Hets 2021

604 N. Maggio et al. / Experimental Neurology 247 (2013) 595–604 Maggio, N., Shavit, E., Chapman, J., Segal, M., 2008. Thrombin induces long-term poten- Shikamoto, Y., Morita, T., 1999. Expression of factor X in both the rat brain and cells of tiation of reactivity to afferent stimulation and facilitates epileptic seizures in rat the central nervous system. FEBS Lett. 463, 387–389. hippocampal slices: toward understanding the functional consequences of cere- brovascular insults. J. Neurosci. 28, 732–736. Siller-Matula, J.M., Schwameis, M., Blann, A., Mannhalter, C., Jilma, B., 2011. Thrombin as a multi-functional enzyme. Focus on in vitro and in vivo effects. Thromb. Haemost. Maggio, N., Cavaliere, C., Papa, M., Blatt, I., Chapman, J., Segal, M., 2013. Thrombin 106, 1020–1033. regulation of synaptic transmission: implications for seizure onset. Neurobiol. Dis. 50, 171–178 (Feb). Sokolova, E., Reiser, G., 2008. Prothrombin/thrombin and the thrombin receptors PAR-1 and PAR-4 in the brain: localization, expression and participation in neurodegenera- Mannaioni, G., Orr, A.G., Hamill, C.E., Yuan, H., Pedone, K.H., McCoy, K.L., Berlinguer tive diseases. Thromb. Haemost. 100, 576–581. Palmini, R., Junge, C.E., Lee, C.J., Yepes, M., Hepler, J.R., Traynelis, S.F., 2008. Plas- min potentiates synaptic N-methyl-D-aspartate receptor function in hippocam- Striggow, F., Riek, M., Breder, J., Henrich-Noack, P., Reymann, K.G., Reiser, G., 2000. The pal neurons through activation of protease-activated receptor-1. J. Biol. Chem. protease thrombin is an endogenous mediator of hippocampal neuroprotection 283, 20600–20611. against ischemia at low concentrations but causes degeneration at high concentra- tions. Proc. Natl. Acad. Sci. U. S. A. 97, 2264–2269. McCoy, K.L., Gyoneva, S., Vellano, C.P., Smrcka, A.V., Traynelis, S.F., Hepler, J.R., 2012. Protease-activated receptor 1 (PAR1) coupling to G(q/11) but not to G(i/o) or Turgeon, V.L., Salman, N., Houenou, L.J., 2000. Thrombin: a neuronal cell modulator. G(12/13) is mediated by discrete amino acids within the receptor second intracel- Thromb. Res. 99, 417–427. lular loop. Cell. Signal. 24, 1351–1360. Wang, J., Jin, H., Hua, Y., Keep, R.F., Xi, G., 2012a. Role of protease-activated receptor-1 Mosnier, L.O., Zlokovic, B.V., Griffin, J.H., 2007. The cytoprotective protein C pathway. in brain injury after experimental global cerebral ischemia. Stroke 43, 2476–2482. Blood 109, 3161–3172. Wang, Y., Zhang, Z., Chow, N., Davis, T.P., Griffin, J.H., Chopp, M., Zlokovic, B.V., 2012b. Rezaie, A.R., 2005. The occupancy of endothelial protein C receptor by its ligand mod- An activated protein C analog with reduced anticoagulant activity extends the ulates the par-1 dependent signaling specificity of coagulation proteases. IUBMB therapeutic window of tissue plasminogen activator for ischemic stroke in rodents. Life 63, 390–396. Stroke 43, 2444–2449. Riewald, M., Ruf, W., 2005. Protease-activated receptor-1 signaling by activated protein C Yang, Y., Rosenberg, G.A., 2011. Blood–brain barrier breakdown in acute and chronic in cytokine-perturbed endothelial cells is distinct from thrombin signaling. J. Biol. cerebrovascular disease. Stroke 42, 3323–3328. Chem. 280, 19808–19814. Zlokovic, B.V., Griffin, J.H., 2011. Cytoprotective protein C pathways and implications for stroke and neurological disorders. Trends Neurosci. 34, 198–209. 99



‫‪Late diagnosis of fetal central nervous system anomalies‬‬ ‫‪following a normal second trimester anatomy scan‬‬ ‫‪Prenatal Diagnosis | 2013‬‬ ‫ההשתתפות בפרוייקט ח\"ץ אפשרה לי לחנוך ולעבוד עם סטודנט מוכשר‬ ‫מנחה‪ :‬פרופ' יואב ינון‬ ‫במיוחד בשם אלעד בן מאיר שבינתיים הפך לרופא מוכשר במחלקת ילדים‬ ‫אולטרסאונד גינקולוגי‬ ‫בשיבא‪.‬‬ ‫‪[email protected]‬‬ ‫העבודה עם אלעד במסגרת הפרוייקט היתה מפרה לשני הכיוונים והובילה‬ ‫לפרסומם של שני מאמרים חשובים בנושא של אנגיוגנזיס בהיריון‪.‬‬ ‫עבורי ואני בטוח שגם עבור אלעד ההשתתפות בפרוייקט היתה חוויה‬ ‫מעצימה‪.‬‬ ‫אלעד בן מאיר‬ ‫מנחה‪ :‬פרופ' אלדד קטורזה‬ ‫אונ' תל אביב‬ ‫דימות העובר‪ ,‬יילוד וגינקולוגיה‬ ‫השתתף כסטודנט בפרויקט ח“ץ‬ ‫מנהל פרויקט ח\"ץ‬ ‫מנהל מכון גרטנר‬ ‫בין השנים ‪2011-2012‬‬ ‫‪[email protected]‬‬ ‫‪[email protected]‬‬ ‫‪101‬‬

