Evidence Based Chronic Pain Management

Chapter 22 evaluated sustained pain-free status (pain free after 29–34%). In contrast, the mean TG were inferior to 2 hours, no use of rescue medication and no recur- sumatriptan 100 mg for naratriptan 2.5 mg (TG 22%; rence within 24 hours). In addition, tolerability versus 95% CI 18–26%), and for frovatriptan (TG 16%; 95 CI placebo was evaluated in these systematic reviews. 8–25%). Similar results were found in two other meta- analyses [39, 49, 51]. For details, see Tfelt-Hansen 2006 Systematic reviews of one drug for [52]. There are minor differences most likely due to the treatment of migraine differences in the RCT included in the meta-analyses Naratriptan 2.5 mg was superior to placebo (number [37, 39, 49]. needed to treat (NNT) for headache relief 4.6) in a systematic review [43]. For a comment on the Based on these three systematic reviews, one can magnitude of the effect of naratriptan, see below. conclude that subcutaneous sumatriptan 6 mg is the Rizatriptan 10 mg was superior to placebo with a most effective triptan. Eletriptan 80 mg (for headache NNT of 2.7 (95% confidence interval (CI) 2.4–2.9) relief and pain-free status), and rizatriptan 10 mg (for [49]. Rizatriptan 10 mg is a first-line drug for the pain-free status) are somewhat better than sumatriptan treatment of migraine attacks. Almotriptan 12.5 mg 100 mg, whereas naratriptan 2.5 mg and frovatriptan was superior to placebo with a NNT of 3.4 [44]. Both 2.5 mg are inferior to sumatriptan 100 mg. eletriptan 40 mg and 80 mg were superior to placebo [41]. Eletriptan 40 mg is a first-line treatment option A large meta-analysis of 53 RCT with oral triptans in acute migraine attacks with a NNT of 2.9 (95% was published in The Lancet [38] and later in detail CI 2.6–3.3 [41] and eletriptan 80 mg (NNT 2.6, 95% in Cephalalgia [12]. Unpublished studies were also CI 2.4–3.0) can be tried in especially severe migraine included in this meta-analysis [38]. Some of the results attacks. In one systematic review the efficacy param- are shown in Figure 22.1. The authors’ interpretation eter pain-free response at 2 hours was used [42] and of the meta-analysis combined with evaluated com- sumatriptan 100 mg was superior to placebo with parative RCT are shown in Table 22.1. Headache relief NNT of 5.1 (95% CI 3.9–7.1). Sumatriptan 50 mg at 2 hours was the primary per protocol endpoint was not superior to placebo for this parameter but in nearly all triptan RCT and as illustrated in Figure only a small number of patients (n = 124) were ana- 22.1A, headache relief after 2 hours, compared with lyzed [42]. In a large meta-analysis, sumatriptan 100 mg sumatriptan, was higher for rizatriptan 10 mg 50 mg was superior to placebo with a therapeutic and eletriptan 80 mg, lower for naratriptan 2.5 mg gain (TG) (percentage relief with active drug minus and frovatriptan 2.5 mg. For the placebo-subtracted percentage relief with placebo) of 18% for pain-free response, the same as TG, only eletriptan 80 mg was status [12]. In a recent large RCT, sumatriptan 50 mg superior to sumatriptan 100 mg. Pain-free status at (49%) taken in the mild phase of a migraine attack 2 hours, recommended as the primary efficacy meas- was superior to placebo (24%) for pain-free status at ure by the International Headache Society [53], was 2 hours [50]. a secondary endpoint in most RCT. Compared with sumatriptan 100 mg, naratriptan 2.5 mg showed lower Meta-analyses of drugs used for the absolute pain-free rates whereas eletriptan 80 mg, acute treatment of migraine almotriptan 12.5 mg, and rizatriptan 10 mg showed One systematic review presented a meta-analysis of higher values. However, only eletriptan 80 mg and riza- the seven oral triptans (sumatriptan, zolmitriptan, triptan 10 mg showed higher values than sumatriptan naratriptan, rizatriptan, almotriptan, eletriptan, and for TG for pain-free status (placebo-subtracted val- frovatriptan) and of subcutaneous, intranasal, and ues) (see Fig. 22.1B). Sustained pain-free status was rectal sumatriptan [37]. The meta-analysis was based higher for rizatriptan 10 mg (26%, 95% CI 24–27%), on headache relief, and the mean TG with 95% CI eletriptan 80 mg (25%, 95% CI 23–27%), and almo- was calculated. Based on this meta-analysis, subcuta- triptan 12.5 mg (27%, 95% CI 23–30%) compared neous sumatriptan 6 mg (TG 51%; 95% CI 49–53%) to sumatriptan 100 mg (20%, 95% CI 18–21%). and eletriptan 80 mg (TG 42%; 95% CI 37–47%) were For adverse events, sumatriptan 100 mg had a mean superior to sumatriptan 100 mg (TG 32%; 95% CI placebo-subtracted rate of any adverse events of 13% (95% CI 8–18%). Rates for the other triptans overlap, except for lower values for almotriptan 12.5 mg and 282

Headache Response at 2 h Pain free at 2 h Placebo-subtracted Absolute (%) Placebo-subtracted Absolute (%) 0 20 40 60 80 0 10 20 30 40 50 25 mg 25 mg Sumatriptan 50 mg Sumatriptan 50 mg 100 mg 100 mg 2.5 mg 2.5 mg Zolmitriptan Zolmitriptan 5 mg 5 mg Naratriptan 2.5 mg Naratriptan 2.5 mg 5 mg 5 mg Rizatriptan Rizatriptan 10 mg 10 mg Eletriptan 20 mg 20 mg 40 mg Eletriptan 40 mg 80 mg 80 mg Almotriptan 12.5 mg Frovatriptan 2.5 mg Almotriptan 12.5 mg (a) (b) Figure 22.1 Headache response (A) and pain-free status after 2 hours (B) after seven oral triptans. The shaded area indicates the 95% CI for sumatriptan 100 mg for both absolute responses and placebo-subtracted results. From Ferrari et al. [38] with permission from the publisher. Table 22.1 Comparison of the main efficacy and tolerability measures for oral triptans versus sumatriptan 100 mg based on the results of the meta-analysis and direct comparative trials, modified from Ferrari et al. 2002 [12]. In bold parentheses is shown my personal judg- ment of these items for rizatriptan, eletriptan, and almotriptan based on the meta-analysis [12] and some later published comparative RCT [60, 61] Initial 2-h relief Sustained Tolerability pain-free status Sumatriptan 50 mg ϭ ϭ ϭ Ϫ ϭ/Ϫ ϩ Sumatriptan 25 mg ϭ ϭ ϭ Zolmitriptan 2.5 mg ϭ ϭ ϭ Zolmitriptan 5 mg Ϫ Ϫ ϩϩ Naratriptan 2.5 mg ϭ ϭ ϭ Rizatriptan 5 mg ϩ(ϭ/ϩ) ϩ(Ϫ/ϭ) ϭ(ϭ) Rizatriptan 10 mg Ϫ ϭ Eletriptan 20 mg ϭ/ϩ(ϩ) ϭ/Ϫ(ϩ/ϭ) ϭ(ϭ) Eletriptan 40 mg ϩ(ϩ)(ϩ) ϩ(ϩ) Ϫ(Ϫ) Eletriptan 80 mg ϭ(ϭ) ϩ(ϭ) ϩϩ(ϩϩ) Almotriptan 12.5 mg ϭ, no difference when compared with 100 mg sumatriptan; ϩ, better than sumatriptan; Ϫ, inferior when compared with sumatriptan. 283

Chapter 22 naratriptan 2.5 mg. The rates for almotriptan and and TG for pain-free status as sumatriptan 100mg (see naratriptan did not differ from placebo. Fig. 22.1). The authors of the meta-analysis concluded that In one large comparative RCT [61] published after at marketed doses, all oral triptans were effective the meta-analysis [38], almotriptan 12.5 mg (18%) and well tolerated. Rizatriptan 10 mg, eletriptan was inferior to sumatriptan 50 mg (25%) for pain-free 80 mg, and almotriptan 12.5 mg provide the high- status after 2 hours. Apparently, almotriptan 12.5 mg est likelihood of consistent success [38]. I agree is better than sumatriptan 100 mg for sustained pain- that all triptans are more effective than placebo but free status, based on the meta-analysis [12, 38]. This the TG for both naratriptan 2.5 mg (22%) [38] and was, however, not the case in a comparative RCT [12] frovatriptan 2.5 mg (18%) [38] are low (see Fig. and in a cross-over RCT [61] almotriptan 12.5 mg 22.1A), and lower than the TG for the combina- had a lower sustained pain-free response (12.9%) tion of aspirin plus metoclopramide (31%) [40]. than sumatriptan 50 mg (17.6%) [62]. The better These two triptans are also more costly than this tolerability of almotriptan 12.5 mg than sumatriptan combination. was confirmed in two trials [61, 63]. In one RCT, treatment-related AE occurred more frequently with Rizatriptan 10 mg is rated ϩ for initial 2 hour sumatriptan 50 mg (15.5%) than with almotriptan relief compared with sumatriptan 100 mg (see Table 12.5 mg (9.1%) [61] and in another RCT, the inci- 22.1). The TG for headache relief for rizatriptan dence of AE was higher with sumatriptan 100 mg 10 mg was similar to sumatriptan (see Fig. 22.1A) (22.2%) than with almotriptan 12.5 mg (8.7%). The and in a large comparative RCT, the two drugs were incidence of AE was the same for almotriptan as for comparable for this efficacy measure [54]. For pain- placebo 12.5 mg (6.1%) [63]. free status after 2 hours, however, the TG was higher for rizatriptan 10 mg than for sumatriptan 100 mg. My personal ratings of rizatriptan 10 mg , elet- The same was found in a comparative RCT [54]. In riptan 40 mg and 80 mg, and almotriptan 12.5 mg addition, rizatriptan 10 mg (n ϭ 1114) was supe- versus sumatriptan 100 mg are shown in bold rior (5%: 95% CI 0.6–9%) to sumatriptan 50 mg parentheses in Table 22.1. Rizatriptan 10 mg should (n ϭ 1116) for pain-free status after 2 hours in a have ϩ/ϭ rating for initial 2-hour relief and for sus- combined analysis of two trials [55, 56]. Rizatriptan tained pain-free status whereas the rest of the rating 10 mg was superior to sumatriptan 100 mg for sus- is unchanged. Eletriptan 40 mg should have ϩ for tained pain-free status [38] but this was not the case initial relief based on the three comparative RCT [60] in the comparative RCT [12]. and the same is the case for sustained pain-free status. Eletriptan 80 mg is unchanged. Almotriptan 12.5 mg After the publication of the meta-analysis [12, 38], a is in my view comparable to sumatriptan 100 mg for new large placebo-controlled RCT comparing eletriptan initial relief and sustained pain-free status but its bet- 40mg and sumatriptan 100mg was published [57]. This ter tolerability, comparable to placebo, has been con- study together with two previous RCT [58, 59] consti- firmed in two RCTs [61, 63]. tutes a large head-to-head comparative database [60]. The headache relief rate after 2 hours was higher for Recent results of RCTs with triptans eletriptan 40mg (67%) than for sumatriptan 100 mg and comparisons with ergotamine (57%) [60] and the 2-hour pain-free response was also In two recently published RCT [64] with rapid- higher for eletriptan (35%) than for sumatriptan (25%) release/fast-disintegrating tablets, sumatriptan 100 mg [60]. The sustained pain-free response was higher for (n ϭ 1101) had a TG for headache relief of 30% (95% eletriptan (22%) than for sumatriptan (15%) [60]. The CI 26–34%) versus placebo (n ϭ 1113), the same frequency of adverse effects (AE) was similar for the two as sumatriptan 100 mg in the meta-analysis [12, 38] drugs [60]. In the meta-analysis eletriptan 80mg, a dose (see Fig. 22.1A). For pain-free status after 2 hours, the not routinely used, was superior to sumatriptan 100mg mean TG was 32% (95% CI 28–36%) [64] which is for TG for headache relief, pain-free status after 2 hours higher than the 20% TG for conventional sumatriptan (see Fig. 22.1) and sustained pain-free status [12,38]. tablets (see Fig. 22.1B). Unfortunately, there are no This was also the case in the comparative studies [12]. Almotriptan 12.5mg had the same initial 2-hour relief 284

Headache Table 22.2 Clinical efficacy, scientific proof for efficacy, and potential for adverse events, rated on a scale from ϩ to ϩϩϩϩ, for some drugs used in acute migraine treatment Drug Clinical Scientific proof for Adverse event potential efficacya efficacy Subcutaneous sumatriptan 6 mg ϩϩϩϩ ϩϩϩϩ* ϩϩϩ Sumatriptan 100 mg ϩϩϩ ϩϩϩϩ* ϩϩ Sumatriptan 50 mg ϩϩϩ ϩϩϩϩ* ϩ Rizatriptan 10 mg ϩϩϩ ϩϩϩϩ* ϩϩ Zolmitriptan 2.5 mg ϩϩϩ ϩϩϩϩ ϩϩ Naratriptan 2.5 mg ϩϩ ϩϩϩϩ* Eletriptan 40 ϩϩϩ ϩϩϩϩ* 0b Frovatriptan 2.5 mg ϩϩ ϩϩϩ Almotriptan 12.5 mg ϩϩϩ ϩϩϩϩ* ϩϩ Oral ergotamine 2 mgc ϩ ϩ ϩ Rectal ergotamine 2 mgd ϩϩϩ ϩϩ Effervescent aspirin ϩϩϩ ϩϩϩ 0b Aspirin plus metoclopramidee ϩϩϩ ϩϩϩ Naproxenf ϩϩ ϩϩϩ ϩϩ Ibuprofenf ϩϩ ϩϩϩ ϩϩϩ ϩ ϩ ϩ ϩ * Systematic review available. a Based on a combination of the published literature and personal experience; b no more adverse events than placebo; c for details, see Tfelt-Hansen et al. 2000 [37]; d see www.GSK.com; esee Tfelt-Hansen et al. 1995 [71]; f see Tfelt-Hansen & Rolan 2006 [103]. comparative RCT with this new oral formulation of attacks of long duration, should be administered by sumatriptan. the rectal route [69]. Treatment in the mild phase of a migraine attack The nonsteroidal anti-inflammatories aspirin, results in higher TG for 2-hour pain-free status. This tolfenamic acid, naproxen, ketoprofen, ibuprofen, and measure increased to 48% after rizatriptan 10 mg [65] diclofenac potassium were all superior to placebo in and 25–44% after sumatriptan 100 mg [50]. It has the treatment of migraine attacks [70]. therefore been suggested that migraine attacks should be treated early. Effervescent aspirin 1000mg was as effective as sumatriptan 50mg in one RCT [110]. Finally, in three RCT oral sumatriptan 100 mg, rizatriptan 10 mg, and eletriptan 40 mg and 80 mg The combination of lysine acetylsalicylate 1640 mg were superior to oral ergotamine plus caffeine [66– plus metoclopramide 10 mg was as effective as 68]. In contrast, in one RCT rectal ergotamine plus sumatriptan 100 mg in one placebo-controlled RCT caffeine was superior to rectal sumatriptan 25 mg [71]. Diclofenac potassium 100 mg was comparable (Clinical Trial Register, GSK: www.gsk.com). to sumatriptan 100 mg [111]. Tolfenamic acid 200 mg was inferior to sumatriptan 100 mg [52]. An overview of acute treatment drugs for migraine is shown in Table 22.2. Systematic reviews of drugs for migraine prophylaxis Other drugs for migraine In one systematic Cochrane review of propranolol In a consensus paper on ergotamine it was stated that [45] it was concluded that propranolol is effective for the evidence for its use in migraine is there but is not short-term migraine prophylaxis, evidence on long- up to current standards [69]. The oral bio-availability term effects is lacking, and propranolol seems to be as of ergotamine is less than 1% and if used in migraine effective and safe as a variety of other drugs used for 285

Chapter 22 Table 22.3 Clinical efficacy, scientific proof for efficacy, and potential for adverse events, rated on a scale from ϩ to ϩϩϩϩ, for some drugs used in migraine prophylaxis. Modified from Tfelt-Hansen [104] Drug Clinical efficacya Scientific proof for Adverse event efficacy potential β-Blockers (propranolol*, metoprolol, ϩϩϩϩ ϩϩϩϩ ϩϩ atenolol, nadolol, bisoprolol) ϩϩ or ϩϩϩ ϩϩϩ ϩϩϩ Antiepileptics ϩϩϩ ϩϩϩϩ ϩϩϩ Sodium valproate /divalproex* Topiramate* ϩϩϩϩ ϩϩ ϩϩϩϩ ϩϩ ϩϩ ϩϩϩ Antiserotonin drugs Methysergideb ϩϩϩ ϩϩϩϩ ϩϩϩ Pizotifenb ϩ ϩ ϩ Calcium antagonists ϩϩ ϩϩ ϩϩ Flunarizinec ϩϩ ϩϩ ϩϩ Verapamil ϩϩ ϩϩ ϩϩϩ NSAIDs ϩϩ ϩϩ ϩϩ Naproxen ϩϩ ϩϩ ϩ Tolfenamic acid ϩ ϩ ϩ ϩϩ ϩ ϩϩ Miscellaneous Amitriptyline Lisinopril Candesartan Clonidine Dihydroergotamine A systematic review is available. a Based on a combination of the published literature and personal experience; b for details see Tfelt-Hansen & Saxena 2006 [105–107]; c for details, see Toda & Tfelt-Hansen 2006 [108]. migraine prophylaxis. Propranolol is thus a first-line propranolol 160 mg [74]. Topiramate is a second-line drug for migraine prophylaxis. drug for migraine prophylaxis. There are, however, adverse events, sedation and cognitive impairment In another systematic Cochrane review the efficacy with topiramate and its place in migraine prophylaxis of feverfew in migraine prophylaxis was judged as depends on how migraine patients in clinical practice unproven [46]. will tolerate the drug. Anticonvulsants were evaluated for migraine proph- For a review of preventive therapy in migraine, see ylaxis in one systematic Cochrane review [47]. It was Yoon et al. 2005 [75]. An overview of the prophylactic concluded that sodium valproate/divalproex sodium is drugs for migraine is shown in Table 22.3. superior to placebo. Valproate/divalproex sodium are second-line drugs for migraine prophylaxis because of Clinical trials in tension-type adverse events, sedation and weight gain. There is no headache RCT comparing valproate with propranolol. Acute treatment of tension-type headache has been Topiramate has been evaluated in three large RCT investigated in episodic tension-type headache, and in and topiramate 100 mg was superior to placebo in all [72–74]. Topiramate 100 mg was comparable to 286

Headache prophylactic treatment of chronic tension-type head- air [91]. In one RCT subcutaneous sumatriptan 6 mg ache, antidepressant drugs and botulinum toxin have was superior to placebo [92]. In another, intranasal been investigated. sumatriptan 20 mg was superior to placebo [93] and intranasal zolmitriptan 5 mg and 10 mg was superior to Aspirin and paracetamol are the analgesics used placebo in one RCT [94]. Subcutaneous sumatriptan most commonly in the treatment of acute tension- 6 mg is the first-line drug for cluster headache attacks type headache [76]. In the most recent RCT, 452 but it is very expensive. patients treated episodes of tension-type headache with aspirin (500 mg or 1000 mg), paracetamol In one prophylactic RCT, verapamil was compara- (500 mg or 1000 mg) or placebo. Headache relief after ble to lithium [95] and in another, verapamil 360 mg 2 hours was 76% after aspirin 1000 mg and 71% after was superior to placebo [96]. paracetamol 1000 mg. Both were superior to placebo despite a high placebo response of 55%. The follow- Research needed to improve ing NSAID were superior to placebo in RCT on the the evidence-based management treatment of acute tension-type headache: ibupro- of migraine fen [77], ketoprofen [78, 112] naproxen [79, 80] and diclofenac [81]. Caffeine has long been used as an Clinical experience shows that triptans are more effec- analgesic adjuvant [76]. In a RCT the combination of tive in migraine than over-the-counter drugs such as aspirin and caffeine was superior to paracetamol and aspirin [97]. In RCT, however, triptans and aspirin are placebo [77]; in another RCT, the combination of comparable [98] and this was also the case in a recent caffeine and ibuprofen was superior to ibuprofen and systematic review including 991migraine attacks [99]. placebo [113]. The combination of caffeine, aspirin When headache was severe the results were similar and paracetamol was more effective than the single [99]. It has been suggested [97] that patients treated substances in one large RCT. with triptans in clinical practice may be relatively more responsive to triptans and relatively less respon- Antidepressants are the drugs most commonly used sive to other agents than those who participate in in chronic tension-type headache [82]. Amitriptyline clinical trials. In patients recruited from general prac- was superior to placebo in most RCT [82–86] but in tice, however, the pain-free response [100] was simi- the largest RCT there was no effect of amitriptyline lar to other studies [38]. Thus a selection bias is most and amitriptylinoxide [87]. Mirtazapine 15–30 mg likely not the reason for the discrepancy in results daily was superior to placebo in one RCT [88] in clinical practice and controlled trials. The lack of whereas the the selective serotonin reuptake inhibi- superiority of triptans in controlled clinical trials has tor (SSRI) citalopram was not more effective than meant that the WHO has not included triptans on the placebo [86]. list of essential drugs. After positive open studies, botulinum toxin has Stricter success criteria [51] have been suggested been studied in chronic tension-type headache but in order to demonstrate superiority of triptans. It has the conclusion of these RCT is that botulinum toxin been recommended to use the criterion of painfree sta- is not more effective than placebo [89].Thus in the tus after 2 hours [53], preferably with administration most recent large RCT including 298 randomized in the mild phase [65]. This resulted in 70% being pain patients, botulinum toxin A admistered in doses from free after rizatriptan 10 mg [65]. In addition, sustained 50 U to 150 U was not different from placebo [90]. pain-free status, which occurs in 60% of patients after rizatriptan 10 mg [65], could be used. However, in the Clinical trials in cluster systematic review of aspirin versus sumatriptan, these headache two efficacy measures were quite comparable (pain- free status 27–29% and sustained pain-free status In contrast to migraine and tension-type headache, 22–24%) [99]. Based on these results, stepwise care, there are very few RCT in cluster headache. This is starting with aspirin, has been recommended [99]. most likely due to the fact that cluster headache is a rather infrequent disease (see above). There is one What efficacy can one achieve with a triptan? RCT demonstrating the superiority of O2 versus plain The highest effect was 87% pain-free status after 287

