiopulmonary other factors in postoperative pulmonary com- cardiac surgery: Report of the American Col- exercise testing as a screening test for perioper- plications: A blinded prospective study of coro- lege of Cardiology/American Heart Association ative management of major surgery in the nary artery bypass patients. Mayo Clin Proc Task Force on Practice Guidelines. Committee elderly. Chest 1999;116:355–362. 1989;64:609–616. on Perioperative Cardiovascular Evaluation for Noncardiac Surgery. Circulation 1996;93: 14 Lockwood P: Lung function test results and the 4 Markos J, Mullan BP, Hillman DR, Musk AW, 1278–1317. risk of post-thoracotomy complications. Respi- Antico VF, Lovegrove FT, Carter MJ, Finu- ration 1973;30:529–542. cane KE: Preoperative assessment as a predic- 9 Smith TP, Kinasewitz GT, Tucker WY, Spill- tor of mortality and morbidity after lung resec- ers WP, George RB: Exercise capacity as a pre- 15 Moghissi K, Connolly CK: Resection rates in tion. Am Rev Respir Dis 1989;139:902–910. dictor of post-thoracotomy morbidity. Am Rev lung cancer patients. Eur Respir J 1996;9:5–6. Respir Dis 1984;129:730–734. 5 Miller JI Jr: Physiologic evaluation of pulmo- 16 Damhuis RA, Schutte PR: Resection rates and nary function in the candidate for lung resec- 10 Bechard D, Wetstein L: Assessment of exercise postoperative mortality in 7,899 patients with tion. J Thorac Cardiovasc Surg 1993;105:347– oxygen consumption as preoperative criterion lung cancer. Eur Respir J 1996;9:7–10. 351; discussion 351–342. for lung resection. Ann Thorac Surg 1987;44: 344–349. 17 Brutsche MH, Spiliopoulos A, Bolliger CT, 6 Didolkar MS, Moore RH, Takita H: Evalua- Licker M, Frey JG, Tschopp JM: Exercise ca- tion of the risk in pulmonary resection for pacity and extent of resection as predictors of bronchogenic carcinoma. Am J Surg 1974;127: surgical risk in lung cancer. Eur Respir J 2000; 700–703. 15:828–832. Preoperative Exercise Testing in Lung 239 Resection Candidates

18 Morice RC, Peters EJ, Ryan MB, Putnam JB, 34 Olsen GN, Block AJ, Tobias JA: Prediction of 49 Dales RE, Dionne G, Leech JA, Lunau M, Ali MK, Roth JA: Exercise testing in the evalu- postpneumonectomy pulmonary function us- Schweitzer I: Preoperative prediction of pul- ation of patients at high risk for complications ing quantitative macroaggregate lung scanning. monary complications following thoracic sur- from lung resection. Chest 1992;101:356–361. Chest 1974;66:13–16. gery. Chest 1993;104:155–159. 19 Gerson MC, Hurst JM, Hertzberg VS, Doogan 35 Bolliger CT, Perruchoud AP: Functional evalu- 50 Corris PA, Ellis DA, Hawkins T, Gibson GJ: PA, Cochran MB, Lim SP, McCall N, Adolph ation of the lung resection candidate. Eur Res- Use of radionuclide scanning in the preopera- RJ: Cardiac prognosis in noncardiac geriatric pir J 1998;11:198–212. tive estimation of pulmonary function after surgery. Ann Intern Med 1985;103:832–837. pneumonectomy. Thorax 1987;42:285–291. 36 Olsen GN, Block AJ, Swenson EW, Castle JR, 20 Bolliger CT, Soler M, Stulz P, Gradel E, Mull- Wynne JW: Pulmonary function evaluation of 51 Bolliger CT, Wyser C, Roser H, Soler M, Perru- er-Brand J, Elsasser S, Gonon M, Wyser C, Per- the lung resection candidate: A prospective choud AP: Lung scanning and exercise testing ruchoud AP: Evaluation of high-risk lung resec- study. Am Rev Respir Dis 1975;111:379–387. for the prediction of postoperative perfor- tion candidates: Pulmonary haemodynamics mance in lung resection candidates at increased versus exercise testing. A series of five patients. 