DOI: 10.1002/pd.4163 ORIGINAL ARTICLE Late diagnosis of fetal central nervous system anomalies following a normal second trimester anatomy scan Y. Yinon1*†, E. Katorza1†, D. I. Nassie1, E. Ben-Meir1, L. Gindes1, C. Hoffmann2, S. Lipitz1, R. Achiron1 and B. Weisz1 1Department of Obstetrics and Gynecology, Sheba Medical Center, Tel-Aviv University, Israel 2Department of Radiology, Sheba Medical Center, Tel-Hashomer, Tel-Aviv University, Israel *Corresponding to: Yoav Yinon. E-mail: [email protected] †These authors contributed equally to this work. ABSTRACT Objective The aim of this study was to describe the nature of central nervous system (CNS) anomalies diagnosed during the third trimester following a normal anatomy scan at 21–24 weeks of gestation. Methods Retrospective cohort study of all pregnant women referred to the fetal medicine unit at Sheba Medical Center between 2005 and 2011 due to fetal CNS anomalies detected at the late second and third trimesters following a normal anatomy scan at 21–24 weeks of gestation. Results During the study period, 47 patients were diagnosed with fetal CNS anomalies at a median gestational age of 31.1 weeks (range 24–38). The four most common anomalies found included intracranial cysts (19%), mild ventriculomegaly (15%), absence or dysgenesis of the corpus callosum (10%), and intracerebral hemorrhage (10%). Other CNS anomalies detected in this group of patients included hydrocephalus, Dandy walker malformation, large cysterna magna, microcephalus with lissencephaly, craniosynestosis, periventricular pseudocysts, global brain ischemia, cerebellar hypoplasia, and sub-ependymal nodule. Conclusions Fetal brain continues to evolve throughout gestation, and therefore, some of the CNS anomalies can be diagnosed only during late second and third trimesters of pregnancy. Consequently, in patients who have a third trimester scan for any reason, assessment of the fetal CNS should be considered. © 2013 John Wiley & Sons, Ltd. Funding sources: None Conflicts of interest: None declared INTRODUCTION In Israel, many women undergo an early anatomy scan at 14– Malformations of the central nervous system (CNS) comprise a 16 weeks of gestation, which is partly covered by the health major cause of developmental delay and neurological deficits. insurance funds followed by a late anatomy scan at 22 weeks of The prevalence of congenital anomalies of the CNS varies gestation, which is fully covered, whereas other chooses to do only among different studies, some of which suggest an incidence one anatomy scan at 22 weeks of gestation. of about 3–5 per 1000 births.1,2 We present in this study a cohort of patient referred to our Antenatal sonography has proven to be a useful tool in the center due to fetal CNS anomalies and aimed to describe the evaluation of the fetal brain and spine enabling prenatal diagnosis nature of CNS anomalies diagnosed at the late second and of many of the CNS anomalies.3 However, compared with other third trimesters following a previous normal anatomy scan at organs, fetal brain continues to evolve throughout gestation, 21–24 weeks of gestation. mainly the gray and white matter, and the development of certain structures is gestational age dependant.4 Consequently, some of METHODS the CNS anomalies can be diagnosed only during late second All pregnant women referred to the fetal medicine unit at and third trimesters of pregnancy.4 Late diagnosis of severe CNS Sheba Medical Center during a 7-year period between January anomalies may lead to third trimester termination of pregnancy, 2005 and December 2011 due to suspected fetal CNS which involves many ethical, moral, and religious issues as well anomalies were reviewed. This is a case-series study focusing as psychological burden for the parents but is currently acceptable on patients diagnosed with CNS anomalies at the late second in a number of western countries.5–7 However, even in countries and third trimesters following a normal anatomy scan at 21– where late termination is illegal, antenatal diagnosis of CNS 24 weeks. Our unit serves as a tertiary referral center evaluating anomalies is crucial and may improve the neonatal outcome as around 500 pregnancies per year for fetal abnormalities. All some cases may require delivery at a tertiary center. Prenatal Diagnosis 2013, 33, 929–934 © 2013 John Wiley & Sons, Ltd. 102