Chapter 22 subcutaneous naratriptan 10 mg [51, 101]. So inher- 10. Friedman AP. Characteristics of tension headache: a pro- ently the triptans as a group are most likely very file of 1420 cases. Psychosomatics 1979; 20: 418–422. effective drugs in migraine when the right drug and the optimum form of administration are used. 11. Rasmussen BK, Jensen R, Olesen J. A population-based Similarly, subcutaneous sumatriptan (76%) was analysis of the diagnostic criteria of the International superior to intravenous acetylsalicylic acid lysinate Headache Society. Cephalalgia 1991; 11: 130–134. (44%) for pain-free status after 2 hours [102]. 12. Ferrari MD, Goadsby PJ, Roon KI, Lipton RB. Triptans Why the presumed superiority of triptans is not (serotonin, 5-HT1B/1D agonists) in migraine: detailed apparent in current RCT remains an enigma. In my results and methods of a meta-analysis of 53 trials. opinion, one should probably try the three-way cross- Cephalalgia 2002; 22: 633–658. over design with a triptan, a NSAID and a placebo. In this design patients can be classified as responding bet- 13. Russell M. Genetic epidemiology of migraine and cluster- ter to one of the active drugs or with equal response. headache. Cephalalgia 1997; 17: 683–701. Patients’ characteristics in the two groups with active drugs can then be analyzed as possible predictors of 14. Aurora SK, Welch KMA. Brain hyperexcitability in response with each drug. If such preferential predic- migraine: evidence from transcranial magnetic stimula- tors are identified, they can then be used in clinical tion studies. Curr Opin Neurol 1998; 11: 205–209. practice. In addition, on can ask for preference, a clini- cally most relevant parameter, with this design. 15. Fabricius M, Fuhr S, Bhatia R, et al. Cortical spread- ing depression and peri-infarct depolarization in acutely References injured human cerebral cortex. Brain 2006; 129: 778–790. 1. Olesen J, Larsen B, Lauritzen M. Focal hyperemia followed 16. Hadjikhani N, Sanchez del Rio M, Wu O, et al. by spreading oligemia and impaired activation of rCBF in Mechanisms of migraine aura revealed by functional MRI classic migraine. Ann Neurol 1981; 9: 344–352. in human visual cortex. Proc Natl Acad Sci USA 2001; 98: 4687–4692. 2. Humphrey PP, Feniuk W. Mode of action of the anti- migraine drug sumatriptan. Trends Pharmacol Sci 1991; 17. Iversen HK, Nielsen TH, Olesen J, Tfelt-Hansen P. Arterial 12: 444–446. responses during migraine headache. Lancet 1990; 336: 837–839. 3. Ophoff RA, Terwindt GM, Vergouwe MN, et al. Familial hemiplegic migraine and epidodic ataxia type-2 are caused 18. Friberg L, Olesen J, Iversen HK, Sperling B. Migraine pain by mutations in the Ca2+ channel gene CACNL1A4. Cell associated with middle cerebral artery dilatation: reversal 1996; 87: 543–552. by sumatriptan. Lancet 1991; 338: 13–17. 4. Rasmussen BK, Jensen R, Schroll M, Olesen J. Epidemio- 19. Zwetsloot CP, Caekebeke JF, Ferrari MD. Lack of asymme- logy of headache in a general population – a prevalence try of middle cerebral artery blood velocity in unilateral study. J Clin Epidemiol 1991; 44: 1147–1157. migraine. Stroke 1993; 24: 1335–1338. 5. Göbel H, Petersen-Braun M, Soyka D. The epidemiology of 20. Iversen HK, Thomsen LL, Olesen J. Headache induced headache in Germany: a nationwide survey of a representa- by a nitric oxide donor (nitroglycerin) responds to tive sample on the basis of the headache classification of sumatriptan. A human model for development of the International Headache Society. Cephalalgia 1994; 14: migraine drugs. Cephalalgia 1996; 16: 412–418. 97–106. 21. Thomsen LL, Kruuse C, Iversen HK, Olesen J. A nitric 6. Rasmussen BK, Stewart WF. Epidemiology of migraine. In: oxide donor (nitroglycerin) triggers genuine migraine Olesen J, Tfelt-Hansen P, Welch KMA (eds) The Headaches, attacks. Eur J Neurol 1994; 1: 73–80. 2nd edn. Lippincott, Williams and Wilkins, Philadelphia, 2000: 227–233. 22. Olesen J, Iversen HK, Thomsen LL. Nitric oxide hyper- sensitivity. A possible molecular mechanisms of migraine 7. Schwartz BS, Stewart WF, Simon D, Lipton RB. Epidemiology pain. Neuroreport 1993; 4: 1027–1030. of tension-type headache. JAMA 1998; 279: 381–383. 23. Ahn AH, Goadsby PJ. Animal models of headache. In: 8. Goadsby P, Tfelt-Hansen P. Cluster headache: introduction Olesen J, Goadsby PJ, Ramadan NM, Tfelt-Hansen P, and epidemiology. In: Olesen J, Goadsby PJ, Ramadan NM, Welch KMA (eds) The Headaches, 3rd edn. Lippincott, Tfelt-Hansen P, Welch KMA (eds) The Headaches, 3rd edn. Williams and Wilkins, Philadelphia, 2006: 221–230. Lippincott, Williams and Wilkins, Philadelphia, 2006: 743–745. 24. Tvedskov JF, Lipka K, Ashina M, Iversen HK, Schifter S, 9. Olesen J, Goadsby PJ, Ramadan NM, Tfelt-Hansen P, Olesen J. No increase of calcitonin gene-related peptide Welch KMA (eds) The Headaches, 3rd edn. Lippincott, in jugular blood during migraine. Ann Neurol 2005; 58: Williams and Wilkins, Philadelphia, 2006. 561–568. 25. Lassen LH, Haderslev PA, Jacobsen VB, Iversen HK, Sperling B, Olesen J. CGRP may play a causative role in migraine. Cephalalgia 2002; 22: 54–61. 26. Olesen J, Diener HC, Husstedt IW, et al. Calcitonin gene- related peptide receptor antagonist BIBN 4096 BS for the acute treatment of migraine. N Engl J Med 2004; 350: 1104–1110. 27. Ho T, Mannix L, Rapoport A, et al. Efficacy and toler- ability of a novel, oral CGRP antagonist, MK-0974, in the acute treatment of migraine. Cephalalgia 2007; 27: 759. 288

Headache 28. Ulrich V, Gervil M, Olesen J. The relative influence of Issue 2. Art. No.: CD003225. DOI: 10.1002/14651858. environment and genes in episodic tension-type headache. CD003225.pub2. Neurology 2004; 62: 2065–2069. 46. Pittler MH, Ernst E. Feverfew for preventing migraine. Cochrane Database of Systematic Reviews 2004, Issue 1. Art. 29. Rasmussen BK. Migraine and tension-type headache in No.: CD002286. DOI: 10.1002/14651858.CD002286.pub2. a general population: precipitating factors, female hor- 47. Chronicle EP, Mulleners WM. Anticonvulsant drugs for mones, sleep pattern and relation to lifestyle. Pain 1993; migraine prophylaxis. Cochrane Database of Systematic 53: 65–72. Reviews 2004, Issue 3. Art. No.: CD003226. DOI: 10.1002/14651858.CD003226.pub2. 30. Rasmussen BK. Migraine and TTH in a general popu- 48. Pilgrim AJ. Methodology of clinical trials of sumatriptan lation: psychosocial factors. Int J Epidemiol 1992; 21: in migraine and cluster headache. Eur Neurol 1991; 31: 1138–1143. 295–299. 49. Oldman A, Smith LA, McQuay HJ, Moore RA. Rizatriptan 31. Mitsikostas DD, Thomas AM. Comorbidity of headache for acute migraine. Cochrane Database of Systematic and depressive disorders. Cephalalgia 1999; 19: 211–217. Reviews 2007, Issue 1. Art. No.: CD003221. DOI: 10.1002/14651858.CD003221.pub2. 32. Jensen R, Rasmussen BK, Olesen J. Cephalic muscle ten- 50. Winner P, Landy S, Richardson M, Ames M. Early inter- derness and pressure pain threshold in headache. A popu- vention in migraine with sumatriptan tables 50 mg versus lation study. Pain 1993; 52:193–199. 100 mg: a pooled analysis of data from six clinical trials. Clin Ther 2005; 27: 1785–1794. 33. Bendtsen L, Jensen R, Olesen J. Decreased pain detection 51. Tfelt-Hansen P. Maximum effect of triptans in migraine? and tolerance thresholds in chronic tension-type head- A comment. Cephalalgia 2008; 28(7): 767–768. ache. Arch Neurol 1996; 53: 373–376. 52. Tfelt-Hansen P. A review of evidence-based medicine and meta-analytic reviews in migraine. Cephalalgia 2006; 26: 34. Jensen R, Bendtsen L, Olesen J. Muscular factors are of 1265–1274. importance in tension-type headache. Headache 1998; 53. International Headache Society Clinical Trial 38:10–17. Subcommittee. Guidelines for controlled trials of drugs in migraine. Cephalalgia 2000; 20: 765–786. 35. Jensen R. Pathophysiological mechanisms of tension-type 54. Tfelt-Hansen P, Teall J, Rodriguez F, et al, Rizatriptan 030 headache. A review of epidemiological and experimental Study Group. Oral rizatriptan versus oral sumatriptan: studies. Cephalalgia 1999; 19: 602–621. a direct comparative study in the acute treatment of migraine. Headache 1998; 38: 748–755. 36. Goadsby PJ. Pathophysiology of cluster headache: a 55. Goldstein J, Ryan R, Jiang K, et al, Rizatriptan Protocol trigeminal autonomic cephalalgia. Lancet Neurol 2002; 1: 046 Study Group. Crossover comparison of rizatriptan 251–257. 5 mg and 10 mg versus sumatriptan 25 mg and 50 mg in migraine. Headache 1998; 38: 737–747. 37. Tfelt-Hansen P, de Vries P, Saxena PR. Triptans in 56. Kolodny A, Polis A, Battisti W, Johnson-Pratt L, migraine. A comparative review of pharmacology, phar- Skobieranda F, Rizatriptan Protocol 052 Study Group. macokinetics and efficacy. Drugs 2000; 60: 1259–1287. Comparison of rizatriptan 5 mg and 10 mg tablets and sumatriptan 25 mg and 50 mg tablets. Cephalalgia 2004; 38. Ferrari MD, Roon KI, Lipton RB, Goadsby PJ. Oral 24: 540–546. triptans (serotonin 5-HT1B/!D agonists) in acute migraine: 57. Mathew NT, Schoenen L, Winner P, Muirhead N, Sikes a meta-analysis of 53 trials. Lancet 2001; 358: 1668–1675. CR. Comparative efficacy of eletriptan 40 mg versus sumatriptan 100 mg. Headache 2003; 43: 214–222. 39. Gawel MJ, Worthington I, Maggisano A. A systematic 58. Goadsby PJ, Ferrari MD, Olesen J, et al, Eletriptan Steering review of the use of triptans in acute migraine. Can J Committee. Eletriptan in acute migraine: a double-blind, Neurol Sci 2001; 28: 30–41. placebo-controlled comparison to sumatriptan. Neurology 2000; 54: 156–163. 40. Oldman AD, Smith LA, McQuay HJ, Moore RA. 59. Sandrini G, Färkilä M, Burgess G, Forster E, Haughie S. Pharmacological treatments for acute migraine: quantita- Eletriptan vs. sumatriptan: a double-blind, placebo-con- tive systematic review. Pain 2002; 97: 247–257. trolled, multiple migraine attack study. Neurology 2002; 59: 1210–1217. 41. Smith LA, Oldman A, McQuay HJ, Moore RA. Eletriptan 60. Diener HC, Ryan R, Sun W, Hettiarachchi J. The 40-mg for acute migraine. Cochrane Database of Systematic dose of eletriptan: comparative efficacy and tolerability ver- Reviews 2007, Issue 1. Art. No.: CD003224. DOI: sus sumatriptan 100 mg. Eur J Neurol 2004; 11: 125–134. 10.1002/14651858.CD003224.pub2. 61. Spierings ELH, Gomez-Mancilla B, Grosz D, Rowland CR, Whaley FS, Jirgens KJ. Oral almotriptan vs oral 42. McCrory DC, Gray RN. Oral sumatriptan for acute migraine. Cochrane Database of Systematic Reviews 2003, Issue 3. Art. No.: CD002915. DOI: 10.1002/14651858. CD002915. 43. Ashcroft DM, Millson D. Naratriptan for the treatment of acute migraine: meta-analysis of randomised controlled trials. Pharmacoepidemiol Drug Saf 2004; 13: 73–82. 44. Dahlöf CG, Pascual J, Dodick DW, Dowson AJ. Efficacy , speed of action and tolerability of almotriptan in the acute treatment of migraine: pooled individual patient data from four randomized, double-blind, placebo-controlled clinical trials. Cephalalgia 2006; 26: 400–408. 45. Linde K, Rossnagel K. Propranolol for migraine prophy- laxis. Cochrane Database of Systematic Reviews 2004, 289

Chapter 22 sumatriptan in the abortive treatment of migraine. 75. Yoon MS, Savidor I, Diener HC, Limmroth V. Evidence- A double-blind, randomized, parallel-group, optimum- based medicine in migraine prophylaxis. Exp Rev Neurother dose comparison. Arch Neurol 2001; 58: 944–950. 2005; 5: 333–341. 62. Cabarrocas X. Reply to Tfelt-Hansen. Cephalalgia 2004; 24: 688–689. 76. Mathew NT, Ashina M. Acute pharmacotherapy of ten- 63. Dowson AJ, Massiou H, Lainez JM, Cabarrocas X. sion-type headaches. In: Olesen J, Goadsby PJ, Ramadan Almotriptan is an effective and well-tolerated treatment NM, Tfelt-Hansen P, Welch KMA (eds) The Headaches, for migraine pain: results of a randomized, double-blind, 3rd edn. Lippincott, Williams and Wilkins, Philadelphia, placebo-controlled clnical trial. Cephalalgia 2002; 22: 2006: 727–733. 453–461. 64. Sheftell FD, Dahlof CG, Brandes JL, Agosti R, Jones MW, 77. Schachtel BP, Furey SA, Thoden WR. Nonprescription Barrett PS. Two replicate randomized, double-blind, pla- ibuprofen and acetaminophen in the treatment of tesion- cebo-controlled trials of the time to onset of pain relief in type headache. J Clin Pharmacol 1996; 36: 1120–1125. the acute treatment of migraine with a fast-disintegrating/ rapid-release formulation of sumatriptan tablets. Clinic 78. van Vergen JMA, Schoemaker RC, Jacobs LD, et al. Self- Therapeut 2005; 27: 407–417. medication of single headache episode with ketoprofen, 65. Mathew NT, Kailasam J, Meadors L. Early treatment of ibuprofen, or placebo home-monitored with an elec- migraine with rizatriptan: a placenbo-controlled study. tronic headache diary. Br J Clin Pharmacol 1996; 42: Headache 2004; 44: 669–673. 475–481. 66. Multinational Oral Sumatriptan and Cafergot Comparative Study Group. A randomized, double-blind comparison 79. Miller DS, Talbot CA, Simpson W, Korev A. A compari- of sumatriptan in the acute treatment of migraine. Eur son of naproxen sodium, acetaminophen and placebo in Neurol 1991; 31: 314–322. the treatment of muscle contraction headache. Headache 67. Christie S, Gobel H, Mateos V, Allen C, Vrijens F, 1987; 27: 392–396. Shivaprakash M. Crossover comparison of efficacy and preference for rizatriptan 10 mg versus ergotamine/caf- 80. Sevelius H, Segre M, Bursick R. Comparative analge- feine in migraine. Eur Neurol 2003; 49: 20–29. sic effects of naproxen sodium, aspirin, placebo. J Clin 68. Diener HC, Jansen JP, Reches A, Pascual J, Pieti D, Steiner Pharmacol 1980; 20: 322–329. TJ, Eletriptan and Cafergot Comparative Study Group. Efficacy, tolerability and safety of oral eletriptan and 81. Kubizek F, Ziegler G, Gold MS, et al. Low-dose diclofenac ergotamine plus caffeine (Cafergot) in the acute treatment potassium in the treatment of episodic tension-type head- of migraine: a multicentre, randomized, double-blind, ache. Eur J Pain 2003; 7: 155–162. placebo-controlled comparison. Eur Neurol 2002; 47: 99–107. 82. Bendtsen L, Mathew NT. Prophylactic pharmacotherapy of 69. Tfelt-Hansen P, Saxena PR, Dahlof C, et al. Ergotamine in tension-type headache, In: Olesen J, Goadsby PJ, Ramadan the acute treatment of migraine – European consensus. NM, Tfelt-Hansen P, Welch KMA (eds) The Headaches, Brain 2000; 123: 9–18. 3rd edn. Lippincott, Williams and Wilkins, Philadelphia, 70. Tfelt-Hansen P, Rolan P. Nonsteroidal antiinflammatory 2006: 735–741. drugs in the acute treatment of migraine. In: Olesen J, Goadsby PJ, Ramadan NM, Tfelt-Hansen P, Welch KMA 83. Lance JW, Curran DA. Treatment of chronic tension head- (eds) The Headaches, 3rd edn. Lippincott, Williams and ache. Lancet 1964; 1: 1236–1239. Wilkins, Philadelphia, 2006: 449–457. 71. Tfelt-Hansen P, Henry P, Mulder K, Scheldewaert R G, 84. Diamond S, Baltes BJ. Chronic tension headache-treated Schoenen J, Chazot G. The effectiveness of combined oral with amitriptyline – a double-blind study. Headache 1971; lysine acetylsalicylate and metoclopramide compared 11: 110–116 with oral sumatriptan for migraine. Lancet 1995; 346: 923–926. 85. Göbel H, Hamouz V, Hansen C, et al. Chronic tension- 72. Brandes JL, Saper JR, Diamond M, et al. MIGR-002 Study type headache: amitriptyline reduces clinical headache- Group. Topiramate for migraine prevention: a randomized duration and experimental pain sensitiviity but does not controlled trial. JAMA 2004; 291: 965–973. alter pericranial muscle activity readings. Pain 1994; 59: 73. Silberstein SD, Neto W, Schmitt J, Jacobs D, MIGR-001 241–249. Study Group. Topiramate in migraine prevention: results of a large controlled trial. Arch Neurol 2004; 61: 490–495. 86. Bendtsen R, Jensen R, Olesen J. A non-selective (amitriptyl- 74. Diener HC, Tfelt-Hansen P, Dahlöf C, et al. Topiramate ine), but not a selective (citalopram), serotonin reuptake in migraine prophylaxis. Results from a placebo-control- inhibitor is effective in the prophylactic treatment of led trial with propranolol as an active control. Eur Neurol chronic tension-type headache. J Neurol Neurosurg 2004; 251: 953–950. Psychiatr 1996; 61: 285–290. 87. Pfaffenrath V, Diener HC, Isler H, et al. Efficacy and tol- erability of amitriptylinoxide in the treatment of chronic tension-type headache: a multi-centre controlled study. Cephalalgia 1994; 14: 149–155. 88. Bendtsen L, Jensen R. Mirtazapine is effective in the pro- phylactic treatment of chronic tension-type headache. Neurology 2004; 62: 1706–1711. 89. Evers S, Olesen J. Botulinum toxin in headache treatment: the end of the road? Cephalalgia 2006; 26: 789–771. 90. Silberstein SD, Gobel H, Jensen R, et al. Botulinum toxin type A in the prophylactic treatment of chronic 290

Headache tension-type headache: a multicentre, randomized, pla- 103. Tfelt-Hansen P, Rolan P. β-Adrenoceptor blocking drugs in cebo-controlled, parallel-group study. Cephalalgia 2006; migraine prophylaxis. In: Olesen J, Goadsby PJ, Ramadan 26: 790–800. NM, Tfelt-Hansen P, Welch KMA (eds) The Headaches, 91. Fogan L. Treatment of cluster headache. A double-blind 3rd edn. Lippincott, Williams and Wilkins, Philadelphia, comparison of oxygen v air inhalation. Arch Neurol 1985; 2006: 519–528. 42: 362–363. 92. Ekbom K, Sumatriptan Cluster Headache Study Group. 104. Tfelt-Hansen P. Prioritizing prohylactic treatment of Treatment for acute cluster headache with sumatriptan. migraines. In: Olesen J, Goadsby PJ, Ramadan NM, Tfelt- N Engl J Med 1991; 325: 322–326. Hansen P, Welch KMA (eds) The Headaches, 3rd edn. 93. van Vielt JA, Bahra A, Martin V, et al. Intranasal Lippincott, Williams and Wilkins, Philadelphia, 2006: sumatriptan in cluster headache: randomized, double-blind 567–568. placebo-controlled study. Neurology 2003; 60: 630–633. 94. Cittadini E, May A, Straube A, Evers S, Bussone G, Goadsby 105. Tfelt-Hansen P, Saxena PR. Antiserotonin drugs in PJ. Effectiveness of intranasal zolmitriptan in acute cluster migraine prophylaxis. In: Olesen J, Goadsby PJ, Ramadan headache: a randomized, placebo-controlled, double-blind NM, Tfelt-Hansen P, Welch KMA (eds) The Headaches, crossover study. Arch Neurol 2006; 63: 1537–1542. 3rd edn. Lippincott, Williams and Wilkins, Philadelphia, 95. Bussone G, Leone M, Percarisi, et al. Double-blind com- 2006: 529–537. parison of lithium and verapamil in cluster headache. Headache 1990; 30: 411–317. 106. Tfelt-Hansen P, Saxena PR. Antiserotonin drugs in migraine 96. Leone MJ, d’Amico D, Frediani F, et al. Verapamil in the prophylaxis. In: Olesen J, Goadsby PJ, Ramadan NM, prophylaxis of episodic cluster headache: a double-blind Tfelt-Hansen P, Welch KMA (eds) The Headaches, 3rd study. Neurology 2000; 64: 1382–1385. edn. Lippincott, Williams and Wilkins, Philadelphia, 2006: 97. Lipton RB, Bigal ME, Goadsby PJ. Double-blind clinical 529–537. trials of oral triptans vs other classes of acute migraine medication – a review. Cephalalgia 2004; 24: 321–332. 107. Tfelt-Hansen P, Saxena PR. Ergot alkaloids in the acute 98. Saxena PR, Tfelt-Hansen P. Triptans, 5HT1B/1D agonists treatment of migraine. In: Olesen J, Goadsby PJ, Ramadan in the acute treatment of migraine. In: Olesen J, Goadsby NM, Tfelt-Hansen P, Welch KMA (eds) The Headaches, PJ, Ramadan NM, Tfelt-Hansen P, Welch KMA (eds) The 3rd edn. Lippincott, Williams and Wilkins, Philadelphia, Headaches, 3rd edn. Lippincott, Williams and Wilkins, 2006: 459–467. Philadelphia, 2006: 469–503. 99. Lampl IC, Voelker M, Diener HC. Efficacy and safety of 108. Toda N, Tfelt-Hansen P. Calcium antagonists in migraine 1,000 mg effervescent aspirin: individual patient data prophylaxis. In: Olesen J, Goadsby PJ, Ramadan NM, meta-analysis of three trials in migraine headache and Tfelt-Hansen P, Welch KMA (eds) The Headaches, 3rd migraine accompanying symptoms. J Neurol 2007; 254: edn. Lippincott, Williams and Wilkins, Philadelphia, 2006: 705–712. 539–544. 100. Tfelt-Hansen P, Bach F, Daugaard D, Tsiropoulos I, Riddersholm B. Treatment with sumatriptan 50 mg in the 109. Goadsby PJ, Edvinsson L. Human in vivo evidence mild phase of migraine attacks in patients with infrequent for trigeminovascular activation in cluster headache. attacks. A randomised, double-blind, placebo-controlled Neuropeptide changes and effects of acute attacks thera- study. J Headache Pain 2006; 7: 389–394. pies. Brain 1994; 117(3); 427–434. 101. Dahlöf C, Hogenhuis L, Olesen J, Petit H, Ribbat J, Schoenen J et al. Early clinical experience with subcutaneous 110. Acute treatment of migraine attacks: efficacy and safety naratriptan in the acute treatment of migraine: a dose- of a nonsteroidal anti-inflammatory drug, diclofenac- ranging study. Eur J Neurol 1998; 5: 469–477. potassium, in comparison to oral sumatriptan and placebo. 102. Diener HC. Efficacy and safety of intravenous acetylsali- The Diclofenac-K/Sumatriptan Migraine Study Group. cylic acid lysinate compared to subcutaneous sumatriptan Cephalalgia 1999; 19(4): 232–240. and parenteral placebo in the acute treatment of migraine. A double-blind, double-dummy, randomized, multicenter, 111. Diener HC, Eikermann A, et al. Efficacy of 1,000 mg effer- parallel group study. The ASASUMAMIG Study Group. vescent acetylsalicylic acid and sumatriptan in treating Cephalalgia 1999; 19: 581–588. associated migraine symptoms. Eur Neurol 2004; 52(1): 50–56. Epub 2004 Jul 5. 112. Dahlöf CG, Jacobs LD. Ketoprofen, paracetamol and pla- cebo in the treatment of episodic tension-type headache. Cephalalgia 1996; 16(2): 117–123. 113. Diamond S, Balm TK, Freitag FG. Ibuprofen plus caffeine in the treatment of tension-type headache. Clin Pharmacol Ther 2000 Sep; 68(3): 312–319. 291