37 Pate P, Tenholder MF, Griffin JP, Eastridge risk for complications. Chest 1995;108:341– Respiration 1994;61:181–186. CE, Weiman DS: Preoperative assessment of 348. the high-risk patient for lung resection. Ann 21 American College of Physicians: Preoperative Thorac Surg 1996;61:1494–1500. 52 Wyser C, Stulz P, Soler M, Tamm M, Muller- pulmonary function testing. Ann Intern Med Brand J, Habicht J, Perruchoud AP, Bolliger 1990;112:793–794. 38 Kristersson S, Lindell SE, Svanberg L: Predic- CT: Prospective evaluation of an algorithm for tion of pulmonary function loss due to pneu- the functional assessment of lung resection can- 22 Boushy SF, Billig DM, North LB, Helgason monectomy using 133Xe-radiospirometry. didates. Am J Respir Crit Care Med 1999;159: AH: Clinical course related to preoperative and Chest 1972;62:694–698. 1450–1456. postoperative pulmonary function in patients with bronchogenic carcinoma. Chest 1971;59: 39 Pierce RJ, Copland JM, Sharpe K, Barter CE: 53 Bolliger CT, Jordan P, Soler M, Stulz P, Gradel 383–391. Preoperative risk evaluation for lung cancer E, Skarvan K, Elsasser S, Gonon M, Wyser C, resection: Predicted postoperative product as a Tamm M, et al: Exercise capacity as a predictor 23 Colman NC, Schraufnagel DE, Rivington RN, predictor of surgical mortality. Am J Respir of postoperative complications in lung resec- Pardy RL: Exercise testing in evaluation of Crit Care Med 1994;150:947–955. tion candidates. Am J Respir Crit Care Med patients for lung resection. Am Rev Respir Dis 1995;151:1472–1480. 1982;125:604–606. 40 Eugene J, Brown SE, Light RW, Milne NE, Stemmer EA: Maximum oxygen consumption: 54 Foster ED, Davis KB, Carpenter JA, Abele S, 24 Keagy BA, Lores ME, Starek PJ, Murray GF, A physiologic guide to pulmonary resection. Fray D: Risk of noncardiac operation in pa- Lucas CL, Wilcox BR: Elective pulmonary lo- Surg Forum 1982;33:260–262. tients with defined coronary disease: The Coro- bectomy: Factors associated with morbidity nary Artery Surgery Study (CASS) registry ex- and operative mortality. Ann Thorac Surg 41 Pollock M, Roa J, Benditt J, Celli B: Estima- perience. Ann Thorac Surg 1986;41:42–50. 1985;40:349–352. tion of ventilatory reserve by stair climbing. A study in patients with chronic airflow obstruc- 55 Hall DR: Regional lung function after pneu- 25 Wernly JA, DeMeester TR, Kirchner PT, Mye- tion. Chest 1993;104:1378–1383. monectomy. Thorax 1974;29:425–431. rowitz PD, Oxford DE, Golomb HM: Clinical value of quantitative ventilation-perfusion 42 Pierce RJ, Pretto JJ, Rochford PD, McDonald 56 Weisman IM: Cardiopulmonary exercise test- lung scans in the surgical management of bron- CF, Hanan JA, Barter CE: Lobar occlusion in ing in the preoperative assessment for lung chogenic carcinoma. J Thorac Cardiovasc Surg the preoperative assessment of patients with resection surgery. Semin Thorac Cardiovasc 1980;80:535–543. lung cancer. Br J Dis Chest 1986;80:27–36. Surg 2001;13:116–125. 