930 Y. Yinon et al. patients underwent a thorough work-up, which consisted of a Table 1 Pregnancy characteristics of patients with normal second detailed anatomy scan, a dedicated neurosonography trimester scan who were later found to have central nervous (neurosonogram) and amniocentesis as indicated. Many of system abnormalities (n = 47)a the patients underwent fetal brain magnetic imaging resonance (MRI) in order to confirm the sonographic diagnosis Maternal age 28 (21–42) and to obtain additional information. Once a fetal CNS In-vitro fertilization 6 (13) anomaly was diagnosed, the patients were counseled by a Twins 3 (6.5) multidisciplinary team including a pediatric neurologist, Abnormal nuchal translucency 1(2) genetic counselor, and a neurosurgeon when required. Gestational age at diagnosis (weeks) Following prenatal diagnosis and counseling and depending Fetal CMV infection 31.1 (24–38) on the severity of findings, some patients elected to have Abnormal karyotype 1 (2) termination of pregnancy and underwent the procedure Non-CNS anomalies 0/28 following approval of the ethics committee. Maternal records Termination of pregnancy 8 (17) and sonographic data of all patients with fetal CNS anomalies, including the actual films when available, were reviewed. Data 20 (42) regarding nuchal translucency, early and late anatomy scans performed during pregnancy, nature of CNS and non-CNS CMV, cytomegalovirus; CNS, central nervous system. anomalies, gestational age at diagnosis, fetal karyotype, and aData are expressed as median (range) or as number (percentage). the presence of fetal infection were collected and entered into our database. Table 2 Central nervous system anomalies (n = 47)a The study was approved by the local institutional ethical Diagnosis N (%) committee. Data are presented as median (range), and frequencies are expressed as percentages. Intracranial cysts 9 (19) RESULTS Arachnoid cysts 6 During the study period, 840 patients were referred to our unit due to suspected CNS anomalies. Forty seven of these patients Cavum vergae cysts 2 were diagnosed with CNS anomalies during the late second or third trimester following a normal anatomy scan at 21– Choroid plexus cyst 1 24 weeks and consist of our study cohort. The fetal CNS anomalies in this cohort were diagnosed at a median Mild ventriculomegalyb 7 (15) gestational age of 31.1 weeks (range 24–38), 20 (42%) of them were diagnosed after 32 weeks of gestation. In 45 patients, the Absence or dysgenesis of corpus callosum 5 (10) initial detection of a CNS anomaly was achieved incidentally on a routine scan performed for fetal growth assessment or Intra-cerebral hemorrhage 5 (10) biophysical score. In one case, a neurosonogram and MRI were performed at 27 weeks of gestation following complaints of Periventricular pseudocysts 4 (8.5) reduced fetal movements revealing global brain ischemia, and one patient was diagnosed with sub-ependymal nodule Hydrocephalus 3 (6) following an incidental detection of cardiac rhabdomyomas. In 39 patients (83%), fetal MRI was performed and confirmed Dandy walker malformation/vermian agenesis 3 (6) the sonographic findings. The pregnancy characteristics of these patients are described in Table 1. Eight (17%) of the Large cysterna magna/Blake pouch 3 (6) patients had other non-CNS anomalies including renal anomalies (3), cardiac abnormalities (3), one patient with an Microcephalus with lissencephaly 3 (6) abdominal cyst, and one patient with single umbilical artery and polyhydramnios. Twenty eight patients underwent Craniosynostosis 2 (4) amniocentesis, and the fetal karyotype was found to be normal in all of them, and in only one case, the CNS findings were due Global brain ischemia 1 (2) to fetal cytomegalovirus infection. After counseling, 20 (42%) women elected to have termination of pregnancy due to the Cerebellar hypoplasia 1 (2) severity of the findings. Sub-ependymal nodule 1 (2) The CNS anomalies diagnosed in this group of women are detailed in Table 2. The four most common anomalies found aData are expressed as number (percentage). included intracranial cysts (19%), mild ventriculomegaly bFive cases were unilateral, two were bilateral. (defined as 10–15 mm), (15%), absence or dysgenesis (malformation) of the corpus callosum (10%), and Mild ventriculomegaly – atrial width of 10–15 mm. intracerebral hemorrhage (10%). All nine cases of intracranial Hydrocephalus – atrial width >15-mm Large cysterna magna – >10 mm on transcerebellar axial plane. cysts were detected