C HAP TE R 23 Chest pain syndromes Austin Leach and Michael Chester National Refractory Angina Centre, Royal Liverpool and Broadgreen University Hospital, Liverpool, UK Introduction Persistent (continuous) or episodic (paroxysmal) chest pain is usually a result of pathology of the chest Chest pain is a common presenting symptom. In the wall. Brief unpredictable episodes of sharp precor- UK, two of the most prevalent causes of death are dial stabbing pains, usually over the anatomic apex of ischemic heart disease and lung cancer. Patients pre- the heart, are common and are rarely associated with senting with chest pain often fear a potentially fatal underlying pathology. A general sense of chest tightness diagnosis; it follows that the pain clinician’s skills may in the absence of any pathologic process commonly be tested to their limits when attempting to help a accompanies severe anxiety states. However, a diagnosis patient with chest symptoms. of anxiety-related chest tightness should be made with great caution and empathy because anxiety is also a We do not intend to discuss in detail the diagnosis common feature of chest pain of pathologic origin. and management of acute chest pain syndromes–these may represent emergency life-threatening situations Anatomy whose management rightly focuses on interrupting the natural history of the underlying disease rather The chest, more correctly referred to as the thorax, is than symptom control. However, many patients and an irregularly shaped cylinder with a narrow opening healthcare professionals may think and behave as if superiorly and a larger opening inferiorly. The supe- the patient is developing an immediately life-threat- rior thoracic opening is in continuity with the neck ening condition such as acute myocardial infarction and the inferior thoracic opening is separated from when, in reality, their symptoms are a manifestation the abdomen by the diaphragm. The musculoskeletal of a chronic pain syndrome. It is important for the wall of the thorax is flexible and consists of segmen- patient, and their advisors, to have some basic rules tally arranged vertebrae, ribs, muscles and the ster- that help to differentiate between fluctuations in sta- num. The thoracic cavity enclosed by the thoracic wall ble angina and the development of thrombus-related is subdivided into three major components: the left unstable coronary syndromes (see below). and right pleural cavities, each surrounding a lung; the mediastinum, a median, longitudinal soft tissue Managing patients with severe angina can provoke partition containing the heart, the oesophagus, the anxiety for the clinician. It is important to note that trachea, the major systemic blood vessels and a variety overall annual all-cause mortality is very low in this of major nerves [2]. population [1]. True unstable angina or myocardial infarction is uncommon in refractory angina and in The thorax has three main functions. our experience the majority of patients can distinguish • Breathing: contraction of the diaphragm and between a fluctuation in stable angina and an infarction. changes in the lateral and anterior dimensions of Evidence-Based Chronic Pain Management. Edited by the thoracic wall caused by movement of the ribs C. Stannard, E. Kalso and J. Ballantyne. © 2010 Blackwell alter the volume of the thoracic cavity – these are Publishing. key elements in breathing. 292

Chest pain syndromes • Protection of vital organs. In our experience, well-informed patients tend to defer • The mediastinum acts as a conduit for structures invasive procedures until they have tried low-risk, noninvasive options. connecting thoracic organs to other body regions, and for structures passing completely through the This modern approach to angina management has thorax from one body region to another. profound implications for the incidence and preva- lence of chronic refractory angina. Using the European Angina pectoris Society of Cardiology’s disease-centered definition of refractory angina, we conservatively estimate there The term “angina pectoris” first appeared in print in are at least 50,000ϩ sufferers in the NHS [4]. The 1768, in a paper from William Heberden’s Medical prevalence of refractory angina would be considerably Transactions published by the Royal College of greater if the North American standard of consent Physicians. The term comes from the Greek word were applied since it is reasonable to assume that a ankhon (“strangling”) and the Latin pectus (“chest”). prudent patient would wish to know about all available Commonly omitted from Heberden’s original descrip- evidence-based alternatives (currently only available to tion of the symptoms is the accompanying sensation refractory angina patients) before consenting to a of angor animi (from the Latin for “mental anguish of potentially lethal palliative revascularization proce- the animating spirit”) which describes the perception dure. The proportion of patients who would try rela- of imminent death. tively noninvasive therapies in preference to invasive therapies can be judged by the very high proportion Prevalence of patients who “choose” the less effective multivessel Chronic stable angina is an increasingly common con- PCI option over coronary artery bypass graft (CABG). dition affecting approximately 1.3 million UK citizens [3]. Its prevalence closely correlates with the incidence Incidence of coronary artery disease (CAD) which is the most The estimates of incidence of angina vary in different common underlying pathologic cause for angina. studies. The Scottish Continuous Morbidity Study Although neither condition is “curable,” survival with estimates that 52,000 men and 43,000 women develop CAD has consistently improved over the past two angina across the UK each year [5]. decades and consequently the prevalence of angina is increasing. Many angina sufferers are successfully Risk factors managed using a combination of standard medical The risk factors for angina are the same as those for treatment plus angioplasty and stent (percutaneous CAD and can be divided into reversible and irrevers- coronary intervention (PCI)) or bypass. ible. The main irreversible risk factors are aging, male gender and a positive family history. There are many The diagnosis of “chronic refractory angina” applies reversible risk factors: smoking; raised cholesterol; to a subset of stable angina patients for whom PCI or inadequate physical activity; poor diabetic control; bypass is not suitable. This diagnosis is generally made obesity; hypertension; a diet lacking adequate fresh by interventional cardiologists and surgeons. The fruit and vegetables, oily fish; inappropriate intake of prevalence of this painful, disabling and costly prob- alcohol. lem is hard to quantify because there is a wide range of opinion on the question of when revascularization is Pathophysiology of angina not suitable. Many interventionally minded cardiolo- The fact that angina is a symptom rather than a dis- gists consider that revascularization is suitable if it is ease in its own right is commonly overlooked. With feasible, whereas there is an emerging body of opin- the rapid rise in the incidence of atheromatous CAD ion within cardiology that palliative revascularization in the latter part of the 20th century, there has been should be a last resort and only after the patient has an increasing tendency for many clinicians to mistak- considered the range of alternatives. This latter view is enly consider angina to be synonymous with coronary consistent with patient-centered care and necessarily disease. In fact, 100 years ago a patient presenting with requires the full participation of the informed patient. 293

Chapter 23 typical symptoms of angina would be far likelier to be referred pain. This hypothesis remains plausible today. suffering from aortic valve stenosis or coronary ostial We go further and propose that angina is only possi- stenosis (with normal coronary arteries) secondary to ble after such abnormal neurologic pathways develop. syphilitic aortitis. Painless or “silent” ischemia is common. Patients who have not experienced prior angina commonly suffer a Angina pectoris is a visceral pain. As for other pains painless myocardial infarction but subsequently go on of visceral origin, the mechanisms that generate the to develop exertional angina days, weeks or months sensation of angina are poorly understood. In the later. On detailed questioning, patients often describe 17th century, at the dawn of modern medical theory, an event when they experienced other typical features William Harvey, best known for his description of the of a myocardial infarct such as acute dyspnea, nausea circulation, was also considering the nature of sensa- and profuse sweating. This observation is consist- tion and feeling. He was one of the first scientists to ent with the hypothesis that the sensation of angina consider the blood as a body part, and speculated requires the development of previously nonexistent that the blood itself was capable of feeling: “even if it or quiescent neurologic connections. However, the were devoid of sensation this would not disqualify the neurophysiologic processes that cause new spinal blood from forming a part and even a very principal connections to activate between visceral and somatic part of a body endowed with sensibility. For neither fibers (i.e. converting “silent” ischemia into “noisy” does the brain nor the spinal marrow nor the crystal- (painful) ischemia) remain obscure. While elegant line or the vitreous humour of the eye, feel anything, animal studies suggest a role for the vagus in afferent though by common consent these are parts of the cardiac signaling, clinical evidence in humans sug- body. Nay even the heart itself, the most distinguished gests a far more prominent role for sympathetic tracts part of the body, appears to be insensible.” Here [5]. Temporary cardiac sympathectomy has been Harvey reports the case of the young son of Viscount shown to produce freedom from angina for periods Montgomery whose heart had been laid bare in conse- greatly exceeding duration of action of local anes- quence of a severe fall in early childhood and tolerated thetic agents. A case of temporary vagus nerve block palpation and other forms of stimulation of his heart carried out at the National Refractory Angina Centre, without discomfort. which produced the anticipated tachycardia, did not result in relief from angina [6]. There is little evidence for the existence of visceral nociceptors, and the current understanding is that Positron emission tomography studies in patients signaling in the “visceral pain pathway” is initiated by with angina pectoris show activation of many brain sympathetic afferent activity. It is likely that activa- areas that are also active when somatic pain fibers tion of these afferent autonomic nerves results from are stimulated: the hypothalamus, the periaqueduc- ischemia in the myocardium and/or the coronary tal gray matter, thalamus and prefrontal cortex. Silent arterial endothelium. Whether the source of sympa- myocardial ischemia seems to be associated with a thetic activation is ischemia or the cellular response failure of transmission of signals from the thalamus to ischemia is uncertain. Furthermore, the gross anat- to the frontal cortex [7]. Once cortical stimulation omy of the cardiac autonomic nerve plexuses shows has occurred there will be a conscious perception of that there is potential for a considerable volume of pain. The pain will seem to originate in the part of neurologic “cross-talk” between other visceral inputs the body represented by the area of sensory cortex to the thoracic sympathetic chain (such as esophagus, that has been activated, and the emotional response stomach and lungs) prior to forming connections generated will depend upon the significance of the within the spinal cord. sensation attached by the limbic system. It is likely, though unproven, that the sensation of angor animi Autonomic afferent fibers enter the spinal cord arises from activity in the amygdalo-hippocampal and synapse in laminae 1 and 2. Once again, there is apparatus. In part, this seems responsible for the the potential for cross-talk, not only among visceral (physiologically entirely inappropriate) increase in afferent fibers but also between visceral and somatic catecholamine activity that invariably accompanies afferent fibers. In 1909 Mackenzie postulated that an episode of anginal pain. “interconnections” forming between visceral and somatic fibers could explain the phenomenon of 294

Chest pain syndromes MYOCARDIAL ISCHEMIA Normal coronary arteries are end arteries. In the (whatever its causes) normal condition, myocytes subtended by a coro- nary artery are entirely dependent on that artery and GATE Chemical Cardiac cannot “borrow” from a neighboring artery. In the stimuli receptors abnormal condition of ischemia, myocytes release vascular growth factors (neovascularization) and Mechanical Cardiac change cellular activity to minimize the effect of stimuli? afferent nerves ischemia (ischemic preconditioning) [8]. New ves- sel growth and the development of collateral blood Peripheral inhibitory Dorsal horn supply in ischemic muscle, be it skeletal or myocar- stimuli (DNIC) convergent neurones dial, have long been recognized. In cases of ischemic heart disease, coronary angiography may show total Stimuli summation Spinothalamic tract proximal occlusion of a main coronary whilst radio- Descending nuclide studies show normal myocardial perfusion and no evidence of myocardial damage. This can only inhibitory stimuli be possible with full collateralization of the coronary circulation that has been achieved prior to occlusion SECOND GATE? Thalamus of the native artery. The processes of full collaterali- Thalamocortical tract zation and ischemic preconditioning take time which Pain perception is may explain why young, apparently fit men drop influenced by psychosocial Cortical decoding dead from the first infarct while elderly patients with a long history of angina tend to have small surviveable background infarcts. Ϯ ANGINAL PAIN Cardiac syndrome X Cardiac syndrome X affects around 2% of post- Figure 23.1 Proposed neural pathway for angina perception. menopausal women and is rarely seen in younger women and men. The diagnosis of cardiac syndrome Because of the common assumption that the pres- X requires a triad of typical angina, objective evidence ence of angina means the existence of coronary of ischemia, e.g. positive treadmill test, nuclear scan or atheroma, it is but a short step to assume that severe stress echo and angiographically “normal” coronary angina equals severe (i.e. dangerous, life-threatening) arteries. Ironically, because the orthodox simplistic coronary disease. The fact that angina is a visceral cardiac view of angina arising from obstructing coro- pain is commonly forgotten by patients and health- nary disease was not applicable, angina arising in the care professionals alike, who apply overly simplistic absence of coronary atheroma has attracted dispropor- “rules” regarding somatic pain, i.e. a more or less pro- tionate academic attention. Unfortunately, the general portional relationship between injury, signal intensity understanding of the condition remains poor and there and perception. is no consensus on a clear and effective therapeutic strategy. Consequently sufferers usually feel neglected Angiographic studies demonstrate that there is or even dismissed by their healthcare advisors. no correlation between the severity, or otherwise, of the pathologic process and the intensity of the Syndrome X patients typically describe painful, sensation of angina that is generated. It is estimated disabling angina that is often triggered by physical or that up to one-third of myocardial infarctions are emotional stress. The level of effort required to trig- painless. There is no evidence for the existence of ger an episode of pain is often highly variable with cardiac nociceptors, and it is illogical to presume clusters of “good” and “bad” periods. Most syndrome that the heart alone among the viscera should X sufferers who experience significant angina have, develop them. The “rules” that appear to govern at one time or another, been left with the impression nociception (i.e. that there is a relationship between that their doctors do not believe them. This creates an stimulus intensity and perception) do not apply in angina. 295

Chapter 23 additional emotional burden that is easily recognized Case scenario by healthcare professionals who understand pain. A 53-year-old plumber has recently experienced a The genesis of ischemia in syndrome X remains return of his symptoms of a sensation of pressure in obscure. Clinicians may describe microvascular angina, the central region of his chest. Over the past 4 months often telling patients that “the small arteries are dis- it has increased in frequency and now he has daily eased.” While there is evidence of some abnormalities episodes of pain. As the sensation begins, it is rela- in the endothelium of a large proportion of syndrome tively mild but builds progressively over 2–3 minutes X patients, this benign problem may be misunderstood until it reaches a plateau. It can remain for between 5 by patients who frequently imagine a continuous block- and 20 minutes. It varies in severity and when severe, ing up of blood vessels that are too small to operate on. it is accompanied by an ache in his mandible, pallor, A second problem in syndrome X relates to the way sweating and shortness of breath. It is sometimes ischemic signals are handled, with an apparent failure provoked by effort, sometimes by emotion, and occa- to control transmission at the thalamus where gating of sionally develops without any obvious associated ascending pain signals usually takes place [8]. Confusion activity. When the pain appears without provocation it and a lack of empathy from healthcare professionals is accompanied by anxiety that borders on panic. experienced by syndrome X sufferers often exaggerate maladaptive responses at a higher cortical level. The sensation is identical to the symptoms he originally experienced 6.5 years earlier when he had Syndrome X-like angina mistakenly supposed that he had indigestion. The It is important to appreciate that the diagnostic cri- “indigestion” had persisted for 2 weeks until one night teria for true syndrome X were invoked to assist in when it had become intolerable and his wife, con- research and there is a large subset of angina sufferers cerned about his pallor and general appearance, had who just fail to fulfil the diagnostic criteria for true called for an ambulance. On arrival at the Accident and syndrome X. Thus patients with mild coronary dis- Emergency department, an acute myocardial infarc- ease, ischemia and typical angina are often excluded. tion had been diagnosed, and after thrombolysis and Similarly, we see many patients with classic symptoms a period of observation, first on the Coronary Care and angiographically normal arteries but in whom Unit and subsequently on a general ward, he had ischemia has not been demonstrated. These patients been allowed home. He was advised to stop smok- feel, and often are, neglected. They often respond well ing, which he did. to a sympathetic hearing and a working diagnosis of syndrome X- like angina. Subsequently, a treadmill test following the modi- fied Bruce protocol showed signs of easily inducible As with all forms of pain of presumed cardiac ori- myocardial ischemia in the anterior and lateral chest gin, education and demystification form the core of leads, and an echocardiogram demonstrated a dys- good management. kinetic apical segment. Coronary angiography and ventriculography showed mildly impaired left ven- Prinzmetal angina – vasospastic angina tricular function with an ejection fraction of 50% and This rare angina variant is characterized by angina a number of significant (i.e. >70% luminal occlusion) with ECG changes that mimic acute myocardial coronary narrowings affecting the left main stem and infarction. As its name suggests, the cause is coro- left anterior descending artery proximal to the first nary spasm and is usually self-limiting, though case diagonal branch. The right coronary system had minor examples of spasm proceeding to myocardial infarc- atheromatous disease. He was referred for urgent cor- tion have been described. Vasospasm is often calcium onary artery bypass graft surgery which was carried channel dependent and patients usually respond to a out a few weeks later. He made an excellent recovery combination of education, demystification, relaxation from his surgery and was delighted to find that his training and high-dose calcium channel blockade. symptoms of angina had completely disappeared. Some colleagues have reported using temporary sym- Four months after his surgery he was able to return pathectomy despite the theoretical risks of inducing to work. spasm with this technique. Six years on, he is distressed by the return of his symptoms. He had been told that the bypass opera- tion “would last 10 years” and he fears that he will be forced into early retirement, which has implications 296

Chest pain syndromes regarding his pension arrangements and his ability to following acute coronary occlusion is the prior- help subsidize his daughter’s university course fees. ity following the diagnosis of angina. Such assess- A number of tests are carried out. He has taken aspi- ment, after taking a detailed history and conducting rin and a statin daily since his original presentation, a full clinical examination, generally includes some and he makes sure that he eats the recommended form of exercise testing with ECG and/or echocar- daily five helpings of fruit and vegetables, and some diographic analysis and if there is evidence of sig- mackerel or salmon once or twice a week. His serum nificant reversible myocardial ischemia, assessment cholesterol is 4.3 mmol/l and echocardiography of the coronary anatomy, generally by performing shows his left ventricular function is normal, with an angiography. ejection fraction of 60%. After coronary angiography (which shows that two of the three bypass grafts are The European Society of Cardiology summarizes still functioning well), he is told that another bypass the aims of treatment as: operation is not feasible and that the only treatment • to improve prognosis by preventing myocardial inf- to be offered is drug therapy. arction and death He feels that the cardiac specialist has told him, in • to minimize or abolish symptoms and/or their code, that there is “nothing more that can be done for him” (besides giving him more pills), which he finds effects. a depressing outlook. Each episode of chest pain he experiences reminds him that he is “living on bor- Prognostic interventions rowed time” and he believes that the pain represents There is overwhelming consensus based on large ongoing damage to his heart muscle. He recalls that studies that prognostic improvement may be brought when he was an inpatient 6 years previously a nurse about by appropriate lifestyle behaviors such as: opti- had told him that each episode of anginal pain was mizing weight; adoption of a “Mediterranean” diet “a bit like a mini heart attack.” His feelings of hope- comprising at least five portions of fresh fruit and lessness are increased when his cardiologist sug- vegetables daily and 2–3 portions of oily fish weekly; gests a visit to the pain clinic for advice regarding smoking cessation; programmed exercise (30 minutes control of his symptoms. of moderate to strenuous exercise five times a week); blood pressure control [10]. In addition, statins and Management of angina pectoris antiplatelet drugs have been shown to improve prog- The presentation of angina is often dramatic and nosis in patients with coronary artery disease and are occasionally life threatening. Unheralded myocar- recommended as standard in all CAD patients [11]. dial infarction due to coronary occlusion follow- In patients with impaired ventricular function or ing atheromatous plaque rupture may be the first diabetics with well-preserved ventricular function, symptomatic manifestation of coronary disease, angiotensin-converting enzyme inhibitors (ACEI) are but more commonly the initial presentation is with also strongly recommended [11, 12]. Current guide- increasing effort-related angina. Few patients cor- lines recommend the use of ß-blocking agents for the rectly self-diagnose angina and their first medical first year following myocardial infarction (MI) and in diagnosis is often incorrect. In our 12-year experi- patients with heart failure. ence of conducting detailed interviews with over 1500 chronic refractory angina patients, only three It is important to recognise thrombotic plaque correctly self-diagnosed the origin of their “anginal” events because emergency treatment to restore symptoms as cardiac. In many cases medical profes- coronary flow during a thrombotic occlusion event sionals also misdiagnose chest pain [9]. The initial will reduce the size of MI and death. Early refer- diagnosis is based on the history and depends heav- ral is important and there are well-established care ily on the location of symptoms. The assessment of pathways for suspected MI that are triggered by a risk of sudden death from myocardial infarction 999 call. Increasingly angioplasty is replacing intra- venous thrombolytic therapy as the treatment of choice for acute myocardial infarction but there is a limited window of opportunity and reopening an occluded vessel more than 48 hours later increases adverse outcomes [12]. The characteristic clinical 297

Chapter 23 presentation of MI (suddenly worsening angina Following the development of cardiopulmonary coupled with nausea, sweating, breathlessness and bypass techniques, CABG using autologous ves- angor animi) is often difficult to distinguish from sels became a useful and effective method of reliev- severe prolonged angina episodes. Current guide- ing symptoms. As described above, CABG improves lines recommend that a prolonged episode of prognosis in selected patients.. However, it is an angina that does not show signs of subsiding after expensive and time-consuming procedure which 15–20 minutes should be treated as a myocardial carries a notable perioperative morbidity and mor- infarction. tality. The introduction of coronary angioplasty and the development of stent placement techniques have In patients with atheromatous coronary dis- failed to live up to initial expectations of improved ease, only CABG surgery has strong support- prognosis and fewer symptoms. The proven reduc- ing evidence for prognostic improvement. Three tion in angina associated with angioplasty is short- large multicenter randomized trials, the Veterans lived and there is no benefit in terms of reduced Administration Cooperative Study (VA Study) [12], cardiovascular risk or need for further revasculari- the European Coronary Surgery Study (ECSS) [14], zation [17,18]. Two large randomized controlled and the Coronary Artery Surgery Study (CASS) [15], trials (RITA-2 and COURAGE) compared initial have compared the strategy of initial bypass surgery angioplasty strategies with conservative manage- with that of initial medical management in regard ment. RITA-2 showed that patients randomized to to long-term survival and symptoms for patients angioplasty experienced less angina at 3 months fol- with mild or moderate symptoms. A meta-analy- lowing randomization but were no better off at 3 sis of these three major randomized trials as well as years [18]. However, this temporary improvement other smaller trials has confirmed the survival ben- was paid for by a substantially increased risk (78%) efit achieved by surgery at 10 postoperative years for of experiencing a major adverse cardiac event. The patients with three-, two- or even one-vessel disease recent COURAGE trial compared PCI plus inten- that included a stenosis of the proximal left anterior sive pharmacologic therapy and lifestyle interven- descending (LAD) coronary artery [16]. The sur- tion (optimal medical therapy) with optimal medical vival rate of these patients was improved by surgery therapy alone in 2287 patients [19]. Like RITA-2, whether they had normal or abnormal left ventricle the patients randomized to PCI benefited initially (LV) function. For patients without a proximal LAD but these benefits were not sustained. COURAGE stenosis, bypass surgery improved the mortality rate is important because it demonstrates that PCI is no only for those with three-vessel disease or left main better than optimal medical therapy (OMT) in pre- stenosis. venting major adverse cardiac events and that it is safe for stable angina patients who do not require Many patients presenting with refractory angina prognostic bypass surgery to defer a decision to have previously undergone bypass surgery and undergo PCI until a trial of OMT. In a recent cost- no randomized studies have been undertaken to effectiveness analysis, the cost per quality-adjusted determine whether repeat bypass surgery improves life-year (QALY) for angioplasty was £47,000, prognosis. greatly exceeding the widely used “justification for treatment” threshold of £30,000 per QALY [19]. There is no evidence that prognosis is improved by Importantly, it is now commonly accepted that, for angioplasty in chronic stable angina. stable angina patients, a decision for angioplasty can be safely deferred until conservative management has Symptomatic interventions been tried. Since the latter half of the 20th century the sympto- matic management of angina pectoris has concen- This has led to a reappraisal of the management trated on drug therapies and revascularization. Only of chronic stable angina where no prognostic gain a small number of drugs have been shown to improve can be made by revascularization. By considering prognosis in coronary disease (see above). Several angina as a pain management problem rather than an classes of drugs are effective in relieving angina with- ischemia problem, and by applying evidence-based pain out altering prognosis; the use of these drugs is often limited by unpleasant side effects. 298