26 Nagasaki F, Flehinger BJ, Martini N: Compli- 43 Nakagawa K, Nakahara K, Miyoshi S, Kawa- 57 Bunn PA Jr, Mault J, Kelly K: Adjuvant and cations of surgery in the treatment of carcino- shima Y: Oxygen transport during incremental neoadjuvant chemotherapy for non-small cell ma of the lung. Chest 1982;82:25–29. exercise load as a predictor of operative risk in lung cancer: A time for reassessment? Chest lung cancer patients. Chest 1992;101:1369– 2000;117:119S–122S. 27 Ferguson MK, Little L, Rizzo L, Popovich KJ, 1375. Glonek GF, Leff A, Manjoney D, Little AG: 58 Siegenthaler MP, Pisters KM, Merriman KW, Diffusing capacity predicts morbidity and mor- 44 Miyoshi S, Nakahara K, Ohno K, Monden Y, Roth JA, Swisher SG, Walsh GL, Vaporciyan tality after pulmonary resection. J Thorac Car- Kawashima Y: Exercise tolerance test in lung AA, Smythe WR, Putnam JB Jr: Preoperative diovasc Surg 1988;96:894–900. cancer patients: The relationship between exer- chemotherapy for lung cancer does not increase cise capacity and postthoracotomy hospital surgical morbidity. Ann Thorac Surg 2001;71: 28 Wang J, Olak J, Ultmann RE, Ferguson MK: mortality. Ann Thorac Surg 1987;44:487–490. 1105–1111. Assessment of pulmonary complications after lung resection. Ann Thorac Surg 1999;67: 45 Olsen GN, Weiman DS, Bolton JW, Gass GD, 59 Schilder RJ, Goldberg M, Millenson MM, 1444–1447. McLain WC, Schoonover GA, Hornung CA: Movsas B, Rogatko A, Rogers B, Langer CJ: Submaximal invasive exercise testing and Phase II trial of induction high-dose chemo- 29 Ali MK, Mountain CF, Ewer MS, Johnston D, quantitative lung scanning in the evaluation for therapy followed by surgical resection and ra- Haynie TP: Predicting loss of pulmonary func- tolerance of lung resection. Chest 1989;95: diation therapy for patients with marginally tion after pulmonary resection for bronchogen- 267–273. resectable non-small cell carcinoma of the lung. ic carcinoma. Chest 1980;77:337–342. Lung Cancer 2000;27:37–45. 46 Nakahara K, Monden Y, Ohno K, Miyoshi S, 30 Loddenkemper R, Gabler A, Gobel D: Criteria Maeda H, Kawashima Y: 1985:A method for 60 Sonett JR, Krasna MJ, Suntharalingam M, of functional operability in patients with bron- predicting postoperative lung function and its Schuetz J, Doyle LA, Lilenbaum R, Gamliel Z: chial carcinoma: Preoperative assessment of relation to postoperative complications in pa- Safe pulmonary resection after chemotherapy risk and prediction of postoperative function. tients with lung cancer. 1992 update. Ann Tho- and high-dose thoracic radiation. Ann Thorac Thorac Cardiovasc Surg 1983;31:334–337. rac Surg 1992;54:1016–1017. Surg 1999;68:316–320. 31 Bolliger CT, Jordan P, Soler M, Stulz P, Tamm 47 Janicki JS, Gupta S, Ferris ST, McElroy PA: 61 Rusch VW, Benfield JR: Neoadjuvant therapy M, Wyser C, Gonon M, Perruchoud AP: Pul- Long-term reproducibility of respiratory gas for lung cancer: A note of caution. Ann Thorac monary function and exercise capacity after exchange measurements during exercise in pa- Surg 1993;55:820–821. lung resection. Eur Respir J 1996;9:415–421. tients with stable cardiac failure. Chest 1990; 97:12–17. 62 Epstein SK, Faling LJ, Daly BD, Celli BR: 32 Van Mieghem W, Demedts M: Cardiopulmo- Inability to perform bicycle ergometry predicts nary function after lobectomy or pneumonec- 48 Holden DA, Rice TW, Stelmach K, Meeker increased morbidity and mortality after lung tomy for pulmonary neoplasm. Respir Med DP: Exercise testing, 6-min walk, and stair resection. Chest 1995;107:311–316. 1989;83:199–206. climb in the evaluation of patients at high risk for pulmonary resection. Chest 1992;102: 33 Pelletier C, Lapointe L, LeBlanc P: Effects of 1774–1779. lung resection on pulmonary function and exer- cise capacity. Thorax 1990;45:497–502. 240 Diacon/Bolliger