incidentally on a routine scan and included arachnoid cysts in six cases (Figure 1) of which four were supra-tentorial and two were at the posterior fossa, two cases of cavum Vergae cysts, and one case of choroid plexus cyst. The five cases of agenesis of the corpus callosum included four cases of complete agenesis of the corpus callosum and one case of dysgenesis of the corpus callosum. All of them were isolated with no other CNS pathology except ventriculomegaly. Five of the patients in our study were diagnosed with intra- cerebral hemorrhage of which four had evidence of intra- ventricular bleeding (Figure 2) with ventricular dilatation confirmed by fetal MRI, and in one patient, the bleeding was located in the head of the caudate nucleus. Fetal intra-cerebral Prenatal Diagnosis 2013, 33, 929–934 © 2013 John Wiley & Sons, Ltd. 103

Late diagnosis of fetal CNS anomalies 931 Figure 1 Large unilateral tempro-parieto-occipital arachnoid cyst diagnosed at 32 weeks of gestation. (a) para-sagittal ultrasonography plane. (b) coronal T2 MR image hemorrhage was detected at 24 weeks of gestation in one case DISCUSSION and at 31–34 weeks of gestation in the other four cases. Other Unlike most fetal organs, the fetal brain continues to develop CNS anomalies that were detected in the late second or third throughout gestation. Therefore, some of the CNS anomalies trimester following a normal anatomy scan at 21–24 weeks may appear late in gestation, thus remaining undiagnosed by included periventricular pseudocysts (8.5%), hydrocephalus the routine anatomy scan performed at 18–22 weeks of gestation. (6%), Dandy walker malformation (6%), large cysterna magna In this paper, we describe a cohort of patients referred to our (defined as above 10 mm on transcerebellar axial plane), center due to CNS anomalies aiming to characterize the nature (6%), microcephaly with lissencephaly (6%) (Figure 3), of anomalies that were diagnosed late in gestation after 24 weeks. craniosynostosis (4%), global brain ischemia (2%) cerebellar In our cohort, 47 women were diagnosed with CNS anomalies hypoplasia (2%), and sub-ependymal nodule (2%). The patient after 24 weeks of gestation, following a previous normal anatomy with microcephaly was diagnosed at 33 weeks of gestation scan at 21–24 weeks of gestation. The most common anomalies following detection of small head circumference measuring diagnosed in this group included intracranial cysts, mild three standard deviations below the mean for gestational age. ventriculomegaly, absence or dysgenesis of the corpus callosum, The patient with sub-ependymal nodule was diagnosed following and intracerebral hemorrhage. Hydrocephalus, Dandy Walker prenatal detection of multiple cardiac rhabdomyomas at 33 weeks malformation, large cysterna magna, microcephalus with of gestation (Figure 4). An earlier scan at 31 weeks of gestation was lissencephaly, craniosynostosis, and periventricular pseudocysts otherwise normal except a thick inter-ventricular septum. The were also detected late in gestation following a previous normal combination of multiple rhabdomyomas and sub-ependymal anatomy scan. This group of patients, who were diagnosed nodule in the fetal brain indicated that this fetus was most likely late in gestation after a previous normal scan, represents a affected by tuberous sclerosis. major challenge for prenatal counseling and management, as Figure 2 Intra-ventricular hemorrhage at 32 weeks of gestation with involvement of the brain parenchyma (IVH grade 4). (a)Axial ultrasonography plane demonstrating enlarged lateral ventricles with hyperechogenic borders and disruption of the normal brain structure, (b) T2 MR axial plane at the same level. The bleeding is well depicted outside the borders of the lateral ventricle with destruction of the surrounding brain tissue (white arrow) Prenatal Diagnosis 2013, 33, 929–934 © 2013 John Wiley & Sons, Ltd. 104

932 Y. Yinon et al. Figure 3 Microcephaly with abnormal cortical development at 34 weeks of gestation. (a) Para-sagittal T2 MR view of the fetal brain demonstrating significant delay of cortex maturation with increased fluid at the sub-arachnoid space and decreased biometry. (b) T2 MR coronal plane image confirming the diagnosis. (c) Axial ultrasonographic view demonstrating abnormal sulcation and operculization (white arrows).Although sulci and insular operculization are visible, they are dilated for the gestational age Figure 4 Ultrasonography of a fetus at 34 weeks of gestation. (a) multiple cardiac rhabdomyomas (white arrows). (b) Echogenic nodule located near the foramen of Monroe (white arrow), suggesting subependymal nodule as commonly seen in tuberous sclerosis complex Prenatal Diagnosis 2013, 33, 929–934 © 2013 John Wiley & Sons, Ltd. 105