Chest pain syndromes management interventions, significant improvements pain for those taking statins when compared to those in quality of life can be made. The American College who do not [22]. Although complex lipid-lowering of Cardiology (ACC) makes reference to this fact in strategies may be suggested [23], some patients may, its Guidelines for the Management of Stable Angina: after appropriate deliberation, choose to accept a small increase in risk by stopping statin therapy (which may Because the presentation of ischemic heart disease be offset by other lifestyle changes) for a reduction in is often dramatic and because of impressive recent unpleasant statin-related side effects. technological advances, healthcare providers tend to focus on diagnostic and therapeutic inter- Harmful misconceptions are common amongst ventions, often overlooking critically important angina suffers and their family and friends [2,23]. The aspects of high quality care. Chief among these majority of patients and carers believe that each epi- neglected areas is the education of patients [20]. sode of pain represents ongoing myocardial damage. In the face of such a widespread and damaging misun- A recent Health Care Commission audit confirmed that derstanding about angina, it is unsurprising that many many cardiac patients leave hospital without receiving angina patients and their carers seek to protect them- basic education [21]. Hard data on the effectiveness of selves by avoiding circumstances that might provoke education are hard to obtain because it is part of rou- angina. Mistaken beliefs are often reinforced by concen- tine medical care, recommended by all clinical guide- trating therapy exclusively on the “biologic” component, lines and GMC professional standards guidelines. i.e. the coronary narrowing. Identifying and correcting the patient’s misunderstandings is a vital first step. Patient-centered management of angina Modern healthcare systems encourage health profes- In that telling quotation from the ACC’s guidelines sionals to practice “patient-centered medicine.” Much (see above), it is clear that in modern cardiologic confusion exists about what this much-used (and practice, much valuable advice regarding lifestyle, much-abused) phrase actually means. We understand weight loss, smoking cessation and fitness improve- it to mean that patients should be enabled to define ment is overlooked in favor of a “disease-centered” the objectives of treatment, and the doctor and other approach to revascularization. In fact, many palliative healthcare professionals should form a “therapeu- revascularization procedures may be avoided once the tic alliance” with patients to help them achieve those patient understands the absence of prognostic gain. objectives. This should necessarily include education With an emphasis on optimizing quality of life and about the condition and the various treatment options using the available evidence, a simple and pragmatic so that patients can be active participants in their algorithm can be constructed (Fig. 23.2). long-term management. This approach to therapy will be familiar to most practitioners in pain medicine. Patient education The rewards achievable with patient education should If the annual expenditure on coronary disease is not be underestimated. Changing harmful miscon- considered, surprisingly little evidence exists in sup- ceptions improves quality of life [24] and it has been port of the many and varied recommended treat- known for some time that cognitive rehabilitation ments for control of anginal symptoms. programs improve symptoms and produce worth- while reductions in palliative coronary revasculariza- Most patients agree that the fundamental objectives tion. In the 1990s Lewin et al. found that the majority of treatment of angina pectoris are to survive as long of patients listed for bypass surgery improved so much as possible, with as high a quality of life as possible. that they changed their minds after attending a cogni- When questioned, the majority stated a preference for tive behavioral rehabilitation program [24]. Similarly, a shorter, happier life rather than a longer, more mis- Ornish showed that 70% of patients successfully erable one. This approach is highly relevant to patients avoided palliative revascularization by enrolling in a who, for example, develop intolerable muscular fatigue comprehensive rehabilitation program. This compared and pain caused by statins. The majority of patients favorably with a control group of patients who had taking statins are not significantly affected, although recently undergone “successful” revascularization in there is a multivariate odds ratio of 1.5 for any muscular whom only 75% avoided further revascularization [25]. 299

Chapter 23 Assess risk of acute Prognostic disease myocardial infarction: (see above): Initial Refer for CABG presentation • ECG with cardiac- • Treadmill test Nonprognostic sounding pain • Echocardiogram disease: – make • Coronary Palliative treatment differential diagnosis angiography Palliative management of angina: • Educate patients about their condition, including lifestyle advice (diet, smoking, exercise, etc.) • Optimize medication (NB: ‘optimal’ medication is rarely ‘maximal’ medication) • Transcutaneous nerve stimulation • Temporary sympathetic nerve block • Opioid analgesics • External enhanced counter-pulsation (EECP) • Consider review of coronary anatomy and angioplasty ϩ stent Figure 23.2 National Refractory Angina Centre (NRAC) treatment algorithm. In our center a brief cognitive intervention delivered angina sufferers, it also offers potentially enormous to 69 consecutive refractory angina patients resulted cost savings in terms of reduction in unscheduled in a clinically and statistically significant improve- admissions, avoidable palliative revascularization, with ment in symptoms, anxiety and depression and qual- concomitant reduction in periprocedural complica- ity of life [26]. In a larger consecutive series of 271 tions such as stroke, myocardial infarction and death. patients enrolled over a 5-year period we compared admissions and myocardial infarct rates before and Current all-comer mortality for first-time bypass after enrolment. A rising admission rate of 15.5 days surgery is 1.7% [29]. Major adverse event rate fol- per patient per year fell to 10 days per patient per year lowing planned PCI is difficult to determine with within a month of enrolment (P Ͻ 0.001). Thirty two precision because myocardial infarctions that are MIs were recorded in the year prior to enrolment not associated with the development of a Q wave compared to eight after (P Ͻ 0.001) [27]. on the ECG are not routinely recorded. Excluding these important events reduces the PCI complica- The key to brief cognitive intervention is mutual tion rates to around 0.5% [30]. In the North West respect and a shared understanding of the patients’ of England, the Cheshire and Merseyside and North experiences and objectives [28]. It must not be so Wales Cardiac Network introduced Stable Angina brief as to compound patients’ and carers’ confusion Guidelines in October 2007 that emphasize the by adding new information to pre-existing miscon- importance of incorporating cognitive intervention ceptions. It is crucial to identify core beliefs by asking into routine rehabilitation before medication and straightforward open-ended questions, taking time palliative revascularization [31]. to gently explore the patients’ and carers’ beliefs, and the provenance of those beliefs. Challenging beliefs Lifestyle advice and offering evidence-based alternatives or opinion Relaxation training. Catecholamine release invariably should be underpinned by an understanding of the accompanies angina and contributes to worsening patient’s beliefs. Not only does this time-consuming of the symptoms (pain, breathlessness, nausea and approach represent a worthwhile and effective inter- sweating). Calming techniques that enable patients vention in terms of quality of life improvement for to reduce endogenous adrenaline release and increase 300

Chest pain syndromes endogenous endorphin can effectively abort episodes dose-dependent relationship between alcohol intake of angina and/or prevent them becoming severe. and blood pressure [33]. On a population basis, the Research shows that cardiac events (mostly revascular- relationship between alcohol intake and coronary ization) were significantly less likely in patients trained mortality is a U or J shape. A recent long-term study in relaxation techniques. Learning a technique of stress of 45–64 year olds showed that the 6% who took up management was associated with a relative risk of 0.26 moderate alcohol were 38% less likely to develop of a cardiac event compared with control patients CAD compared to their persistently nondrink- receiving usual care [32]. ing counterparts [34]. The consistent association between moderate alcohol intake (2 units a day for Trust and understanding are required before some women and 3 units a day for men) and cardiovascu- patients will accept advice on calming techniques, lar risk reduction is attributed to beneficial effects on especially if they have formed the suspicion that HDL-cholesterol, fibrinogen and glucose intolerance doctors think their symptoms are “all in the head.” [35]. Despite these positive attributes, the European Fortunately, most patients will have worked out for Society of Cardiology guidelines on cardiovascular themselves that “getting worked up” worsens episodes disease (CVD) prevention do not recommend mod- of angina and many will have already adopted basic erate alcohol intake to prevent the development of relaxation techniques through trial and error. CVD. The reason for this appears to relate to con- cerns about abuse and toxicity. Exercise. Many patients and their carers mistakenly believe that undertaking exercise is more dangerous Smoking. All the evidence points to a direct causal than remaining sedentary and it is important to rec- relationship between inhaled cigarette smoke and ognize and challenge irrational fears about exercise. an increased risk of major cardiovascular events. Patients and carers will often relate stories of being told Unlike some lung diseases and cancers, cardiovas- to “take things easy” or “avoid overdoing it” following cular risk is not linearly related to dose and a single a heart attack and will talk of otherwise fit acquaint- cigarette significantly increases risk of MI. Patients ances who “dropped dead while jogging” as confir- and carers should be urged to consider complete mation of their view that exercise is best avoided. The cessation. Passive smoking should be discouraged. overwhelming expert consensus view is that regular exercise reduces cardiovascular risk and enhances well- Optimizing medication being. The standard advice of 30 minutes of moderate Many patients have fundamental misunderstand- to strenuous exercise five times a week should be given, ings about their antianginal medication. In many with additional advice on a slow progressive symptom- cases they mistakenly believe that symptomatic treat- avoiding program using pacing and goal-setting tech- ments carry an additional prognostic benefit. The niques [11]. Patients who understand their condition only groups of drugs for which there is clear evidence and the wide-ranging beneficial effects of exercise on that myocardial infarction and cardiac death rates are pain, weight, blood pressure, well-being and cardiovas- reduced are antiplatelet drugs (aspirin or clopidogrel) cular risk are more likely to adhere to advice. and statins [5]. β-Blockers and ACEI improve prog- nosis if there is measurable left ventricular dysfunc- Diet. A healthy diet is important. Most patients tion. β-Blockers are indicated for a year following an think that a regimen that avoids saturated fat is uncomplicated MI but there is no evidence of risk adequate. The current recommendations focus on reduction beyond a year after infarction. Thus for the the importance of a cardioprotective diet of at least majority of stable angina patients with normal ven- five portions of fresh fruit and vegetables a day and tricular function, most prescribed drugs (i.e. calcium two portions of oily fish a week. Compared with channel blockers; nitrates; potassium channel block- teetotallers and moderate to heavy drinkers, a daily ers; the new Iφ channel blocker ivabradine) and opio- consumption of modest amounts of alcohol (1–2 ids are for symptom palliation only. units for women and 2–3 units for men) is associ- ated with the lowest cardiovascular risk. Of these, It is important that the prescription of pallia- the most controversial is alcohol. There is a clear tive therapy should be subjected to a “cost–benefit” 301

Chapter 23 analysis with respect to symptomatic improvement wall. Episodes of angina that begin in the chest and measured against unpleasant side effects. Because extend to other areas may be controlled or abolished most antianginal drugs operate through their by TENS with appropriately positioned skin electrodes. vasodilating properties, there is potential for drug A case report describes how TENS may compromise interaction causing unpleasant side effects such as respiratory function in debilitated patients and leads postural hypotension, dependant edema and persist- should not be applied to the throat [37]. TENS for ent headache. The patient should be encouraged and angina is most successful when the area of pain referral helped to determine for himself whether his quality can be encapsulated by four skin electrodes (personal of life has been enhanced or reduced by a particu- communication, D Trenbath). Frequency and ampli- lar drug. With care, and a logical approach to thera- tude need to be adjusted to individual preference. peutic trial, drug intake can be limited to only those medications which control symptoms effectively Temporary sympathetic nerve block with an acceptable side-effect profile. Maximal med- In 1920 Jonnesco published a case report describing ication is often confused with optimal medication. It the beneficial effect on intractable angina of surgical may take some months for patients to optimize their extirpation of the left stellate ganglion. This treat- medication successfully. ment was used to treat patients with severe angina and heart failure. Because of high surgical morbidity Transcutaneous electrical nerve stimulation (TENS) the procedure was subsequently modified to alcohol In the 1960s Wall & Melzack proposed the gate con- ablation, a treatment which remained in use until the trol theory of pain transmission based on their obser- 1960s. The reason for the abandonment of this low- vations of neural activity in the dorsal horn of the risk and effective symptomatic treatment is obscure. spinal cord [34]. Since then TENS has become a pop- ular, safe and effective method of controlling various The growing recognition of neural plasticity in types of pain. Its benefits in treating angina pain are the peripheral and central nervous system has led to well established, although its use is not widespread in techniques causing a permanent neurologic lesion the UK. This may be due to a widely held, although becoming less popular among pain clinicians, espe- totally fallacious belief that TENS is potentially haz- cially in the knowledge that an anesthesia dolorosa ardous to angina sufferers because of potential inter- pattern may develop with time. The well-recognized ference in the heart’s conduction pathways. phenomenon of local anesthetic nerve block pro- ducing pain relief of a duration far exceeding the There is evidence that TENS improves quality longevity of the drug has led to a re-examination of of life in patients with angina, and that it has been temporary stellate ganglion block and upper thoracic shown to increase coronary flow.Three groups of paravertebral block in the management of chronic patients with heart conditions (34 with syndrome stable angina. Temporary cardiac sympathectomy X, 15 with CAD, 16 following heart transplant) had may result in many weeks of freedom from pain with coronary blood flow velocity (CBFV) estimated fol- low risk of serious complications without requiring lowing TENS. The first two groups showed signifi- fluoroscopy. Following an initial case report con- cant increases in CBFV, but the third group did not, firming prolonged relief from angina following stel- leading the researchers to conclude that TENS-related late ganglion block, a 2-year unblinded prospective changes in CBFV were mediated by neural mecha- audit examined the effects of a total of 227 stellate nisms [35]. This adds to the earlier observation (in ganglion blocks performed on 46 patients with coro- two groups of 13 and 23 patients) that TENS ther- nary artery disease and refractory angina. Mean dura- apy applied in three 1-hour daily sessions increases tion of freedom from pain was 3.5 weeks (although work capacity, reduces ST segment depression during one patient reported more than 40 pain-free weeks), exercise, reduces frequency of angina episodes and with a low complication rate (2%). Only one patient reduces nitrate consumption [36]. required overnight hospitalization following a com- plication, and no permanent harm was suffered [38, Transcutaneous electrical nerve stimulation is less 39]. Although a long-term commitment between pain successful in abolishing or reducing pain when the clinician and patient is required for repeated nerve primary area of pain referral extends beyond the chest 302

Chest pain syndromes blocks to be a valid method of symptom control, It is important to include the patient as an equal some patients find it an acceptable strategy. partner in decision making when considering the use of these drugs in managing long-term pain problems Opioid drugs [45]. Most health professionals are accustomed to treat- ing chest pain in patients admitted as an emergency Intrathecal opioid therapy. Implantable pumps deliv- with strong opioid analgesics. There is a growing ering microvolumes of analgesic drugs have been acceptance that it is appropriate to treat patients with developed since the 1990s. The principle behind their severe chronic nonmalignant pain with oral modi- use is the ability to directly target drugs to their site fied-release preparations of opioid drugs. This notion of action rather than relying on intermediary mecha- was first presented in the 1980s by Portenoy and was nisms of transport. There are a number of drawbacks initially regarded as controversial. Concerns among to the use of these devices. physicians and patients of inducing dependence, • Implantation of the drug delivery system requires addiction and ever-escalating doses have been exten- sively considered [40]. There remains a reluctance to an invasive procedure. prescribe opioids among some groups of physicians. • The patient and doctor are committed to a long-term An American study showed that geriatricians were more likely than internal medicine specialists to pre- relationship which can damage patient autonomy. scribe opioids to elderly patients in pain [41]. The • A lengthy titration period may be necessary while British Pain Society has published evidence-based guidelines about the prescription of strong opioid previous medication is adjusted to allow the new drugs in chronic nonmalignant pain [42]. The princi- drug – and route of administration – to \"bed in.\" ples described may be applied as effectively in angina • Equipment costs are considerable and pump replace- management as in any other chronic opioid-sensitive ment may be required. pain condition. Side effects often limit the usefulness • Intrathecal granuloma formation at the catheter of opioids and some authorities have suggested that tip, which may cause spinal cord compression, has this may be overcome by delivering opioids intrathe- been reported. The overall incidence is unclear, cally because lower doses are required. but in a series describing 41 cases of catheter tip granuloma in chronic pain patients receiving As more is understood about endogenous opioid intrathecal drugs, the use of high concentrations cell membrane receptors and their interactions with of morphine was most strongly linked to granu- intracellular processes, it is becoming clearer that the loma development [46]. mechanisms of tolerance, dependence and addiction • Long-term outcomes are uncertain. A prospective case are varied and complex [43]. The use of terms such study followed 38 intrathecal pump recipients for 3 as “addiction” and “dependence” may become emo- years after implantation and concluded that although tive, and the inclusive phrase “problem drug use” has intrathecal analgesic administration may improve been preferred in some quarters. The growing list of pain, mood and function, this group of patients con- publications on this topic demonstrates that there tinue to experience severe symptoms [47]. is a small incidence of problem drug use in chronic Importantly, there is little evidence of overall benefit pain patients prescribed strong opioids. A recent to chronic angina patients using intrathecal infu- review of 67 publications on this issue has suggested sion pumps and worldwide experience is limited to an incidence of around 0.2% risk of addiction and a very small number of specialist units [48]. Overall, about 0.6% risk of aberrant drug-related behavior our own experience (at the National Refractory following the prescription of opioids to a chronic Angina Centre) of eight implants is discourag- pain patient with no previous history of drug or ing. Intractable nausea, inadequate pain control alcohol misuse. In patients who have previous his- and testicular suppression are among the problems tory of alcohol and/or drug misuse, the incidences of encountered. addiction and abberrant behavior were higher (3% and 11% respectively) [44]. Spinal cord stimulation (SCS) Following the development of TENS after Melzack & Wall’s pioneering work on spinal gating, interest in 303

Chapter 23 other neurostimulation techniques quickly gained Managing changes in symptoms ground. Analgesia was observed following direct stim- It is important to appreciate that coronary atheroma ulation of the spinal cord by a wire placed in the pos- progresses in a nonlinear fashion. Lesions may remain terior epidural space. Spinal cord simulation has since stable for years before an unpredictable plaque disrup- become an established treatment for neuropathic pain. tion event produces sudden progression or, less often, regression of the lesion. Rapid worsening in the sever- In the opinion of the authers, SCS can be an effec- ity of a coronary lesion causes ischemia at low work- tive technique in the control of refractory angina, but loads, resulting in more frequent and severe angina, there is insufficient evidence to support its cost effec- and it is important to have a plan to deal with such an tiveness in the UK (see NICE puplication: www.nice. event. It is sensible to negotiate with local cardiologists org.uk). For a more detailed examination of the role in advance so that a subsequent sudden deteriora- of SCS in pain medicine see Chapter 29. tion in symptoms is handled in the preferred manner. Generally speaking, cardiologists have a low threshold External enhanced counterpulsation (EECP) for investigating and treating patients who they sus- External counterpulsation by the use of sequen- pect may have a fresh clot in a coronary artery. tially inflated ECG-co-ordinated high-pressure cuffs around calves, thighs and buttocks has been investi- Even if the deterioration is related to an occlusive gated since the 1960s as a noninvasive method of rep- event, it need not cause a myocardial infarction. licating the effects of intra-aortic balloon pumping. Very severe narrowings are at high risk of occlusion Initially it was developed as a treatment for cardio- and such events are often associated with an abrupt genic shock. In a randomized controlled trial EECP deterioration in frequency and severity of symptoms. has been shown to benefit chronic refractory angina However, severe lesions usually occlude without caus- patients and has FDA approval [49]. Nevertheless, the ing myocardial infarction because of the development evidence for cost-effectiveness of EECP in the rou- of coronary collaterals [51]. In practice, worsening tine care of refractory angina patients is lacking and angina is more commonly the result of physical stress, a health technology assessment is under way. In our such as chest or urinary infections. Emotional stress practice, EECP is reserved for patients who would plays a major role in otherwise inexplicable deteriora- otherwise be candidates for redo bypass surgery or tions and is often underplayed by patients who are too spinal cord stimulation. EECP’s major advantage over embarrassed to mention personal matters. An effective the alternatives is that it can be tried without signifi- therapeutic alliance will avoid such problems. cant risk and withdrawn if it is intolerable. Funding involves a time-consuming process in England and Chest wall pain requires approval from the patient’s Primary Care Trust specialist commissioning group. The chest wall comprises skin, subcutaneous tissues, bony structures linked by symphysis, fibrocartilagi- Treatment involves attending for regular ses- nous and synovial joints, muscle, nerves, blood ves- sions to allow a total of 35 hours of therapy. This is sels and parietal pleura. Many conditions may present a major commitment and limits EECP’s usefulness with chest pain, descriptions of which are beyond the because it is only available in a few centers across the scope of this chapter. The most common clinical situ- UK. A recent review of data from a long-term fol- ations encountered by the pain clinician result from low-up registry suggests that clinically significant some kind of physical insult to one or more tissue benefits are seen in 70% of patients up to 3 years types. This may be accidental trauma, postsurgical or after the initial treatment [50]. There is no doubt as a consequence of infection. that EECP increases coronary perfusion during aug- mentation but why the effect lasts for so long after Chest wall trauma is common. Causes range from completion of the course of treatments is unclear. road traffic accidents to falls and brawls. The origi- Whatever the mechanism, EECP is a useful nonin- nal traumatic event may easily be forgotten if there vasive treatment option and is often preferred to is a delay between the event and the onset of pain. invasive therapies such as palliative surgery or spinal We have met patients who were unwilling to men- cord stimulation. tion their suspicions that an earlier trauma might be 304

Chest pain syndromes related to their current pain because a doctor had pre- Around 5% of 1000 refractory angina patients viously dismissed the idea. If rib fracture occurs, the referred with a supposed diagnosis of postrevascu- inflammatory process will cause the area to be painful larization recurrent angina were found to have other until the bone has healed. Injury to adjacent structures causes for their chest pain. A carefully taken his- is common following trauma. Fibromyalgia can result tory invariably revealed significant differences in the from traumatized soft tissues, including muscle and characteristics of their preoperative angina and the periosteum, and because the neurovascular bundle is presenting pain. When recurrent angina happens, even so close to the rib its structures may easily be damaged if it has not been present for many years, it almost by jagged bone fragments. It has been well established always follows the same pattern as it did before sur- that long-term pain experience can be influenced by gery. On careful questioning, postbypass patients with the reported level of pain during the acute phase [52]. noncardiac chest pain will often admit to experienc- ing their “old” angina under conditions of physical or If lung laceration or contusion has complicated the emotional stress. This will help distinguish true recur- injury the parietal pleura will be involved in the inflam- rent post-CABG angina from other chest wall pain matory process, and there is an increased risk of local syndromes. or generalized lung infection. Persistent cough due to pleural or lung injury immediately following trauma Other categories of surgery that commonly result is likely to be a determining factor in pain experience in chronic chest wall pain are thoracotomy and breast during the acute phase. surgery. Although there are numerous anecdotes and references to the frequency of occurrence, there is a Chest wall pain after surgery relative lack of accurate published data. A recent ret- Pain following surgery to the chest wall is common and rospective review of a cohort of 255 patients under- frequently under-reported. As with accidental trauma, going thoracotomy in a 2-year period received 149 persistent pain may arise from a variety of sources, and (58%) responses. The overall incidence of chronic it may be extremely difficult to elucidate the precise chest wall pain was 52% (32% rated “mild,” 16% cause of pain. Furthermore, there may be a tendency rated “moderate,” 3% rated “severe”) [55]. These pain to misdiagnose a new postoperative chronic pain as a syndromes frequently have a significant neuropathic recurrence of the original condition. For example, it is component, and may be helped by the various treat- well known that chest wall dysesthesia is common fol- ments for neuropathic pain described elsewhere. lowing CABG surgery. In a retrospective study of a 2-year cohort of CABG patients, 44% of 380 respond- Since the mid 1990s an incidence of around 40% ers to a questionnaire admitted to persistent dysesthesia of pain after breast surgery has been accepted as accu- of the chest wall [53]. This dysesthesia may be related rate. A cohort of 138 women previously reporting to the harvesting of the internal mammary artery pain after breast surgery was reviewed at a mean of 9 as part of the revascularization procedure, although years (SD 1.8 years) after initial operation; 113 (82%) incidence of pain is similar at 1 year (approximately women responded, of whom 59 (52%) reported per- 30%) with or without use of the mammary artery and sistent pain [56]. regardless of surgical technique, which were the find- ings of a prospective study of 349 consecutive patients Future directions for research undergoing CABG. Fourteen (4%) of these patients rated their pain as persistently greater than 50 mm on Despite calls for a comprehensive health strategy a visual analog scale [54]. going back over a decade, refractory angina remains in the shadows of stable angina. The clinical and Because there may be a significant delay between time financial implications of the condition are poorly of surgery and onset of pain, and because CABG is often understood and a comprehensive epidemiologic review successful in abolishing anginal symptoms for a variable is urgently needed. The lack of interest by inde- period, a proportion of cases of chest wall pain following pendent funding bodies has resulted in the creation CABG will be misdiagnosed as recurrent angina pectoris, of an artificial evidence base funded by the manufac- and the patient may be subjected to a series of unneces- turing and pharmacology industries that appears to sary and potentially hazardous investigations. support physical and pharmacologic interventions. 305

Chapter 23 No industry-sponsored trial in refractory angina has 7. Yellon D, Downey J. Preconditioning the myocardium: from required patients to have undergone optimal reha- cellular physiology to clinical cardiology. Physiol Rev 2003; bilitation. Similarly, none has used rehabilitation 83(4): 1113–1151. as a comparator. Randomized controlled trials that compare one expensive intervention with another will 8. Sekhri N, Feder GS, Junghans C, Hemingway H, Timmis not provide evidence of cost-effectiveness and are of AD. How effective are rapid access chest pain clinics? questionable value. Prognosis of incident angina and non-cardiac chest pain in 8762 consecutive patients. Heart 2007; 93: 458–463. It is hoped that the current UK NHS practice- based commissioning reforms, enabling primary care 9. European guidelines on cardiovascular disease prevention practitioners to identify cost-effective services that are in clinical practice. Eur J Cardiovasc Prev 2007; 14(suppl 2): best suited to the needs of their patients, will lead to S1–S113. an expansion of noninvasive pain management strat- egies for the treatment of this complex and disabling 10. European Society of Cardiology. Guidelines on the man- chronic pain condition. agement of stable angina pectoris: executive summary. Eur Heart J 2006; 27: 1341–1387. Authors’ recommendations 11. Hochman JS, Lamas GA, Buller CE, et al., Occluded Artery After 12 years of developing the UK’s first comprehen- Trial Investigators. Coronary intervention for persistent sive refractory angina service, we are convinced that fear occlusion after myocardial infarction. N Engl J Med 2006; and confusion dominate the clinical picture. Parallel 355(23): 2395–2407. lines of research demonstrate the critical importance of setting aside time to take a detailed biopsychosocial 12. VA Coronary Artery Bypass Surgery Cooperative Study history that identifies and deals with misconceptions. Group. Eighteen-year follow-up in the Veterans Affairs However, few are prepared or able to spend the 2–3 Cooperative Study of Coronary Artery Bypass Surgery for hours that we have found necessary to begin the proc- stable angina. Circulation 1992; 86: 121–130. ess of changing long-held damaging beliefs. We urge colleagues who plan to treat refractory angina patients 13. Varnauskas E. Twelve-year follow-up of survival in the ran- to resist management pressure to fall in line with exist- domized European Coronary Surgery Study. N Engl J Med ing outpatient arrangements. Make your clinic patient 1988; 319: 332. friendly and insist on at least 2 hours initial assessment for a new patient clinic. In our experience this time is 14. Passamani E, Davis KB, Gillespie MJ, Killip T. A rand- invaluable in preventing patients progressing to time- omized trial of coronary artery bypass surgery. Survival of consuming and costly interventions such as sympathec- patients with a low ejection fraction. N Engl J Med 1985; tomy and spinal cord stimulation. 312: 1665–1671. References 15. Yusuf S, Zucker D, Peduzzi P, et al. Effect of coronary artery bypass graft surgery on survival: overview of 10-year results 1. Drake R, Vogl W, Mitchell A (eds). Gray’s Anatomy for from randomised trials by the Coronary Artery Bypass Graft Students Elsevier, Philadelphia, 2005. Surgery Trialists Collaboration [published erratum appears in Lancet 1994; 344: 1446]. Lancet 1994; 344: 563–570. 2. http: //www.heartstats.org/datapage.asp?idϭ6799. 3. Mannheimer C, Camici P, Chester MR, et al. The problem 16. RITA-2 Trial Participants. Coronary angioplasty ver- sus medical therapy for angina: the second Randomised of chronic refractory angina; report from the ESC Joint Intervention Treatment of Angina (RITA-2) trial. Lancet Study Group on the Treatment of Refractory Angina. Eur 1997; 350: 461–468. Heart J 2002; 23: 355–370. 17. Boden WE, O’Rourke RA, Teo KK, et al. Optimal medical 4. http: //www.heartstats.org/uploads/documents%5C2007. therapy with or without PCI for stable coronary disease. N Engl J Med 2007; 356: 1503–1516. chapter2.pdf. 5. Foreman RD. Mechanisms of cardiac pain. Annu Rev 18. Griffin SC, Barber JA, Manca A, et al. Cost effectiveness of clinically appropriate decisions on alternative treatments Physiol 1999; 61: 143–167. for angina pectoris: prospective observational study. BMJ 6. Rosen SD, Camici PG. The brain–heart axis in the percep- 2007; 334: 624. tion of cardiac pain: the elusive link between ischaemia and 19. Gibbons RJ, Chatterjee K, Daley J, et al. ACC/AHA/ACP- pain. Ann Med 2000; 32(5): 350–364. ASIM guidelines for the management of patients with chronic stable angina: a report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines (Committee on Management of Patients With Chronic Stable Angina). J Am Coll Cardiol 1999; 33(7): 2092–2197. 20. Heart patient survey shows too many people are leaving hospital without aftercare and advice, June 2005. www. healthcarecommission.org.uk/newsandevents/pressreleases. cfm/widCall1/customWidgets.content_view_1/cit_id/1998. 21. Buettner C, Davis R, Leveille S, Mittleman M, Mukamal K. Prevalence of musculoskeletal pain and statin use. J Gen Intern Med 2008; 23(8): 1182–1186. 306

Chest pain syndromes 22. Jacobsen T. Towards pain-free statin prescribing: a clinical 40. Ballantyne JC, LaForge KS. Opioid dependence and addic- algorithm for diagnosis and managment of myalgia. Mayo tion during opioid treatment of chronic pain. Pain 2007; Clin Proc 2008; 83(6): 687–700. 129(3): 235–255. 23. Furze G, Lewin R, Murberg TA, Bull P, Thompson D. Does 41. Lin JJ, Alfandre D, Moore C. Physician attitudes toward it matter what patients think? The relationship between opioid prescribing for patients with persistent non-cancer changes in patients’ beliefs about angina and their psycholog- pain. Clin J Pain 2007; 23(9): 799–803. ical and functional status. J Psychosom Res 2005; 59: 323–329. 42. Recommendations for the appropriate use of opioids for per- 24. Lewin B, Cay E, Todd I, Soryal I, Goodfield N, Bloomfield P, sistent non-cancer pain, 2005. www.britishpainsociety.org. Elton R The angina management programme: a rehabilita- tion treatment. Br J Cardiol 1995; 2(8): 221–226. 43. Christie MJ. Cellular neuroadaptations to chronic opioids: tolerance, withdrawal and addiction. Br J Pharmacol 2008; 25. Ornish D. Avoiding revascularization with lifestyle changes: 154(2): 384–396. the Multicenter Lifestyle Demonstration Project. Am J Cardiol 1998; 82(10B): 72T–76T. 44. Fishbain DA, Cole B, Lewis J, Rosomoff HL, Rosomoff RS. What percentage of chronic non-malignant pain patients 26. Moore RK, Groves D, Bateson S, Barlow P, Hammond C, exposed to chronic opioid analgesic therapy develop Leach AA, Chester MR. Health related quality of life of abuse/addiction and/or aberrant drug-related behaviours? patients with refractory angina before and one year after A structured evidence-based review. Pain Med 2008; 9(4): enrolment onto a refractory angina program. Eur J Pain 444–459. 2005; 9: 305–310. 45. Blake S, Ruel B, Seamark C, Seamark D. Experiences of patients 27. http: //heartsurgery.healthcarecommission.org.uk/Survival. requiring strong opioid drugs for chronic non-cancer pain: a aspx. patient-initiated study. Br J Gen Pract 2007; 57(535): 101–108. 28. http: //www.bcis.org.uk/resources/audit/audit_2006. 46. Coffey RJ, Burchiel K. Inflammatory mass lesions associated 29. http: //www.cmcn.nhs.uk/guidelines/stable_angina.htm. with intrathecal drug infusion catheters: report and obser- 30. Blumenthal JA, Jiang W, Babyak MA, et al. Stress manage- vations on 41 patients. Neurosurgery 2002; 50(1): 78–87. ment and exercise training in cardiac patients with myo- 47. Thimineur MA, Kravitz E, Vodapally MS. Intrathecal opioid cardial ischemia: effects on prognosis and evaluation of treatment for chronic non-malignant pain: a 3 year pro- mechanisms. Arch Intern Med 1997; 157: 2213–2223. spective study. Pain 2004; 109(3): 242–249. 31. Xin X, He J, Frontini MG, Ogden LG, Motsami OI, Whelton PK. Effects of alcohol reduction on blood pressure: a meta- 48. Cherry DA, Gourlay GK, Eldredge KA. Management of analysis of randomised controlled trials. Hypertension 2001; chronic intractable angina – spinal opioids offer an alterna- 38: 112–117. tive therapy. Pain 2003; 102(1-2): 163–166. 32. King DE, Mainous AG 3rd, Geesey ME. Adopting moderate alcohol consumption in middle age: subsequent cardiovas- 49. Arora RR, Chou TM, Jain D, et al. The multicenter study cular events. Am J Med 2008; 121: 201–206. of enhanced external counterpulsation (MUST-EECP): 33. Mukamal KJ, Jensen MK, Grønbaek M, et al. Drinking effect of EECP on exercise-induced myocardial ischemia frequency, mediating biomarkers, and risk of myocardial and anginal episodes. J Am Coll Cardiol 1999; 33(7): infarction in women and men. Circulation 2005; 112(10): 1833–1840. 1406–1413. 34. Melzack R, Wall PD. Pain mechanisms: a new theory. 50. Loh PH, Cleland JG, Louis AA, et al. Enhanced exter- Science 1965; 150(699): 971–979. nal counterpulsation in the treatment of chronic refrac- 35. Chauhan A, Mullins PA, Thuraisingham SI, Taylor G, Petch tory angina: a long-term follow-up outcome from the MC, Schofield PM. Effect of transcutaneous electrical nerve International Enhanced External Counterpulsation Patient stimulation on coronary blood flow. Circulation 1994; 89: Registry. Clin Cardiol 2008; 31(4): 159–164. 694–702. 36. Mannheimer C, Carlsson C-A, Emanuelsson H, Vedin A, 51. Davies MJ. The pathophysiology of acute coronary syn- Waagstein F, Wilhelmsson C. The effects of transcutaneous dromes. Heart 2000; 83: 361–366. electrical nerve stimulation in patients with severe angina pectoris. Circulation 1985; 71: 308–316. 52. Curran N, Brandner B. Chest pain following trauma. 37. Mann CJ. Respiratory compromise: a rare complication of Trauma 2005; 7(3): 123–131. transcutaneous electrical nerve stimulation for angina pec- toris. J Accident Emerg Med 1996; 13: 68. 53. Bar-El Y, Gilboa B, Unger N, Pud D, Eisenberg E. 38. Chester M, Hammond C, Leach A. Long-term benefits of Skeletonized versus pedicled internal mammary artery: stellate ganglion block in severe chronic refractory angina. impact of surgical technique on post CABG surgery pain. Pain 2000; 87: 103–105. Eur J Cardiothorac Surg 2005; 27(6): 1065–1069. 39. Moore R, Groves D, Hammond C, Leach A, Chester MR. Temporary sympathectomy in the treatment of chronic 54. Meyerson J, Thelin S, Gordh T, Karlsten R. The incidence of refractory angina. J Pain Symptom Manage 2005; 30(2): chronic post-sternotomy pain after cardiac surgery – a pro- 183–191. spective study. Acta Anaesthesiol Scand 2001; 45(8): 940–944. 55. Pluijms WA, Steegers MA, Verhagen AF, Scheffer GJ, Wilder- Smith OH. Chronic post-thoracotomy pain: a retrospective study. Acta Anaesthesiol Scand 2006; 50(7): 804–808. 56. Macdonald L, Bruce J, Scott NW, Smith WC, Chambers WA. Long-term follow-up of breast cancer survivors with post-mastectomy pain syndrome. Br J Cancer 2005; 92(2): 225–230. 307

This page intentionally left blank

PART 3 Cancer pain

This page intentionally left blank

CHAP TE R 24 Oncologic therapy in cancer pain Rita Janes and Tiina Saarto Department of Oncology, Helsinki University Hospital, Helsinki, Finland Background therapy is well documented. Radiotherapy is effective also in treating pain caused by soft tissue tumors. The Cancer pain becomes more frequent as the malignancy pain-relieving effect of bisphosphonates in the treat- progresses, with one-third reporting pain at the time ment of bone metastases is relatively low, but they are of diagnosis and 60–80% of patients with advanced used to slow down progression of bone metastases and cancer. Tumor growth and metastasis can cause pain in prevent skeletal complications including pain, hyper- any organ but no single site is predestined to be pain- calcemia, fractures and spinal cord compression. ful. The skeleton is the most common site of cancer pain because several of the most prevalent malignan- In this chapter we will focus on evidence for the use cies, i.e. breast, prostate and lung cancer, have a pro- of radiotherapy, radionuclides and bisphosphonates in pensity for bone secondaries and more than half of treating malignant bone pain and that of radiotherapy those who have skeletal disease experience bone pain. in the treatment of pain caused by soft tissue tumors. Mechanisms by which cancer causes pain include tis- sue damage and inflammation, nerve compression and Epidemiology infiltration, increased intracranial pressure, obstruction of hollow organs and distension of capsules surround- Cancer pain affects the majority of patients with ing internal organs such as the liver and spleen. advanced cancer. At the time of diagnosis one in three patients report cancer-related pain whereas 74% Analgesics are the mainstay for treating cancer (53–100%) of those with advanced malignancy have pain. Specific cancer therapies including surgery, can- cancer pain [1]. Pain syndromes were assessed in a cer medicine and radiotherapy rarely relieve pain prospective study of 2266 cancer patients; 30% pre- quickly enough to render the use of analgesics need- sented with one, 39% with two and 31% with three less. Oncologic treatments alleviate cancer pain mainly or more distinct pain syndromes. The majority of by removing or downsizing pain-causing tumors. The patients had pain caused by cancer (85%) or antineo- anti-inflammatory effect of some treatment modalities plastic treatment (17%) and 9% had pain not related such as corticosteroids, radiotherapy and bisphospho- to cancer. Pain could be classified as originating from nates is also of significance in relieving cancer-related nociceptors in the bone (35%), soft tissue (45%) or pain. Radiotherapy is the single most effective onco- visceral structures (33%) or otherwise of neuropathic logic treatment of cancer pain. In the treatment of origin (34%). In most patients, pain syndromes were bone metastases, the pain-relieving efficacy of both located in the lower back (36%), abdominal region external radiotherapy and systemic radionuclide (27%), thoracic region (23%), lower limbs (21%), head (17%) and pelvic region (15%) [2]. Evidence-Based Chronic Pain Management. Edited by C. Stannard, E. Kalso and J. Ballantyne. © 2010 Blackwell Bone is the third most common metastatic site Publishing. after the lung and liver [3]. Several of the most preva- lent malignancies, i.e. breast, prostate and lung cancer, have a propensity for bone [3]. More than two-thirds 311

Chapter 24 of patients with metastatic breast or prostate cancer or osteosclerotic, i.e. osteoblast activity and bone and 30–50% of those with lung cancer will develop formation are increased. In multiple myeloma, bone metastatic bone disease. Skeletal metastases are often lesions are osteolytic whereas osteosclerotic metastases present also in thyroid and renal cancer and multiple predominate in prostate cancer. Patients with other myeloma. The median survival after the appearance cancers usually present with mixed bone lesions. Bone of bone metastases varies between 2 and 4 years for secondaries are seldom solitary, with the exception of patients with cancer arising from the breast, prostate renal cancer. More than 80% of bone metastases are or thyroid and those with multiple myeloma, while the found in the axial skeleton [6]. The most commonly prognosis for lung and renal cancer is shorter, less than affected areas are the spine, pelvis, ribs, proximal a year. Five to 40% are alive at 5 years depending on the thigh, upper arm bone and skull [3, 4]. As metastatic tumor histology and tumor burden [4]. Approximately bone destruction progresses, the risk of pathologic 65–75% of patients with bone secondaries suffer from fracture increases. Eight to 30% of patients with skel- bone pain [5]. etal metastases suffer from pathologic fractures [4, 5]. The underlying malignancy is most often multiple The incidence of pain arising from nonosseous myeloma or breast cancer because of the osteolytic metastases is also high, though less well docu- nature of the metastases and the relatively long prog- mented [6]. Among soft tissue tumors, the prevalence noses [6]. Affected areas include the ribs, pelvis, long of pain in lung cancer is best recorded. Approximately bones and vertebrae. Spinal cord compression occurs 60% (25–62%) of patients with locally advanced inop- in 5–10% of patients with skeletal disease and may erable primary lung cancer suffer from chest pain; cause severe neuropathic pain [3–5, 9]. almost one half have moderate to severe pain [7]. Anecdotally, in one series 84% of pancreatic cancer The pathophysiologic mechanism of pain in patients with inoperable disease suffered from pain [8]. patients with bone metastases not associated with fracture or spinal cord compression is poorly under- Current understanding of relevant stood. Not all bone metastases are painful. About two- pathophysiology thirds of demonstrated sites of bone metastases are painless and every sixth to every third of patients with Cancer pain is caused by various mechanisms includ- bone metastases do not suffer from pain [5, 9]. The ing tissue damage and inflammation, nerve compres- presence of pain does not correlate with the type of sion and infiltration, increased intracranial pressure, tumor, location, number or size of metastases [9]. It obstruction of hollow organs and distension of cap- is poorly understood why pain patterns vary between sules surrounding internal organs such as the liver and patients and even within a patient. Various factors spleen. Pain caused by tissue damage, i.e. (somatic or contribute in causing bone pain: mechanical stress visceral) nociceptive pain, is the most common type of the weakened bone, direct destruction of the bone, of cancer pain. Pure neuropathic pain is less common, microfractures, periosteal distension, nerve entrap- even though neuropathy as a part of the pain is seen ment and tumor pressure on adjacent tissues. The in every third cancer patient suffering from pain. This mechanisms underlying bone cancer pain also involve chapter focuses on the pathophysiology of bone pain. changes in osteoclastic activity leading to increased bone resorption, and inflammatory activity pro- Skeletal metastases are one of the most common voked by cytokine and prostaglandin production by causes of cancer pain [3]. Bone pain is often a dull the cancer cells [10]. Thus bone cancer pain appears continuous ache that increases in intensity with time. to be driven simultaneously by different mechanisms In addition, many patients suffer from incident pain including tumorigenic, inflammatory and neuro- i.e. pain exacerbated by pressure or movement, which pathic mechanisms. The healthy bone is constantly may be difficult to control with analgesics alone due undergoing a complex process of remodeling charac- to the quick changes in the level of pain from one terized by two opposing actions: the resorption of old moment to another along the course of a day. This bone by osteoclasts and the formation of new bone leads to impaired mobility. by osteoblasts. When healthy, there is a steady-state balance or “coupling” of osteoclastic bone resorption Bone metastases are either osteolytic, i.e. there is increased osteoclast activity and bone resorption, 312

Oncologic therapy in cancer pain and osteoblastic bone formation. This balance is lost be the most important mechanism. Significant tumor when tumor cells enter the bone microenvironment. shrinkage would not be expected when using the The tumor induces excessive osteoclast bone resorp- relatively low total doses (e.g. 4 Gy) which have been tion. In lytic bone metastases, increased osteoclast shown to relieve bone pain. Further evidence comes activity and bone resorption decrease bone density from the absence of a clear relationship between radi- and disrupt the skeletal architecture, either at focal osensitivity of the primary tumor and the analgesic sites or generally throughout the skeleton. effect [9, 11]. Mechanisms which explain the rapid pain response may include the inhibitory effect radio- Actively resorbing bone releases a number of therapy has on osteoclast activity and other relevant bone-derived growth factors and cytokines, e.g. pros- host cells such as macrophages, resulting in down- taglandins, bradykinin, substance P, cytokines (e.g. regulation of the release of chemical mediators of the interleukin (IL)-1, IL-6 and tumor necrosis factor inflammatory response [6, 9]. (TNF)), which attract circulating cancer cells to the bone surface and facilitate the tumor cells’ growth External beam radiotherapy using linear accel- and proliferation [9, 11]. Also sensory neurons are erators or cobalt machines is the most common way excited and/or sensitized by these growth factors. to deliver radiotherapy either, as in most cases, to a Malignant cells secrete prostaglandins, cytokines and local field or, when treating painful disseminated growth factors, which increase inflammatory activity bone metastases, to a wide field. Bone-seeking radio- and may also directly excite primary afferent nocicep- isotopes, samarium-153 and strontium-89, are used tors located in bone [9, 11]. This vicious cycle caus- to treat disseminated symptomatic bone disease. ing multidirectional interactions between tumor cells, Several clinical trials have shown that external beam bone cells and nociceptors leads to increased tumor radiotherapy and systemic radionuclide therapy are growth, osteolysis and pain. effective in treating painful bone metastases. There are only a limited number of studies that address Radiotherapy the question of how efficient radiotherapy is in relieving pain caused by soft tissue tumors. Radiotherapy uses high-energy X-rays, γ rays or electrons to deliver ionizing radiotherapy. Ionizing Local field radiotherapy in the treatment radiation causes DNA damage, leading to cell death. of bone metastases While tumor cell death is responsible for shrinkage Radiation therapy alleviates metastatic bone pain effi- of the tumor and thus improvement of pain, the pre- ciently in the majority of patients. It is particularly cise mechanism of action by which radiation results useful in treating metastatic bone pain that is not in pain relief remains uncertain. Tumor cell kill and controlled with adequate pain medication and is often shrinking of the tumor bulk in bone promote bone of particular help in controlling pain exacerbated by healing by enabling osteoblastic repair and restored pressure or movement. integrity of the damaged bone. There is evidence of degeneration and necrosis of cancer cells followed by In two large controlled studies, complete or partial replacement with proliferative fibrous tissue, which pain relief was obtained in 30–60% and 70–80% of then aggregates and becomes calcified, filling in an patients, respectively. The onset of pain relief varied osteolytic lesion [12]. This is evident on a plain X-ray from a few days to 4 weeks and the duration of pain where recalcification of osteolytic lesions is seen in relief between 3 and 6 months [14, 15]. In the Bone 65–85% of treated sites [13]. However, it takes several Pain Trial Working Party prospective randomized months for osteolytic bone lesions to heal whereas the trial, 765 patients with painful skeletal metastases pain-relieving effect is usually seen within a few weeks requiring palliative radiotherapy were entered into a and in some cases within a few days [11]. The rapid study comparing 8 Gy single fraction with a multi- onset of pain relief cannot be explained by the tumor fraction regimen of 20 Gy in five fractions or 30 Gy in cell kill and shrinking of the tumor mass alone [11]. 10 fractions. At baseline 29% of patients had mild pain, The perceived absence of a dose response relationship 44% moderate and 23% severe pain. There were no also suggests that downsizing of the tumor may not differences in pain relief: 78% of patients both in the single fraction and multifraction schedules obtained 313