Late diagnosis of fetal CNS anomalies 933 some of the anomalies are severe and are associated with major ventriculomegaly, corpus callosum anomalies, and intracranial handicap in the future. Therefore, some of the patients will cysts.16 Migration disorders, vermian dysgenesis, macrocephaly, elect to have termination of pregnancy. However, termination brain calcifications, and microcephaly were also found in their of pregnancy beyond viability raises many ethical questions, cohort of patients. and in many countries, it is not allowed6,8. It is important to know whether these anomalies, which were detected late in It is possible that some of the anomalies that were diagnosed the second or third trimester, could have been detected late in gestation in our cohort were missed during the second earlier during the second trimester anatomy scan or whether trimester ultrasound examination. Filly et al. have described 112 their late detection is inevitable due to the ongoing development fetuses with brain anomalies and found that most of the of the fetal brain throughout gestation. Seven of our patients prenatally diagnosed CNS anomalies could be suspected by were diagnosed with mild ventriculomegaly at the late second examining the atrium of the lateral ventricle and the cisterna or third trimester following a normal anatomy scan at 22 weeks magna.17 However, there are several CNS abnormalities that of gestation. In most cases, isolated mild ventriculomegaly is do not manifest until the third trimester such as acquired associated with favorable outcomes. However, in a minority lesions secondary to stroke, hemorrhage and infections and of cases, it could be the earliest manifestation of brain damage developmental anomalies which become apparent only after the due to primary cerebral maldevelopment or due to destructive end of the proliferation and migration period such as lesions secondary to hypoxia or infection.2 A review of 234 cases lissencephaly, heterotopia, macrocephaly, and microcephaly.16 with mild ventriculomegaly, of which most were detected Whenever a CNS anomaly is detected by ultrasound during the early in pregnancy, revealed chromosomal aberrations in 4% of third trimester, fetal MRI has an important role as adjunctive tool cases and neurologic sequelae in 11% of cases.9 Although the to sonography for the additional information given to parents and optimal management of isolated mild ventriculomegaly is for the possibility to change the diagnosis, counseling, and uncertain, especially if initially detected during the third trimester, management of pregnancy.18 fetal karyotyping, aCGH, and a TORCH screen are usually recommended. Furthermore, ventriculomegaly might indicate Our study was not designed to determine the incidence of fetal more severe cerebral lesions that are more apparent on MRI than brain anomalies diagnosed late in gestation following a previous ultrasound, and therefore, fetal MRI should be considered normal second trimester scan. Moreover, as a referral center, our following antenatal detection of ventriculomegaly.10,11 population is obviously biased and do not represent the general population. Nevertheless, during a 7-year period, 47 referred Absence/dysgenesis of the corpus callosun and brain cysts patients were diagnosed with brain abnormalities after 24 weeks were also relatively common in our cohort of patients. The corpus of gestation following a previous normal second trimester scan, callosum completes its formation between 18 and 20 weeks of resulting in termination of pregnancy in 42% of the cases due to gestation. Consequently, prenatal diagnosis of complete agenesis the severity of the findings. of the corpus callosum is possible at 22 weeks of gestation by mid-coronal and mid-sagittal scans. Nevertheless, agenesis of Our study shows that CNS evaluation during a third trimester the corpus callosum may be partial in which the dorsal portion ultrasound exam may