Chapter 24 at least 50% pain reduction, and 57% and 58% had a patients at some time during the trials. There were complete response. There were no differences in the no differences in the proportions of patients achiev- time to first improvement in pain, time to complete ing these outcomes between single or multiple frac- pain relief or time to first increase in pain at any time tion schedules. The number needed to treat (NNT) up to 12 months from randomization, nor in the class to achieve complete relief at 1 month was 4.2 (95% of analgesics used. Retreatment was twice as common confidence interval (CI) 3.7–4.7). No pooled esti- after a single fraction compared with multifraction mates of speed of onset of relief, or of its duration, therapy. There were no significant differences in the could be obtained. In the largest trial (759 patients) incidence of nausea, vomiting, spinal cord compression 52% of those who had complete relief had achieved it or pathologic fracture between the two groups. Overall within 4 weeks, and the median duration of complete survival at 12 months was 44% [15]. relief was 12 weeks. Adverse effect reporting was poor. There were no obvious differences between the vari- A Dutch study randomized 1171 patients to receive ous fractionation schedules in the incidence of nausea either a single 8 Gy fraction (n ϭ 585) or 24 Gy in and vomiting, diarrhea or pathologic fractures. The six fractions (n ϭ 586). The primary tumor was in authors concluded that radiotherapy is clearly effec- the breast in 39% of the patients, prostate 23%, lung tive at reducing pain from painful bone metastases. 25% and other locations 13%. Bone metastases were There was no evidence of any difference in efficacy located in the spine (30%), pelvis (36%), femur (10%), between different fractionation schedules, nor indeed ribs (8%), humerus (6%) and other sites (10%). The of a dose response with total dose of radiation [16]. main endpoint was pain measured on a pain scale from 0 (no pain at all) to 10 (worst imaginable pain). Another 949 patients with breast or prostate can- Median survival was 7 months. Response was defined cer and moderate to severe pain in 1–3 locations as a decrease of at least two points compared to the were entered into a randomized phase III study that initial pain score. The difference in response between compared 8 Gy single fraction with 30 Gy 10 fraction the two treatment groups proved not significant. schedule. Median survival was 9 months. No differ- Overall, 71% experienced a response at some time ences were seen in total or partial pain relief (15% during the first year. With regard to pain medica- versus 18% and 50% versus 48%, respectively). Acute tion, quality of life and side effects, no differences grade II–IV toxicity was more common among the between the two treatment groups were found. The patients receiving 30 Gy (17% versus 10%); late tox- total number of retreatments was 188 (16%), with 147 icity was rare in both arms (4%). The incidence of (25%) in the single fraction and 41 (7%) in the multi- subsequent pathologic fracture was 5% for the 8 Gy fraction group. It was shown that the level of pain was arm and 4% for the 30 Gy arm. The retreatment rate an important reason for retreatment. There were also was statistically significantly higher in the 8 Gy arm indications that doctors were more willing to re-treat (18%) than in the 30 Gy arm (9%) (P Ͻ 0.001) [17]. patients in the single fraction group because time to retreatment was substantially shorter in this group The efficiency of reirradiation for painful and the preceding pain score was lower. Unexpectedly, bone metastases more pathologic fractures were observed in the single Two groups have reported on reirradiation as part of fraction group, but the absolute percentage was low. the treatment of painful bone metastases. Jeremic et al. In a cost analysis, the costs of the 24 Gy in six frac- investigated the effectiveness of 4 Gy single frac- tions and the 8 Gy single fraction treatment schedules tion given for retreatment after previous single frac- were calculated at 2305 and 1734 euros respectively. tion radiotherapy and arrived at the conclusion Including the costs of retreatment reduced this 25% that the single fraction was effective and well toler- cost difference to only 8%. The saving of radiotherapy ated. Of the 135 patients retreated, 109 patients were capacity, however, was considered a major economic treated because of pain relapsing after 4 Gy (group I, advantage of the single dose schedule [14]. n ϭ 34), 6 Gy (group II, n ϭ 39) or 8 Gy (group III, n ϭ 36), while 26 patients were reirradiated after ini- In a Cochrane analysis, radiotherapy produced tial nonresponse (group I, n ϭ 12; group II, n ϭ 8; complete pain relief at 1 month in 395/1580 (25%) group III, n ϭ 6). Of the 109 patients reirradiated for patients, and at least 50% relief in 788/1933 (41%) 314

Oncologic therapy in cancer pain pain relapse, 74% responded and 31% had complete the efficacy of a single 8 Gy fraction with 20 Gy response. Patients with previous complete response in five fractions for this type of pain. Two hundred were more likely to achieve complete response. It is and seventy two patients from 15 centers were ran- worth noting that among the 26 patients who initially domized. Eligible patients had radiologic evidence did not respond, there were 12 (46%) responses. There of bone metastases from a known malignancy, no were no differences between the three initial treatment other metastases along the distribution of the neuro- groups regarding the efficiency of second radiother- pathic pain and no clinical or radiologic evidence of apy. Toxicity was low and only gastrointestinal [18]. cord/cauda equina compression. Primary endpoints were pain response within 2 months of commence- The Dutch Bone Metastasis Study mentioned ment of radiotherapy and time to treatment failure above studied the effect of 8 Gy single fraction versus (TTF). 24 Gy in six fractions on painful bone metastases [14]. In this study, 24% of patients received retreatment The most common primary cancers were lung (31%), after single fraction versus 6% after multiple fractions prostate (29%) and breast (8%); index sites were spine (P Ͻ 0.001). A new study evaluated factors influenc- (89%), rib (9%), other (2%). The median overall sur- ing retreatment and its effect on response. Of the vival was 4.8 months. The intention-to-treat overall 137 patients who had received a single fraction and response rates (95% CI) for 8 Gy single fraction versus were retreated, 33% were given 8 Gy single fraction and 20 Gy in five fractions were 53% (45–62%) versus 61% 67% 24/4 Gy. The 36 patients having first had multiple (53–70%) (P ϭ 0.18). Corresponding figures for com- fraction therapy received single fraction 8 Gy in 75% plete response were 26% (18–34%) versus 27% (19– of cases and multiple fraction therapy 24/4 Gy in 25%. 35%) (P ϭ 0.89). The estimated median TTF (95% CI) Among the randomized 1171 patients, the response to were 2.4 mo(2.0–3.3 mo) versus 3.7 mo (3.1–5.9 mo) initial treatment was 71% after single and 73% after respectively. The hazard ratio (95% CI) for the com- multiple fraction radiotherapy (P ϭ 0.84). Retreatment parison of TTF curves was 1.35 (0.99–1.85), log-rank raised response to 75% for single fraction; multiple P ϭ 0.056. There were no statistically significant differ- fraction remained unaltered (P ϭ 0.54). The response ences in the rates of retreatment, cord compression or status after initial treatment did not predict occurrence pathologic fracture by arm [20]. of retreatment: 35% single fraction versus 8% multiple fraction nonresponders and 22% single fraction versus Wide-field radiotherapy and systemic 10% multiple fraction patients with progressive pain radionuclide therapy in the treatment were retreated. Logistic regression analyses showed the of disseminated painful bone metastases randomization arm and the pain score before retreat- Wide-field radiotherapy and radio-isotopes are used ment to significantly predict retreatment (P Ͻ 0.001), to alleviate pain caused by widespread painful skeletal i.e. intensive pain and single fraction treatment pre- metastases. They may also be administered with pro- dicted retreatment. Retreatment for nonresponders phylactic intent to reduce the number of new symp- was successful in 66% after single fraction versus 33% tomatic sites. after multiple fractions (P ϭ 0.13); retreatment for progression was successful in 70% after single frac- Wide-field radiotherapy tion versus 57% after multiple fractions (P ϭ 0.24). Scattered painful skeletal lesions may be treated with Overall, retreatment was effective in 63%. Irrespective single fraction and fractionated wide-field or half-body of response to initial treatment, physicians were more irradiation to the upper, lower or mid-body, depend- willing to retreat after a single fraction [19]. ing on the extent of metastases and symptoms. A single fraction of 6 Gy is given to the upper half-body field Radiotherapy in the treatment of in order to avoid pulmonary toxicity and 8 Gy to the neuropathic pain due to bone metastases lower half-body field. Half-body irradiation relieves There are very few data on radiotherapy for bone pain as effectively as local external radiotherapy. metastases causing pain with a neuropathic com- Response rates up to 73% have been reported with ponent. The Trans-Tasman Radiation Oncology 20% of patients reporting complete pain relief. Because Group undertook a randomized trial comparing half-body irradiation is usually given to patients with 315

Chapter 24 far advanced disease, more than half stay free from Systemic radionuclide therapy pain. Half of those who respond obtain pain relief Radionuclides, including strontium-89, samarium- within 48 hours and 80% within a week [21]. 153, rhenium-186 and rhenium-188, combined with bone-seeking agents have been used to deliver Half-body irradiation has also been administered ionizing radiation to widespread skeletal metas- with prophylactic intent. When compared to local tases. Most clinical experience has been gained using external radiotherapy, there is significantly less need strontium-89 and samarium-153, which are admin- for radiotherapy to new painful sites following half- istered intravenously, are localized in sites of active body irradiation [22, 23]. bone turnover, emit short-range β particles, are well tolerated and can be administered on an outpatient The side-effects caused by large single fractions basis. Indications include multiple painful bone sec- include nausea, vomiting, diarrhea, fever, transient ondaries that are positive on bone scan from prostate, increase in bone pain, hematologic toxicity and, breast and lung cancer, i.e. primaries with sclerotic rarely, pneumonitis [21]. Patients are usually admitted or mixed bone secondaries. Finlay et al. concluded in to hospital for intravenous hydration and premedica- their systematic review that radio-isotopes are effective tion consisting of antiemetics and corticosteroids. in providing pain relief with response rates between These side effects may be especially taxing for patients 40% and 95%. Pain relief starts 1–4 weeks after treat- with poor performance status. An interval of at least ment, continues up to 18 months, and is associated 4 weeks is recommended before administering the with a reduction in analgesic use in many patients. other half-body treatment or continuing chemother- Thrombocytopenia and neutropenia are the most apy to avoid severe hematologic toxicity. common toxic effects, but they are generally mild and reversible. Repeat doses are effective in provid- Salazar et al. continued to explore fractionated half- ing pain relief in many patients. The effectiveness of body irradiation to determine whether it was more systemic radionuclide therapy may be greater when efficient and less toxic compared to a single dose. They it is combined with chemotherapeutic agents such as concluded that fractionating half-body irradiation cisplatin [26]. eliminated the need for extensive premedication and patient monitoring required for single fraction [24]. Strontium-89 Finlay et al. identified 38 observational studies reporting A phase III trial for widespread symptomatic on the use of strontium-89 for the management of met- bone cancer from various primaries was carried out astatic bone cancer; of these, 16 studies were prospective in 156 patients to find the fastest and most efficient and had more than 20 patients and were analyzed. Pain method of delivery fractionated half-body irradia- was the main outcome measure but the differing crite- tion. There were three half-body (HBI) arms: (A) 15 ria used complicated the analysis of the data, although Gy/5 fractions/5 days; (B) hyperfractionation 8 Gy/2 complete response and lack of response were straight- fractions/1 day; (C) accelerated hyperfractionation forward to define. The proportion classified as complete 12 Gy/4 fractions/2 days. Pain relief was seen in 91% responders to strontium-89 ranged from 8% to 77% of patients (45% complete response and 46% par- (mean 32%), and the proportion showing no response tial response) within 3–8 days. Sixty three percent of ranged from 14% to 52% (mean 25%). Within this patients in arm A obtained complete pain response, range, 44% of patients had some degree of response to compared to 43% and 32% in arms C and B, which strontium-89 treatment, giving a mean overall response was significant between arms A and B (P ϭ 0.016). of 76%. Delay in the start of response was between 4 days Pain-free survival was 174 (A), 150 (B), and 122 (C) and 28 days, with a response duration up to 15 months. days. Toxicity was acceptable (41% none, 50% mild/ Although a reduction in analgesic use was a major cri- moderate, 12% severe but transitory); more was seen terion for assessing pain response, quantification of this with upper HBI [25]. The authors concluded that, for variable was poorly reported in the studies. In those most primary tumor types (except prostate), deliver- that did report on this variable, a reduction of between ing two HBI daily doses of 3 Gy on 2 consecutive days 71% and 81% was observed [26]. is as effective as delivering a daily dose of 3 Gy for 5 consecutive days. Thus, this is a faster and much more convenient HBI schedule for the palliation of pain in widespread cancer. 316

Oncologic therapy in cancer pain Buchali et al. carried out a double-blind, randomized A third randomized study, performed by the controlled trial of 49 patients with skeletal metastases European Organization for Research and Treatment of from prostatic carcinoma. Patients were assigned to Cancer Genitourinary Group, compared strontium-89 three injections of 75 MBq (2 mCi) strontium-89 with 150 MBq (4 mCi) with palliative local-field radiotherapy intervals of 1 month (25 patients) or to saline as pla- in 203 patients with hormone-refractory prostate can- cebo. There was no significant difference in pain relief cer and found no difference in subjective response (pain between the two groups, which might reflect the sub- score, dose of analgesics, performance status) [31]. optimum treatment regimen with a lower than usual dose per injection of strontium-89 used [27]. Another Quilty et al. compared the effect of strontium-89 study randomly assigned 26 patients with hormone- 200 MBq (5.4 mCi) local-field or half-body radio- refractory prostate cancer to strontium-89 150 MBq therapy. Two hundred and eighty four patients with (4 mCi) or to stable strontium as a placebo. This study prostate cancer and painful bone secondaries were assessed response to treatment by scoring the patients’ first stratified according to suitability for local or half- general condition, mobility, analgesic intake, and pain body radiotherapy, then randomly allocated that form analysis. Strontium-89 was substantially more effective of treatment or strontium-89. After 4, 8 and 12 weeks than placebo, both in patients receiving a single dose pain sites were mapped, toxicity monitored, and all (P ϭ 0.01) and in patients allocated to repeated doses additional palliative treatments recorded. All treat- (P ϭ 0.03). Complete responses were observed only in ments provided effective pain relief at the index sites; patients receiving radio-active strontium. The absolute improvement was sustained to 3 months in 63.6% risk reduction for patients achieving pain relief with after half-body radiotherapy compared with 66.1% strontium-89 was 0.321 (95% CI Ϫ0.035 to 0.678) [28]. after strontium-89, and in 61% after local radiother- apy compared with 65.9% in the comparable stron- The combination of external beam radiotherapy tium-89 group. Fewer patients reported new pain and radio-isotopes has been studied. In a double-blind, sites after strontium-89 than after only external radio- controlled, randomized trial in Canada, 126 patients therapy local or half-body radiotherapy (P Ͻ 0.05). with hormone-refractory prostate cancer received Radiotherapy to a new site was required by 12 patients local-field radiotherapy and were randomly assigned in the local radiotherapy group compared with two to strontium-89 or saline placebo. Strontium-89 was after strontium-89 (PϽ 0.01); there was no signifi- given as a single injection of 399.6 MBq (10.8 mCi), cant difference between half-body radiotherapy (six 2.5 times that approved by the US Food and Drug patients) and strontium-89 (nine patients) in this Administration. There was a significant reduction in respect. Platelets and leukocytes fell by an average analgesic use at 3 months in patients assigned to stron- 30–40% after strontium-89 but sequelae were uncom- tium-89, and significantly fewer new active sites were mon, and other symptoms rare [23]. Dearnaley et recorded. As could be expected, the groups did not al. compared retrospectively the pain response of differ in pain reduction at the index site [29]. A sec- 27 prostate cancer patients receiving half-body irra- ond placebo-controlled, randomized trial reported diation with 51 patients receiving strontium-89 and contrasting results. This study assessed the effective- found no difference. Performance status and extent of ness of concurrent strontium-89 150 MBq (4 mCi) or disease on bone scan were of over-riding importance saline placebo with palliative external radiotherapy in in determining outcome [32]. 95 patients with skeletal metastases from various pri- maries. The endpoint was physician-assessed response The outcome data presented by these studies for at 3 months, which took into consideration the pain strontium-89 are conflicting, although they generally score, dose of analgesics, performance status and imply similar subjective response rates to external additional pain treatment. At 3 and 6 months, no dif- radiotherapy. Variation in the definition of efficacy ferences between treatment arms were observed. The criteria and the differences in the populations of pretreatment use of opiates was independently associ- patients recruited could contribute to the inconsist- ated with short progression-free survival, i.e. patients encies reported. need to be treated before the skeletal disease is too far progressed [30]. There have been efforts to improve pain response and possibly survival by adding chemotherapy to strontium-89. Sciuto et al. assessed the efficacy 317

Chapter 24 of strontium-89 with or without low-dose cisplatin in over 4 weeks after treatment. Mild, transient myelo- 70 patients with hormone-refractory prostate cancer in suppression was the only side effect [35]. a prospective, randomized controlled trial. The propor- tion of patients reporting pain relief and the duration Sartor et al. enrolled 152 men with hormone- of response were significantly higher with combined refractory prostate cancer and painful bone secondar- treatment. Overall pain relief occurred in 91% of ies into a prospective, randomized, double-blind trial patients in the combined arm versus 63% (P Ͻ 0.01) comparing radioactive samarium-153 with nonra- with the median duration of 120 days versus 60 [33]. dioactive samarium. Patients were randomized (2:1) to the radioactive agent. Patients’ diaries were used Samarium-153 to record analgesic use and metastatic bone pain, Resche et al. assessed the therapeutic efficacy of two which was assessed with a visual analog scale and a doses of samarium-153 in alleviating metastatic bone pain descriptor scale. Nonresponders were informed pain from cancers of the prostate (n ϭ 67), breast of the treatment received after 4 weeks of treatment (n ϭ 36), lung (n ϭ 2), and other primaries (n ϭ 9). and, if initially treated with placebo, were allowed to Fifty five were randomly assigned to a single dose of receive samarium-153 in an open-label fashion. Pain 18.5 MBq/kg (0.5 mCi/kg) and 59 patients to 37 MBq/ was measured using validated patient-derived visual kg (1.0 mCi/kg). Treatment produced improvement analog scales and pain descriptor scales. At 3–4 weeks, from baseline in all patient-rated efficacy assessments, analgesic consumption significantly decreased in including degree of pain, level of daytime discomfort, patients assigned to active samarium compared with quality of sleep and pain relief. During the first 4 weeks those assigned placebo (P ϭ Ͻ 0.05). Furthermore, after dose administration, when the patients evaluated both the visual analog scale and the pain descriptor efficacy daily, there were statistically significant changes scale showed significant improvements in the treat- from baseline with the 1.0 mCi/kg dose but not with the ment group at 2–4 weeks, which correlated with the 0.5 mCi/kg dose. The difference between doses in visual reduction in analgesic use (P ϭ 0.0004). The absolute analog pain scores was statistically significant at week 4 risk reduction for complete responders at 4 weeks (P ϭ 0.0476). Values for platelets and white blood cells was 0.29 (95% CI 0.08–0.36). Because nonrespond- reached nadirs at 3 or 4 weeks with both doses and ers were unblinded at week 4, statistical comparisons recovered by 8 weeks. Even at their lowest point, the between the arms beyond week 4 were not possible. values were generally higher than those associated with Mild and transient myelosuppression was the only infectious or hemorrhagic complications [34]. clinically important toxic effect associated with active samarium [36]. Serafini et al. assessed the efficacy of two doses of samarium-153 versus placebo; 118 patients with The different radionuclides have been compared to painful metastases secondary to prostate, breast or each other and no significant differences have been lung cancers were randomly assigned to 18.5 MBq/kg found. Liepe et al. investigated samarium-153, stron- (0.5 mCi/kg), 37 MBq/kg (1.0 mCi/kg) or placebo. tium-83, rhenium-186 and rhenium-188 to deter- Pain intensity was assessed by patients and clini- mine their efficacy and toxicity in pain palliation of cians. An improvement in pain relief was reported bone metastases. Seventy nine patients (18 with breast for both doses compared with the placebo group, and 61 with prostate cancer) were treated (31 patients but the improvement was significantly greater in the with rhenium-188, 15 each with rhenium-186 and 37 MBq/kg group than in the 18.5 MBq/kg group samarium-153, and 18 with strontium-83). All at 4 weeks (67% versus 42%, absolute risk reduction patients were interviewed using standardized sets of 0.256; 95% CI 0.043–0.47). Persistence of pain relief questions before and after therapy weekly for 12 weeks. was seen through week 16 in 43% of patients receiv- Blood counts were taken weekly for 6 weeks and after ing the higher dose. Pain relief was observed within 12 weeks. In total, 73% of patients reported pain relief 1 week in the majority of patients who responded. (77% after rhenium-188, 67% after rhenium-186, 73% Analgesic consumption progressively rose in the pla- after samarium-153, and 72% after strontium-83). cebo group compared with the prestudy rate, whereas Fifteen percent of patients could discontinue their there was a reduction in both active treatment groups analgesics and were pain free. Pain showed a decrease from 3.6ϩ/Ϫ1.7 to a maximum of 2.2ϩ/Ϫ1.8 at 318

Oncologic therapy in cancer pain visual analog scale in 10 steps (P Ͻ 0.01). There were caused by radiotherapy-induced edema; prophylactic no significant differences in the inducement of flare corticosteroids are used. symptoms or marrow toxicity [37]. The side effects caused by large single fractions to Radiotherapy in the treatment of pain wide fields such as upper, mid or lower body include caused by soft tissue tumors nausea, vomiting, diarrhea, fever, transient increase Radiotherapy effectively relieves not only pain in bone pain, hematologic toxicity and, rarely, pneu- secondary to bone metastases, but also pain arising monitis [21]. Patients are usually admitted to hospital from soft tissue tumors. Among soft tissue tumors, for intravenous hydration and premedication con- palliative radiotherapy for inoperable lung cancer is sisting of antiemetics and corticosteroids. These side best documented. Approximately 60% (25–62%) of effects may be especially taxing for patients with poor patients with locally advanced inoperable primary performance status. An interval of at least 4 weeks lung cancer suffer from chest pain; almost one half is recommended before administering the other have moderate to severe pain [7]. The other typical half-body treatment or continuing chemotherapy to local symptoms of inoperable lung cancer are dyspnea avoid severe hematologic toxicity. Toxicity caused by in 80–90% (13–90%), cough in 80–90% (24–95%), half-body irradiation may be more acceptable with and hemoptysis in every third patient (13–47%). fractionated schedules [25]. Radiotherapy alleviates local symptoms in two-thirds (48–94%) of patients. Complete symptom relief is Patients close to death should be given adequate seen in every third patient. Hemoptysis is alleviated medication and good care; they are not candidates for in most of the patients (72–97%), chest pain in two- radiotherapy. thirds (44–88%), and dyspnea (46–72%) and cough (38–82%) approximately in every second patient. The Serious late side effects are rarely seen. This is due duration of symptom control lasts approximately 2–3 both to the short prognosis of patients and the low months [7, 38–44]. doses used in palliative care. Spinal cord injury is a dreaded late side effect. The risk of myelopathy is low Palliative radiotherapy is widely used also in the with conventional fractionation schedules 1.8–2 Gy alleviation of pain from other soft tissue tumors. given to a total dose of 45 Gy [45]. It is less clear how Whereas the clinical impression is that radiotherapy is safe higher single fractions are. According to Macbeth effective in reducing pain caused by soft tissue tumors et al., the cumulative risk of myelopathy was 0.6% at from various primaries, adequate studies are lacking. 1 year and 2.2% at 2 years following two 8.5 Gy frac- tions administered 1 week apart. A single 8 Gy fraction Side effects of external beam radiotherapy appears to be safe with regard to myelopathy [46]. and radio-isotopes Palliative radiotherapy for bone pain and pain-causing Systemic radionuclide therapy is usually well soft tissue tumors is relatively well tolerated. Side tolerated. Thrombocytopenia and neutropenia are the effects are related to total dose, fraction size and the most common toxic effects, but they are generally mild size of the irradiated field. Side effects arise both and reversible. For example, in a cohort of men with from adjacent healthy tissues and the tumor. In the prostate cancer treated with samarium-153, white cell latter case, a flare of previous symptoms may occur counts fell by about 45% and platelet counts fell by 40%, and dying tumor cells may affect the patient’s fluid with corresponding nadirs at 3–4 weeks of 3.8 ϫ 109/l and electrolyte balance and kidney function. Pelvic for the white cell count and 127 ϫ 109/l for the platelet irradiation may cause transient nausea and diarrhea count. Normal counts of white cells and platelets were resulting from bowel irritation and large pelvic fields observed by week 8 [36]. Hematologic toxicity should be may cause myelosuppression; treatment to the cervi- taken into consideration when treating extensively pre- cal or thoracic spine is associated with irritation of treated patients, and in patients requiring chemotherapy, the throat and esophagus. Side effects associated with because bone marrow capacity may be limited. treating nerve compression or brain metastases are increase of pain and associated neurologic symptoms Pain flares and complications of metastatic bone disease such as spinal cord compression have not been over-represented. In the work by Sartor mentioned above where active samarium-153 was compared with inactive samarium, pain flares were seen in 6% of patients in both 319

Chapter 24 groups. Spinal cord compression occurred at the same (chemotherapy or hormone therapy) reduces the risk frequency in both groups (5.9%) [36]. of developing skeletal events, i.e. new skeletal lesions, progression of existing bone lesions, hypercalcemia, Bisphosphonates pain, pathologic fractures, need for palliative radio- therapy or surgery, but no survival benefit has been Mechanism of action demonstrated. The strongest evidence for the effec- Bisphosphonates are potent inhibitors of osteoly- tiveness of bisphosphonates is found in breast cancer sis and bone resorption. They are synthetic analogs and multiple myeloma. There is some evidence of of natural pyrophosphates. In contrast to the natu- their efficacy, especially zoledronic acid, in treatment ral pyrophosphates, bisphosphonates are resistant to of advanced prostate cancer and other solid tumors. breakdown by enzymatic hydrolysis. Bisphosphonates act specifically on bone because of their strong affin- Bisphosphonates in treatment of pain ity for calcium phosphate. According to their structure, secondary to bone metastases bisphosphonates can be divided into those resembling The Cochrane review of bisphosphonates (clodro- the natural pyrophosphates, or nonaminobisphos- nate, pamidronate and etidronate) for the relief of phonates (clodronate and etidronate), and those with pain secondary to bone metastases includes 30 rand- a nitrogen-containing side chain, aminobisphospho- omized controlled trials reported before January 2000 nates (pamidronate, alendronate, zoledronic acid, rise- [50]. Of those, five studies were designed with pain dronate, ibandronate). Bisphosphonates inhibit bone relief as a primary endpoint and six studies were pla- resorption by a variety of mechanisms. When osteo- cebo controlled. Twenty-five studies were available to clasts ingest bisphosphonate-containing bone, their address the primary objective of the review, but only cytoskeleton becomes disrupted and the apoptosis of eight studies (771 patients from six placebo-controlled the osteoclasts is activated. The nonaminobisphospho- and two open-controlled studies) were included nates act as analogs of adenosine triphosphate (ATP) in the analyses of “the best pain response within and inhibit ATP-dependent intracellular enzymes, 12 weeks.” Number needed to treat (NNT) was 11 leading to apoptosis and death of the osteoclasts. The (95% CI 6–36) at week 4 and 7 (95% CI 5–12) at aminobisphosphonates, on the other hand, inhibit week 12. Odds ratio of 2.37 (95% CI 1.61–3.5) for enzymes of the mevalonate pathway by disrupting the “the best pain response within 12 weeks” was in favor signaling functions of key regulatory proteins and lead of the bisphosphonate group. In double-blinded stud- to osteoclast apoptosis. Bisphosphonates also inhibit ies the NNT was somewhat higher (NNT 8). Four the development of osteoclasts from monocyte pre- other placebo- or open-controlled studies including cursors by diminishing the recruitment and activity 619 patients reported no pain-relieving effect with of osteoclasts. These effects on the osteoclasts lead to bisphosphonates, but the data were not in usable form a decrease in bone turnover that is secondary to the for the review. The other 14 placebo- or open-controlled inhibition of bone resorption [47]. studies had no data on the proportion of patients with pain relief during the time frame of interest. Average The mechanism by which bisphosphonates alleviate drug reactions were generally mild. The number needed pain is not known. According to preclinical data, they to harm (NNH) was 16 (95% CI 12–27) for adverse inhibit bone resorption and reduce tumor burden at drug reactions requiring discontinuation of therapy. skeletal sites. Bisphosphonates may have some anti- inflammatory and antinociceptive effects [48, 49]. All Despite the large number of randomized control- these mechanisms might contribute to pain relief. led studies, these conclusions are based on a limited number of studies and patients due to methodologic Bisphosphonates in treatment problems in measuring and reporting pain. The review- of metastatic cancer ers concluded that the evidence suggests that bisphos- Bisphosphonates have successfully been used in the phonates provide modest pain relief for patients with treatment of malignant hypercalcemia and skeletal painful bony metastases [50]. metastases [50–53]. The use of bisphosphonates in addition to systemic treatment of malignant disease Rosen et al. compared zoledronic acid to pamid- ronate in patients with multiple myeloma and breast 320

Oncologic therapy in cancer pain cancer [54]. No difference was found between these no significant pain-relieving effect was seen, even two bisphosphonates in pain-relieving effect. though there was a trend towards it. The review included 1955 patients with skeletal metastatic pros- Another Cochrane review by Pavlakis et al. summa- tate cancer from 10 controlled trials published up to rizes the studies of bisphosphonates in breast cancer June 2005. In seven trials pain was the primary end- [51]. The review was updated in 2005. It included point. Clodronate was used in seven trials, pamid- 11 studies (seven placebo-controlled and three open ronate, etidronate and zoledronic acid in one study studies) that tested the effects of bisphosphonates each. Pain relief was demonstrated in 27.9% of the (compared with placebo or no bisphosphonate) on patients in the bisphosphonate group and in 21.1% pain in women with advanced breast cancer and exist- of the controls in 416 patients from three clodronate ing bone metastases, using a reference pain scale. In six and one etidronate studies. The odds ratio for pain studies (two placebo-controlled studies of IV pamidr- response was 1.54 (95% CI 0.97–2.44, P ϭ 0.07), onate 90 mg, one placebo-controlled study of IV iban- showing a trend of improved pain relief in the dronate 6 mg, one pooled study of oral ibandronate 50 bisphosphonate group, although this was not statis- mg, one oral clodronate 1600 mg, and one open study tically significant. Mean pain change was reported of oral pamidronate 300 mg), a significant difference in one zoledronic acid study (643 patients) and was seen in pain in favor of the bisphosphonate group. one pamidronate study (350 patients). The baseline One study compared zoledronic acid to pamidronate brief pain inventory score was 2.0 (standard devia- and found no differences between these two bisphos- tion (SD) 2.0) in the 4 mg zoledronic acid group, phonates in pain-relieving effect [54]. There were, 2.5 (SD 2.1) in the 8/4 mg group and 2.1 (SD 2.0) however, methodologic weaknesses in measuring and in the placebo group. The mean changes within 15 reporting the pain-relieving effects. months were ϩ0.58 (Ϫ0.29 to 0.87), ϩ0.43 (Ϫ0.16 to 0.70) and ϩ0.88 (Ϫ0.61 to 1.15), respectively. The The Cochrane review of multiple myeloma difference between zoledronic acid 4 mg and placebo included 11 trials published before June 2001 [52]. was not statistically significant (P ϭ 0.134), while Bisphosphonates statistically significantly ameliorated zoledronic acid 8/4 mg was statistically significantly pain in multiple myeloma. From eight eligible studies better than placebo (P ϭ 0.026). After combining the (four clodronate, four pamidronate, two etidronate and two zoledronic acid groups, the mean pain change one ibandronate studies) with a total of 1281 patients, did not alter during the zoledronic acid therapy there were 276 out of 657 patients who reported pain Ϫ0.04 (SD 0.21), while it increased in the placebo on bisphosphonates versus 318 out of 624 controls, 0.59 group ϩ0.42 (SD 0.22). No change was seen with (95% CI 0.46–0.76, P ϭ 0.00005). The absolute risk pamidronate: the mean pain change decreased by reduction was 9% and the NNT to prevent one patient Ϫ0.61 (SD 0.17) with pamidronate and Ϫ0.44 (SD experiencing pain was 11 (95% CI 7–28). However, 0.16) with placebo [53]. analysis of the effect of bisphosphonates on pain was based on clinically heterogeneous data and must be In a study of zoledronic acid in the treatment interpreted with caution. The benefit was most apparent of skeletal metastatic lung cancer and other solid with clodronate and pamidronate. The review included tumors in 773 patients, no significant pain-relieving only one ibandronate study, where ibandronate was effect was demonstrated, even though the propor- given at a dose of 2 mg intravenously, monthly, without tion of skeletal-related events (radiation, surgery, any significant pain-relieving effect. However, the dose pathologic fracture, spinal cord compression, hyper- of ibandronate was suboptimal when compared to the calcemia of malignancy) was significantly reduced by dose used in breast cancer (6 mg intravenously) [55]. zoledronic acid [56]. No studies of zoledronic acid were available. In two open-label pilot studies the analgesic effect of In the Rosen study zoledronic acid was compared ibandronate loading doses has been studied [57, 58]. to pamidronate in patients with multiple myeloma Twenty-five hormone-refractory prostate cancer patients and breast cancer, but myeloma patients have not participated in an open prospective nonrandomized been reported separately [54]. clinical study. Ibandronate was given 6 mg intravenously on 3 consecutive days at the study entry and 6 mg IV In the Cochrane review of bisphosphonates in the treatment of skeletal metastatic prostate cancer, 321

Chapter 24 every 4 weeks thereafter [57]. Significant reduction in NNH was 16 for adverse drug reactions requiring pain score measured by visual analog scale from 6.5 discontinuation of therapy [50]. Intravenous admin- (5–10) to 2.0 (0–4) was achieved in 23 (92%) patients; istration of the aminobisphosphonates can induce nine patients (39%) become completely pain free. Only transient flu-like symptoms with fever, myalgia and two patients did not respond. On average, the first anal- arthralgia. Oral administration of bisphosphonates gesic effect was observed at day 3 (1–5). In another in turn may be accompanied by gastrointestinal dis- study 18 patients were treated with IV ibandronate 4 mg comfort, nausea, dyspepsia, vomiting, diarrhea and on 4 consecutive days (total dose 16 mg) [58]. Within even esophageal ulceration [65]. Some deterioration 7 days there was a significant reduction in bone pain of renal function is reported in patients treated with that remained below baseline at 42 days. However, these bisphosphonates [54, 65, 66]. Forty two patients are preliminary results from open studies and no phase from a group of 446 who were treated with a total of III controlled randomized studies are available. 3115 zoledronic acid doses experienced renal deteri- oration [66]. Eight out of 446 patients required dis- Dosing and treatment schedules continuation of zoledronic acid therapy. Predictive In clinical phase III studies the onset of the analgesic factors for the development of renal deterioration effect of bisphosphonates has started within 4 weeks were age, diagnosis of myeloma or renal cell carci- and the maximal analgesic effect has been achieved noma, cumulative number of doses, concomitant within 8–12 weeks [59–62]. The analgesic effect has therapy with NSAID and current or prior therapy been demonstrated with IV pamidronate, oral and with cisplatin. Though the renal complications IV clodronate, oral and IV ibandronate, and IV zoled- during bisphosphonate therapy are often mild and ronic acid. The best evidence of the pain-relieving apparent only as transient increases in serum creati- effect of pamidronate has been reported with the nine, serum electrolyte and creatinine levels should dose of 90 mg intravenously every 4 weeks. In dose- be monitored during bisphosphonate therapy. The finding studies of pamidronate, lower doses from 45 probability of acute toxicity due to hypocalcemia is to 60 mg have also had some pain-relieving effect, greater with the use of IV aminobisphosphonates but 90 mg was the most effective dose [50, 51]. and supplementation with calcium and vitamin D is The maximum bone-resorbing effect is obtained recommended [65]. using oral clodronate at the dose level 1600–3200 mg. The optimal analgesic dose is not known [63]. Osteonecrosis of the jaw is a recently described The corresponding IV doses are either a single dose rare yet severe adverse side effect of bisphosphonates of 1500 mg or 300 mg on 5 consecutive days [50]. [67, 68]. Patients treated with IV nitrogen-containing Ibandronate 2 mg intravenously was not effec- bisphosphonates (pamidronate or zoledronic acid) tive in the treatment of multiple myeloma whereas represent 94% of published cases. Of interest, clodr- 6 mg IV infusion and 50 mg oral ibandronate sig- onate, a nonaminobisphosphonate, has not been nificantly alleviated pain in breast cancer [59]. Four implicated in the development of osteonecrosis. milligrams of zoledronic acid intravenously is used Most patients (85%) have myeloma or breast cancer. in clinical practice instead of 8 mg because of renal The cumulative incidence of osteonecrosis appears safety issues. There are only a few direct compari- to be 1% within the first year, increasing with the sons between different bisphosphonates. Zoledronic prolongation of treatment (3-year cumulative inci- acid has been demonstrated to be as effective as dence ranging between 4% and 21%), being higher pamidronate [54]. In a small study of 51 patients IV with zoledronic acid than pamidronate treatment. pamidronate 90 mg seemed to alleviate pain more The typical presentation of osteonecrosis is a “non- effectively than oral (1600 mg) or intravenous (1500 healing” extraction socket or exposed jawbone with mg) clodronate [64]. localized swelling and purulent discharge. The man- dible is affected twice as often as the maxilla and Safety of bisphosphonate treatment 60% of cases are preceded by a dental surgical pro- Bisphosphonate therapy is generally well tolerated. cedure. The remaining cases occurred spontaneously. According to the Cochrane review on pain relief, the Oversuppression of bone turnover is probably the primary mechanism with additional contributing 322

Oncologic therapy in cancer pain co-morbid factors like dental surgery or poor oral The latter two may also be administered with pro- hygiene. The management of this side effect is based phylactic intent. Single fraction radiotherapy is as on awareness of the problem and its prevention. effective as multiple fraction therapy in relieving Dental procedures such as extractions should be metastatic bone pain. Retreatment following single kept to a minimum during bisphosphonate treat- fraction treatment is needed somewhat more often; ment and when necessary should prompt temporary retreatment is equally effective. Toxicity caused by interruption of treatment until the lesion is healing wide-field radiation may be more acceptable with satisfactorily. fractionated schedules. The onset of pain relief fol- lowing external radiotherapy varies from a few days Conclusion to 4 weeks and the duration of pain relief is between 3 and 6 months. The response rate seen following Analgesics are the cornerstone of cancer pain treat- systemic radionuclide therapy is similar to that seen ment. Effective systemic oncologic treatment, when after external beam radiotherapy. Treatment is well available, alleviates cancer pain. Radiotherapy is the tolerated, with the exception of mild myelotoxic- single most effective oncologic treatment for cancer ity. Systemic radionuclide therapy seem to be more pain. Approximately 70% of patients with painful effective in treating osteosclerotic and mixed skeletal bone secondaries obtain significant pain relief and metastases. The onset of pain relief takes 2–4 weeks. a complete response is seen in every third patient. Radiotherapy is a useful tool in treating pain caused Oncologic treatments should be implemented early by soft tissue tumors, even though this is less well in the disease before irreversible tissue damage has documented. taken place. Local-field external radiotherapy is the recommended treatment of one to a few painful Pain relief provided by bisphosphonates is mod- bony metastases, while systemic radionuclide ther- est at most. The evidence is best documented for apy and wide-field radiotherapy are recommended myeloma and breast cancer, i.e. in osteolytic or mixed for patients with multiple painful skeletal metastases. metastases. The pain-relieving evidence for prostate and other solid tumors is insufficient (Table 24.1). Table 24.1 Bisphosphonates in treatment of pain caused by bony metastases Tumor type Reference Studies Results Evidence level All tumor types Wong & Wiffen [50] Cochrane review NNT 7 (95% CI 5–12) B Rosen et al. [54] RCT OR 2.37 (95% CI 1.61–3.5) No significant difference between pamidronate and zoledronic acid Multiple Djulbegovic et al. [52] Cochrane review NNT 11 (95% CI 7–28) B myeloma OR 0.59 (95% CI 0.46–0.76) Breast cancer Pavlakis et al. [51] Cochrane review Six out of 10 studies demonstrated pain B relieving effect of bisphosphonates. No significant difference between pamidronate and zoledronic acid Prostate cancer Yuen et al. [53] Cochrane review OR 1.54 (95% CI 0.97–2.44) B Small et al. [69] RCT No significant difference between Saad et.al. [70] RCT pamidronate and placebo, or zoledronic acid 4 mg and placebo Other solid Rosen et al. [56] RCT No significant difference between C tumors zoledronic acid and placebo OR, odds ratio; CI, confidence interval. 323

Chapter 24 Authors’ recommendations 9. Mercadante S. Malignant bone pain: pathophysiology and treatment. Pain 1997; 69: 1–18. • Analgesics are the mainstay in treating cancer pain. • Determine if effective systemic oncologic treatment 10. Fulfaro F, Casuccio A, Ticozzi C, Ripamonti C. The role of bisphosphonates in the treatment of painful meta- is available. static bone disease: a review of phase III trials. Pain 1998; • Radiotherapy is effective for metastatic bone pain 78(3):157–169. (A). External beam radiotherapy is very effec- 11. Poulsen HS, Nielsen OS, Klee M, Rorth M. Palliative irra- tive in relieving metastatic bone pain. The overall diation of bone metastases. Cancer Treatment Rev 1989; response rate is approximately 70–80% and the 16(1): 41–48. complete response rate is 30–60% (A). Single frac- tion radiotherapy is as effective as multiple fraction 12. Harrington KD. Impending pathologic fractures from radiotherapy in relieving metastatic bone pain (A). metastatic malignancy: evaluation and management. Retreatment is effective (B). Radiotherapy is effective Instructional Course Lectures 1986; 35: 357–381. in treating pain caused by soft tissue tumors (C/D). • Radio-isotopes are effective for metastatic bone 13. Garmatis CJ, Chu FCH. The effectiveness of radiation ther- pain (A). apy in the treatment of bone metastases from breast cancer. • Bisphosphonates provide at best modest pain relief Radiology 1978; 126: 235–237. for patients with painful bony metastases (B); this effect is best documented for multiple myeloma 14. Steenland E, Leer JW, van Houwelingen H, et al. The effect and breast cancer (B). of a single fraction compared to multiple fractions on pain- • There is insufficient evidence to support the pain- ful bone metastases: a global analysis of the Dutch Bone relieving effect of bisphosphonates in prostate Metastasis Study. Radiother Oncol 1999; 52(2): 101–109. cancer (B) 15. Yarnold JR on behalf of the Bone Pain Trial Working Party. References 8 Gy single fraction radiotherapy for the treatment of metastatic skeletal pain: randomised comparison with a 1. Hearn J, Higginsson IJ. Cancer pain epidemiology: a sys- multifraction schedule over 12 months of patient follow- tematic review. In: Bruera ED, Portenoy RK (eds) Cancer up. Radiother Oncol 1999; 52: 111–121. Pain, Assessment and Management. Cambridge University Press, Cambridge, 2003: 19–37. 16. McQuay HJ, Collins S, Carroll D, Moore RA. Radiotherapy for the palliation of painful bone metastases. Cochrane 2. Grond S, Zech D, Diefenbach C, Radbruch L, Lehmann KA. Database of Systematic Reviews 1999, Issue 3. Art. No.: Assessment of cancer pain: a prospective evaluation in 2266 CD001793. DOI: 10.1002/14651858.CD001793. cancer patients referred to a pain service. Pain 1996; 64(1): 107–114. 17. Hartsell WF, Scott CB, Bruner DW, et al. Randomized trial of short- versus long-course radiotherapy for palliation of pain- 3. Tubiana-Hulin M. Incidence, prevalence and distribution ful bone metastases. J Natl Cancer Inst 2005; 97: 798–804. of bone metastases. Bone 1991; 12(suppl1): S9–S10. 18. Jeremic B, Shibamoto Y, Igrutinovic I. Single 4 Gy re- 4. Coleman RE. Skeletal complications of malignancy. Cancer irradiation for painful bone metastasis following single 1997; 80 (suppl): 1588–1594. fraction radiotherapy. Radiother Oncol 1999; 52: 123–127. 5. Elte JWF, Bijvoet OLM, Cleton FJ, van Oosterom AT, 19. van der Linden YM, Lok JJ, Steenland E, et al. Dutch Bone Sleeboom HP. Osteolytic bone metastases in breast carci- Metastasis Study Group. Single fraction radiotherapy is noma pathogenesis, morbidity and bisphosphonate treat- efficacious: a further analysis of the Dutch Bone Metastasis ment. Eur J Clin Oncol 1996; 22: 493–500. Study controlling for the influence of re-treatment. Int J Radiat Oncol Biol Phys 2004; 59(2): 528–537. 6. Nielsen OS, Munro AJ, Tannock IF. Bone metastases: patho- physiology and management policy. J Clin Oncol 1991; 9(3): 20. Roos DE, Turner SL, O’Brien PC, et al. Randomized trial 509–524. of 8 Gy in 1 versus 20 Gy in 5 fractions of radiotherapy for neuropathic pain due to bone metastases (Trans-Tasman 7. Hoopwood P, Stephans RJ. Symptoms at presentation for Radiation Oncology Group, TROG 96.05). Radiother Oncol treatment in patients with lung cancer: implications for 2005; 75: 54–63. the evaluation of palliative treatment. Br J Cancer 1995; 71: 633–636. 21. Salazar OM, Rubin P, Hendrickson FR, et al. Single-dose half-body irradiation for palliation of multiple bone 8. Singh SM, Longmire WP Jr, Reber HA. Surgical palliation metastases from solid tumours. Cancer 1986; 58: 29–36. for pancreatic cancer. The UCLA experience. Ann Surg 1990; 212: 132–139. 22. Poulter CA, Cosmatos DH, Rubin P, et al. A report of RTOG 8206: a phase III study of whether the addition of single dose half-body irradiation to standard fraction- ated local field irradiation is more effective than local field irradiation alone in the treatment of symptomatic osseous metastases. Int J Radiat Oncol Biol Phys 1992; 23: 207–214. 23. Quilty PM, Kirk D, Bolger JJ, et al. A comparison of the palliative effects of strontium-89 and external beam radio- therapy in metastatic prostate cancer. Radiother Oncol 1994; 31(1): 33–40. 324

Oncologic therapy in cancer pain 24. Salazar OM, DaMotta NW, Bridgman SM, Cardiges NM, 37. Liepe K, Kotzerke J. A comparative study of 188Re-HEDP, Slawson RG. Fractionated half-body irradiation for pain 186Re-HEDP, 153Sm-EDTMP and 89Sr in the treatment of palliation in widely metastatic cancers: comparison with painful skeletal metastases. Nucl Med Commun 2007 ;28(8): single dose. Int J Radiat Oncol Biol Phys 1996; 36(1): 49–60. 623–630. 25. Salazar OM, Sandhu T, da Motta NW, et al. Fractionated 38. Simpson JR, Francis ME, Perez-Tamayo R, Marks RD, half-body irradiation (HBI) for the rapid palliation of Rao DV. Palliative radiotherapy for inoperable carcinoma of widespread, symptomatic, metastatic bone disease: a rand- the lung: final report of a RTOG multi-institutional trial. Int omized Phase III trial of the International Atomic Energy J Radiat Oncol Biol Phys 1985; 11(4): 751–758. Agency (IAEA). Int J Radiat Oncol Biol Phys 2001; 50(3): 765–775. 39. Teo P, Tai TH, Choy D, Tsui KH. A randomized study on palliative radiation therapy for inoperable non small cell 26. Finlay IG, Mason MD, Shelley M. Radio-isotopes for the carcinoma of the lung. Int J Radiat Oncol Biol Phys 1988; palliation of metastatic bone cancer: a systematic review. 14(5): 867–871. Lancet Oncol 2005; 6: 392–400. 40. Anonymous. Inoperable non-small-cell lung cancer 27. Buchali K, Correns HJ, Schuerer M, Schnorr D, Lips H, (NSCLC): a Medical Research Council randomised trial of Sydow K Results of a double blind study of 89-strontium palliative radiotherapy with two fractions or ten fractions. therapy of skeletal metastases of prostatic carcinoma. Eur Report to the Medical Research Council by its Lung Cancer J Nucl Med 1988; 14: 349–351. Working Party. Br J Cancer 1991; 63(2): 265–270. 28. Lewington VJ, McEwan AJ, Ackery DM, et al. A prospective, 41. Anonymous. A Medical Research Council (MRC) ran- randomised double-blind crossover study to examine the domised trial of palliative radiotherapy with two fractions efficacy of strontium-89 in pain palliation in patients with or a single fraction in patients with inoperable non-small- advanced prostate cancer metastatic to bone. Eur J Cancer cell lung cancer (NSCLC) and poor performance status. 1991; 27: 954–958. Medical Research Council Lung Cancer Working Party. Br J Cancer 1992; 65(6): 934–941. 29. Porter AT, McEwan JB, Powe JE, et al. Results of a rand- omized phase-III trial to evaluate the efficacy of strontium- 42. Macbeth FR, Bolger JJ, Hopwood P, et al. Randomized trial of 89 adjuvant to local field external beam irradiation in the palliative two-fraction versus more intensive 13-fraction radio- management of endocrine resistant metastatic prostate therapy for patients with inoperable non-small cell lung cancer cancer. Int J Radiat Oncol Biol Phys 1993; 25: 805–813. and good performance status. Medical Research Council Lung Cancer Working Party. Clin Oncol 1996; 8(3): 167–175. 30. Smeland S, Erikstein B, Aas M, Skovlund E, Hess SL, Fosså SD. Role of strontium-89 as adjuvant to palliative 43. Abratt RP, Shepherd LJ, Salton DG. Palliative radiation for stage external beam radiotherapy is questionable: results of a 3 non-small cell lung cancer – a prospective study of two mod- double-blind randomized study. Int J Radiat Oncol Biol erately high dose regimens. Lung Cancer 1995; 13(2): 137–143. Phys 2003; 56(5): 1397–1404. 44. Rees GJ, Devrell CE, Barley VL, Newman HF. Palliative 31. Oosterhof GO, Roberts JT, de Reijke TM, et al. radiotherapy for lung cancer: two versus five fractions. Clin Strontium(89) chloride versus palliative local field radio- Oncol 1997; 9(2): 90–95. therapy in patients with hormonal escaped prostate cancer: a phase III study of the European Organisation for Research 45. Schultheiss TE, Stephens LC. Permanent radiation myelop- and Treatment of Cancer, Genitourinary Group. Eur Urol athy. Br J Radiol 1992; 65: 737–753. 2003; 44: 519–526. 46. Macbeth FR, Wheldon TE, Girling DJ, et al. Estimates of 32. Dearnaley DP, Bayly RJ, A’Hern RP, Gadd J, Zivanovic MM, risk in 1048 patients in three randomized trails of pallia- Lewington VJ. Palliation of bone metastases in prostate tive radiotherapy for non-small-cell lung cancer. Clin Oncol cancer. Half-body irradiation or strontium-89? Clin Oncol 1996; 8:176–181. 1992; 4(2): 101–107. 47. Russell RG, Rogers MJ. Bisphosphonates: from the labora- 33. Sciuto R, Festa A, Rea S, et al. Effects of low-dose cisplatin on tory to the clinic and back again. Bone 1999; 25(1): 97–106. 89Sr therapy for painful bone metastases from prostate can- cer: a randomized clinical trial. J Nucl Med 2002; 43: 79–86. 48. Bonabello A, Galmozzi MR, Bruzzese T, Zara GP. Analgesic effect of bisphosphonates in mice. Pain 2001; 91(3): 34. Resche I, Chatal JF, Pecking A, et al. A dose-controlled study 269–275. of 153Sm-ethylenediaminetetramethylenephosphonate (EDTMP) in the treatment of patients with painful bone 49. Walker K, Medhurst SJ, Kidd BL, et al. Disease modify- metastases. Eur J Cancer 1997; 33: 1583–1591. ing and anti-nociceptive effects of the bisphosphonate, zoledronic acid in a model of bone cancer pain. Pain 2002; 35. Serafini AN, Houston SJ, Resche I, et al. Palliation of pain 100(3): 219–229. associated with metastatic bone cancer using samarium- 153 lexidronam: a double-blind placebo-controlled clinical 50. Wong RKS, Wiffen PJ. Bisphosphonates for the relief of trial. J Clin Oncol 1998; 16(4): 1574–1581. pain secondary to bone metastases. Cochrane Database of Systematic Reviews 2002, Issue 2. Art. No.: CD002068. 36. Sartor O, Reid RH, Hoskin PJ, et al. Samarium-153- DOI: 10.1002/14651858.CD002068. Lexidronam complex for treatment of painful bone metas- tases in hormone-refractory prostate cancer. Urology 2004; 51. Pavlakis N, Schmidt RL, Stockler MR. Bisphosphonates for 63: 940–945. breast cancer. Cochrane Database of Systematic Reviews 2005, Issue 3. Art. No.: CD003474. DOI: 10.1002/14651858. CD003474.pub2. 325

Chapter 24 52. Djulbegovic B, Wheatley K, Ross H, et al. Bisphosphonates 62. McCloskey EV, MacLennan IC, Drayson MT, Chapman C, in multiple myeloma. Cochrane Database of Systematic Dunn J, Kanis JA. A randomized trial of the effect of clodr- Reviews 2002, Issue 4. Art. No.: CD003188. DOI: onate on skeletal morbidity in multiple myeloma. MRC 10.1002/14651858.CD003188. Working Party on Leukaemia in Adults. Br J Haematol 1998; 100(2): 317–325. 53. Yuen KY, Shelley M, Sze WM, Wilt T, Mason M. Bisphosphonates for advanced prostate cancer. Cochrane 63. O’Rourke N, McCloskey E, Houghton F, Huss H, Kanis JA. Database of Systematic Reviews 2006, Issue 4. Art. No.: Double-blind, placebo-controlled, dose-response trial of CD006250. DOI: 10.1002/14651858.CD006250. oral clodronate in patients with bone metastases. J Clin Oncol 1995; 13(4): 929–934. 54. Rosen LS, Gordon D, Kaminski M. Zoledronic acid ver- sus pamidronate in the treatment of skeletal metastases in 64. Jagdev SP, Purohit P, Heatley S, Herling C, Coleman RE. patients with breast cancer or osteolytic lesions of multi- Comparison of the effects of intravenous pamidronate ple myeloma: a phase III, double-blind, comparative trial. and oral clodronate on symptoms and bone resorption Cancer J 2001; 7(5): 377–387. in patients with metastatic bone disease. Ann Oncol 2001; 12(10): 1433–1438. 55. Body JJ, Diel I, Lichinitser MR. Intravenous ibandronate reduces the incidence of skeletal complications in patients 65. Conte P, Guarneri V. Safety of intravenous and oral bisphos- with breast cancer an bone metastases. Ann Oncol 2003; phonates and compliance with dosing regimens. Oncologist 14(9): 1399–1405. 2004; 9(suppl 4):28–37. 56. Rosen LS, Gordon D, Tchekmedyian S, et al. Zoledronic 66. McDermott RS, Kloth DD, Wang H, Hudes GR, Langer CJ. acid versus placebo in the treatment of skeletal metastases Impact of zoledronic acid on renal function in patients with in patients with lung cancer and other solid tumours: a cancer: clinical significance and development of a predictive phase III, double-blind, randomized trial – the Zoledronic model. J Support Oncol 2006; 4(10): 524–529. Acid Lung Cancer and Other Solid Tumours Study Group. J Clin Oncol 2003; 21(16): 3150–3157. 67. Woo SB, Hellstein JW, Kalmar JR. Narrative [corrected] review: bisphosphonates and osteonecrosis of the jaws. Ann 57. Heidenreich A, Elert A, Hofmann R. Ibandronate in the Intern Med 2006; 144(10): 753–761. treatment of prostate cancer associated painful osseous metastases. Prostate Cancer Prostatic Dis 2002; 5(3): 231–235. 68. Migliorati CA, Siegel MA, Elting LS. Bisphosphonate- associated osteonecrosis: a long-term complication of 58. Mancini I, Dumon JC, Body JJ. Efficacy and safety of iban- bisphosphonate treatment. Lancet Oncol 2006; 7: 508–514. dronate in the treatment of opioid-resistant bone pain associated with metastatic bone disease: a pilot study. J Clin 69. Small EJ, Smith MR, Seaman JJ, Petrone S, Kowalski MO. Oncol 2004; 22(17): 3587–3592. Combined analysis of two multicenter, randomized, placebo-controlled studies of pamidronate disodium for 59. Cameron D, Fallon M, Diel I. Ibandronate: its role in meta- the palliation of bone pain in men with metastatic pros- static breast cancer. Oncologist 2006; 11(suppl 1): 27–33. tate cancer. J Clin Oncol 2003 Dec 1; 21(23): 4277–84. Epub 2003 Oct 27. 60. Kohno N, Aogi K, Minami H, et al. Zoledronic acid sig- nificantly reduces skeletal complications compared with 70. Saad F, Gleason DM, Murray R, Tchekmedyian S, Venner placebo in Japanese women with bone metastases from P, Lacombe L, Chin JL, Vinholes JJ, Goas JA, Chen B; breast cancer: a randomized, placebo-controlled trial. J Clin Zoledronic Acid Prostate Cancer Study Group. A random- Oncol 2005; 23(15): 3314–3321. ized, placebo-controlled trial of zoledronic acid in patients with hormone-refractory metastatic prostate carcinoma. 61. Hortobagyi GN, Theriault RL, Lipton A. Long-term preven- J Natl Cancer Inst 2002 Oct 2; 94(19): 1458–68. tion of skeletal complications of metastatic breast cancer with pamidronate. J Clin Oncol 1998; 16: 2038–2044. 326

CHAP TE R 25 Cancer pain: analgesics and co-analgesics Rae Frances Bell Pain Clinic/Regional Centre of Excellence in Palliative Care, Haukeland University Hospital, Bergen, Norway Background as tricyclic antidepressants (e.g. amitriptyline) or anticonvulsants (e.g. gabapentin or pregabalin). Pathophysiology Refractory neuropathic pain requires other measures, Pain due to malignancy may be both acute and such as adjuvant treatment with an NMDA receptor chronic. Cancer patients commonly experience sev- antagonist or anesthesiologic techniques such as spi- eral types of pain concurrently. In the majority of nally administered local anesthetic as an adjuvant to cases, one or more types of pain are caused by the opioid. cancer [1]. Tumor expansion can cause pressure on surrounding organs, while tumor infiltration into Intermittent or breakthrough pain nerve plexi and bone, and damage of nerve tissue can Breakthrough pain is common in cancer patients, cause neuropathic pain. Metastatic spread of cancer with bone pain, local tumor invasion in soft tissue, to bone is reported to be one of the most common and brachial plexopathy most frequently reported causes of cancer pain [2]and may cause pain both at [6]. Breakthrough pain usually occurs at the site of rest and on movement. Cancer patients may expe- the background pain and the duration may vary from rience muscular pain due to rapid weight loss and minutes to hours [7]. Intense, short-lasting pain epi- other factors. They are potentially subject to painful sodes and movement-related pain are particularly adverse effects of treatment, such as joint pain fol- difficult to treat effectively with analgesics. Normal- lowing chemotherapy, painful mucositis, and acute release oral opioid or oral transmucosal fentanyl and/or persistent neuropathic pain following radio- citrate have until recently been the most common or chemotherapy. Cancer patients are often exposed pharmacologic treatment options for breakthrough to surgical interventions and experience acute and pain. Different formulations/routes of administra- in some cases chronic postoperative pain [3]. In an tion of opioids such as the fentanyl buccal tablet [8] international survey addressing cancer pain char- or adjuvants such as intranasal ketamine [9] are cur- acteristics, 71.6% of cancer patients with pain were rently under investigation. judged to have nociceptive pain, 34.7% nociceptive visceral pain and 39.7% had neuropathic pain [1]. The potential complexity of the cancer patient’s pain syndrome (see Table 25.1) underscores the Neuropathic pain importance of repeated clinical assessment and pain Neuropathic pain is difficult to treat with opio- diagnosis, together with an individual treatment plan. ids alone and usually requires adjuvant drugs such Prevalence, epidemiology and risk factors Evidence-Based Chronic Pain Management. Edited by Pain is common in patients with cancer. A literature C. Stannard, E. Kalso and J. Ballantyne. © 2010 Blackwell review in patients with lung cancer found that pain Publishing. affected 27% of outpatients (range 8–85%) and 76% 327

Chapter 25 Table 25.1 Cancer pain. Modified from Bell et al. [27] Examples of pain subtypes Possible pain mechanisms Tumor related Sensitization of peripheral nociceptive primary afferents (tumor factors, e.g. enthothelin and prostaglandins, tumor-induced acidosis, inflammation-associated factors); Metastatic soft tissue pain entrapment and nerve injury; invasion of mechanically sensitive tissues (e.g. visceral pain); Metastatic bone pain central sensitization Inflammatory (e.g. mucositis) Peripheral sensitization due to inflammation. Hyperalgesia due to central sensitization Neuropathy Tumor-induced acidosis; injury or infiltration of sensory neurones that innervate the bone Muscle pain marrow; peripheral sensitization of nociceptors [4]; central sensitization [5] osteolysis, Acute postoperative pain pathologic fracture, microfractures Chronic postoperative pain Peripheral sensitization due to inflammation. Hyperalgesia due to central sensitization Nervous tissue compression or lesion leading to central sensitization. Degeneration of sensory neurones and sensitization of primary nociceptive afferents due to chemotherapeutic agents and release of cytokines [4] Tumor factors; central sensitization; bone metastases causing muscle spasm; muscle hypercatabolism; immobilization leading to muscle atrophy; increased muscular tension Acute nociception; peripheral sensitization; nerve damage; (central sensitization) Central sensitization; nerve damage (peripheral sensitization) of patients cared for by palliative care services (range the addition of weaker opioids, and strong pain with 63–88%) [10]. Pain was caused by cancer in 73% the substitution of stronger for weaker opioids. The (range 44–87%) and by treatment in 11% (range utility of the second step on the ladder has been chal- 5–17%). The most common type of pain was noci- lenged, with suggestions to replace step two opioids ceptive, while neuropathic pain accounted for 30%. with stronger opioids. Morphine is the ”gold stand- A study using data from 1021 patients with advanced ard” opioid for cancer pain. lung cancer who were enrolled in three randomized trials of chemotherapy found that 74% of patients Although these guidelines are recommended and reported some pain, with 50% of these stating that commonly followed, there is restricted evidence for this pain was affecting daily activities [11]. practice. A systematic review of studies evaluating the effectiveness of the ladder for cancer pain management Interventions supported by included eight uncontrolled case series studies [13]. The evidence (ranked according conclusion was that the evidence did not permit estima- to evidence level) tion of the effectiveness of the ladder, and the need for randomized controlled efficacy trials was emphasized. Very few interventions in cancer pain are well docu- One prospective open-label 10-year study has assessed mented. Even the guidelines for cancer pain treat- the course of treatment of 2118 cancer pain patients ment lack an evidence base. The current evidence for being treated according to WHO guidelines [14]. treatment guidelines and commonly used interven- tions are discussed below. NSAIDs and paracetamol A Cochrane qualitative systematic review published in World Health Organization pain ladder 2003, with the latest Medline search, concluded that The World Health Organization (WHO) three- NSAIDs are more effective than placebo for the short- step ladder for cancer pain relief [12] advises that term treatment of cancer pain [15]. Forty-two trials mild cancer pain should be treated with nonopioid with 3084 patients were included. Seven of eight trials analgesics (paracetamol and/or nonsteroidal anti- that compared NSAIDs with placebo demonstrated inflammatory drugs (NSAIDs)), moderate pain with that NSAIDs had superior efficacy with no difference in adverse effects. Clear evidence to support superior 328

Analgesics and co-analgesics safety or efficacy of one NSAID was lacking. Only one to placebo for 7±1 days, a total of 4 patients received trial has compared paracetamol to placebo [16]. In placebo treatment [21]. In the majority of the tri- this 6-hour cross-over study in 29 evaluable patients, als included in this review morphine was used as an there was no significant difference in effect between active comparator. The assumption for using mor- paracetamol and placebo. A randomized, placebo- phine in this way is that it has previously been found controlled, cross-over trial in patients with advanced effective in cancer pain compared to placebo. But is cancer already receiving a strong opioid regimen this the case? found that pain and overall well-being were better for patients receiving paracetamol than for those receiv- A Cochrane review concluded that oral morphine ing placebo [17]. is effective for cancer pain [22]. This review attempted to bring all the literature together and included data It is important to note that the majority of stud- from randomized trials, including open-label tri- ies on NSAIDs and paracetamol were of less than als. The authors remarked that the majority of trials 7 days’ duration, so that the conclusions cannot be are equivalency studies designed to show that differ- generalized to long-term treatment of cancer pain. ent formulations of morphine have the same effect, There do not appear to be any long-term studies of and that this makes it difficult to extract information NSAIDs/paracetamol for cancer pain. It is important on the effectiveness of morphine per se. Furthermore, to remember that cancer patients are often at high they underlined that it is unclear whether the trials risk of NSAID-related adverse effects. In general, for are sufficiently powered to detect a clinically mean- safety reasons, for daily treatment lasting more than ingful difference between treatments. Although the 1–2 weeks, paracetamol is to be preferred. If NSAIDs Oxford Quality Scale scores were generally high, are used for long-term treatment then the lowest with a median of 4, it was noted that the quality of effective dose should be used, and co-treatment with reporting was disappointing, especially in regard to a gastroprotective agent (for example, omeprazole) is assessment of pain and pain relief. The trials in this recommended. Cochrane review were not scored for validity, and the relevance of a placebo control for the demonstration Opioids of efficacy is not specifically addressed. Morphine has been shown to be superior to placebo in patients with noncancer-related neuropathic pain In a second Cochrane review [23], 11 trials investi- (postherpetic neuralgia (PHN), painful diabetic neu- gating hydromorphone in cancer pain were included, ropathy (PDN) and phantom limb pain), while oxy- all of which used an active comparator. A recent codone has been investigated in relation to PHN and quantitative systematic review on oxycodone for PDN and found to have a similar number needed to cancer pain [24] found no placebo-controlled trials. treat (NNT) to morphine: 2.5 (95% confidence inter- A Cochrane review on methadone for cancer pain val (CI) 1.9–4.1) [18]. identified five randomized, double-blind trials all of which used an active comparator [25]. An Agency for However, although oral morphine is the “gold Healthcare Research and Quality (AHRQ) Evidence standard” opioid for the treatment of cancer pain, it Report [26] looking at the relative efficacy of anal- is surprising how few placebo-controlled trials have gesics in cancer pain described the need for placebo been performed. In a methodologic review of trials controls in order to avoid overestimation of treat- investigating oral opioids for cancer pain, only one ment effects, at the same time noting that placebo trial had a placebo control, while another trial had a controls in cancer pain trials are “rare.” Table 25.2 is placebo control in the pilot phase of the study [19]. a summary of placebo-controlled trials investigating In the first of these trials which investigated the effect stronger opioids for cancer pain. of a loading dose of morphine elixir added to the first dose of slow-release morphine tablets, a total of Unless there is an as yet unaccessed body of data, 9 patients were treated with a single dose of placebo, these findings raise interesting questions regarding the study duration being 12 hours [20]. In the pilot the current efficacy data for oral opioids in cancer phase of the second trial, where three different for- pain. Even though opioids appear clinically effective mulations of slow-release morphine were compared for cancer pain, the question of how effective is not resolved by the literature. Morphine is accepted as the 329

Chapter 25 Table 25.2 Randomized, double-blind trials in cancer pain patients comparing stronger opioid with placebo. Modified from Bell et al. [27] Study nϭ Drug Route Duration Comments Houde et al. [28] 67 (placebo: 28) Morphine IM 6 hours Double-blind Stambaugh et al. 20 po Up to 6 hours Single dose [16] Butorphanol 30 (acetaminophen) IM 4 days, 1 treatment Single dose Stambaugh et al. (B ϩ A) per day [29] 60 (3 groups) IM Meperidine 7 days Double-blind Stambaugh et al. 43 (40 evaluable) Hydroxyzine 6 hours Single and multiple [30] MϩH 12 hours dose study 20 (19 evaluable) 9 Stambaugh et al. treated with placebo Dezocine Double-blind [31] 172 (152 final day Butorphanol Single dose efficacy data) Cross-over Hoskin et al. [20] 4 treated with placebo Ciramadol po 92 (89 assessable, Codeine Double-blind Broomhead et al. treated with at least [21] one unit of OTFC and Morphine po one unit of placebo) Farrar et al. Morphine po First phase (placebo Double-blind [32] control): 7 Ϯ 1 days) Placebo control only Fentanyl citrate OTM in first phase of study (OTFC) Titration period Ϯ 10 randomly ordered Double-blind treatment units Breakthrough (Pain evaluated for pain 60 minute period) IM, intramuscular; po, oral; OTM, oral transmucosal. Sources: AHRQ [26], Wiffen et al. [22], Quigley [23], Reid [24] and Bell [19]. In addition, searches were performed on PubMed with limits “randomized controlled trial” and search terms “fentanyl AND placebo and cancer” / “methadone AND placebo AND cancer.” gold standard for cancer pain treatment but placebo- Antidepressants for neuropathic pain, controlled efficacy data in cancer pain are lacking. with special reference to cancer pain A Cochrane review on antidepressants for neuropathic Opioids for breakthrough pain pain with final search in December 2004 found that A Cochrane review published in 2006 concluded that antidepressants are effective for the treatment of neu- there is evidence that oral transmucosal fentanyl cit- ropathic pain, with the best evidence being for tricyclic rate (OTFC) is an effective treatment in the manage- antidepressants, with amitriptyline having a NNT of 2 ment of breakthrough pain [33]. Four studies with 393 (CI 1.7–2.5) [35]. Two small placebo-controlled stud- participants were included. Only one trial compared ies on postoperative neuropathic pain after breast can- OTFC to placebo. The literature was small and no cer treatment were included in this review. In the first trials for other opioids were found. One randomized study, a randomized cross-over trial which enrolled double-blind, placebo-controlled trial published the 20 patients, amitriptyline 100 mg provided significant same year investigated fentanyl buccal tablet (FBT) pain relief; however, adverse effects such as tiredness (100–800 µg) for breakthrough pain in cancer patients compromised treatment and caused four patients to [34]. FBT is an effervescent formulation which has discontinue the trial [36]. In the second study which enhanced rate and extent of uptake of fentanyl com- was also a cross-over study and included 15 patients, pared to OTFC. This trial of short duration found the venlafaxine 75 mg gave some pain relief and no sig- fentanyl buccal tablet to be safe and efficacious. nificance in adverse effects compared to placebo [37]. 330

Analgesics and co-analgesics A second systematic review on drug treatment of treatment of neuropathic cancer pain, there are as yet neuropathic pain included 105 studies [18]. Cancer no published trials on this treatment. pain studies are not specifically described in this review. In peripheral neuropathic pain, the lowest Finnerup et al. performed a systematic review on NNT was for tricyclic antidepressants, while there randomized double-blind placebo-controlled trials were limited data for central neuropathic pain. of treatment for neuropathic pain [18]. This review found that gabapentin has documented moderate Two recent randomized placebo-controlled trials effect on pain and quality of life measures includ- have reported duloxetine to be effective and well toler- ing mood and sleep disturbance in mixed neuro- ated as a treatment for PDN [38,39]. In the first of these pathic pain states, PHN, PDN and spinal cord injury. trials, 344 of 457 randomized patients completed the Pregabalin in PHN and PDN was found to have a 12-week study period. Duloxetine 60 and 120 mg/d were combined NNT of 4.2 (CI 3.4–5.4) for doses from found to improve pain measures, with a NNT of 4.1 (CI 150 mg to 600 mg, comparable to the effect of gabap- 2.9–7.2), and to improve daily functioning as measured entin. A subsequent large, randomized study found by the Brief Pain Inventory and quality of life measures. pregabalin to be effective in PHN and PDN with a Patients with psychiatric disorders, including general- NNT of 3.8 (CI 2.6–7.3) [44]. ized anxiety disorder and/or major depressive disorder, and patients with mixed pain disorders were excluded NMDA receptor antagonists for from this trial. In the second trial which had similar neuropathic pain conclusions, 248 of 334 randomized patients completed The oral NMDA receptor antagonists memantine, the 12-week study period. Since tricyclics are generally dextromethorphan and riluzole have been studied poorly tolerated by elderly patients, mainly due to anti- mainly in small trials in neuropathic pain with lit- cholinergic adverse effects, duloxetine may be a more tle or no effect [18]. Randomized placebo-controlled suitable drug for the treatment of neuropathic pain in trials in cancer pain are lacking. this patient group. As yet, trials on duloxetine for neuro- pathic pain in the elderly and in cancer pain are lacking. Ketamine as an adjuvant to opioids for cancer pain Anticonvulsants for neuropathic pain, It is apparent from the literature that this is a com- with special reference to cancer pain mon treatment for refractory cancer pain. However, A Cochrane review from 2005 found that gabapentin trials are lacking. A Cochrane review published in is effective for PHN and PDN, but does not appear 2003 found insufficient evidence to support the use to be superior to carbamazepine, which is a cheaper of ketamine as an adjuvant to opioids for cancer pain option [40]. A Cochrane review on gabapentin for [45]. Only two small placebo-controlled trials with a acute and chronic pain concludes that there is evi- total of 30 patients could be included. One study with dence that gabapentin is effective in neuropathic pain a duration of 3 hours examined the effect of an intra- and that approximately two-thirds of patients who venous bolus of ketamine [46]. The second trial was take either carbamazepine or gabapentin can expect of longer but unspecified duration, and investigated to achieve good pain relief [41]. Fifteen trials (14 pub- the co-administration of intrathecal ketamine with lished reports, one reporting on two trials) with 1468 intrathecal morphine [47]. This review was updated participants were included. Only one study concerned in 2007 and did not identify additional randomized, cancer-related neuropathic pain [42]. This was a controlled trials. 10-day randomized, double-blind, placebo-controlled trial in 121 patients where gabapentin was added Systemic local anesthetics for to existing analgesic treatment. The trial concluded neuropathic pain, with special that gabapentin is effective in improving analgesia in reference to cancer pain patients with neuropathic pain already treated with Two systematic reviews have addressed the effect of sys- opioids, However, the trial was of short duration and temic local anesthetics in chronic pain. The first review, the magnitude of effect limited compared to placebo published in 1998, included 17 trials, of which three [43]. Although pregabalin is currently used for the concerned cancer pain [48]. Two small cross-over trials 331