contribute to the diagnosis of fetal brain is missing to varying degrees. Although complete agenesis is anomalies that were not seen or not present during the second regarded as a malformation, partial agenesis, also referred to as trimester scan. Moreover, some developmental CNS anomalies dysgenesis of the corpus callosum, may represent a disruptive are not apparent during the routine anomaly scan (21– event that can occur at any time during pregnancy .12 Bennett 24 weeks) and can be visualized only in later stages of gestation. et al. reported on 15 fetuses with callosal agenesis confirmed by A third trimester scan can also detect other non-CNS anomalies, third trimester scans, 10 of whom were reported to have a which appear late in gestation, including skeletal dysplasias, completely normal ultrasound scan at 16–22 weeks.13 Therefore, cardiac abnormalities (cardiomyopathies, arrhythmias, tumors), prenatal diagnosis of partial agenesis of the corpus callosum and gastrointestinal malformations such as esophageal atresia.19 may only become detectable in late gestation. In Israel, for example, termination of pregnancy is legal until term if the fetus carries a high risk of disability and pending approval of Similarly, intracranial cysts may not develop until late in a special ethics committee. This has led to a growing demand for gestation. In a series of 54 cases of fetal intracranial cyst, none were accurate diagnosis and prognostication even if achieved only in discovered before 20 weeks of gestation, whereas 55% were the third trimester. In countries where late termination is not an detected between 20 and 30 weeks and 45% were detected after option, late prenatal diagnosis of fetal brain pathologies is the 30th week. All of the cases detected after 30 weeks had a important as well because it might have impact on mode and normal ultrasound at 22 weeks of gestation, indicating that many location of delivery as certain pathologies require delivery at a of these lesions actually develop late in pregnancy.14,15 Postnatally, tertiary center where facilities such as level 3 neonatal intensive the majority of these lesions, if not associated with other care unit and pediatric neurosurgery are available. anomalies, are benign in nature, remain clinically silent or even frequently regress spontaneously.15 In accordance with our data, Therefore, the CNS abnormalities described in our study, Malinger et al. have reported on 34 fetuses with intracranial which were only detected during the third trimester, should pathologies diagnosed following a normal second trimester raise awareness of such lesions and their further management. anatomy scan. The anomalies in this series were diagnosed We agree with others, that if scans are performed in the third at 29 weeks of gestation (range 22–37 weeks) and similar to our trimester for any reason, it may be useful to perform study, the three most common pathologies consisted of anatomical evaluation of structures where late appearing abnormalities are more common including the CNS, heart, genitourinary system, and limb length.18–20 Prenatal Diagnosis 2013, 33, 929–934 © 2013 John Wiley & Sons, Ltd. 106

934 Y. Yinon et al. WHAT’S ALREADY KNOWN ABOUT THIS TOPIC? WHAT DOES THIS STUDY ADD? • Fetal brain continues to evolve throughout gestation and therefore • CNS evaluation during a third trimester ultrasound exam may some of the CNS anomalies can be diagnosed only during the late contribute to the diagnosis of fetal brain anomalies that were not second and third trimester of pregnancy. seen or not present during the second trimester scan. REFERENCES 11. Levine D MR imaging of fetal central nervous system abnormalities. Brain Cogn 2002;50(3):432–48. 1. Garne E, Loane M, Dolk H, et al. Prenatal diagnosis of severe structural congenital malformations in Europe. Ultrasound Obstet Gynecol 12. Volpe P, Paladini D, Resta M, et al. Characteristics, associations and 2005;25(1):6–11. outcome of partial agenesis of the corpus callosum in the fetus. Ultrasound Obstet Gynecol 2006;27(5):509–16. 2. Grandjean H, Larroque D, Levi S. The performance of routine ultrasonographic screening of pregnancies in the Eurofetus Study. Am J 13. Bennett GL, Bromley B, Benacerraf BR. Agenesis of the corpus callosum: Obstet Gynecol 1999;181(2):446–54. prenatal detection usually is not possible before 22 weeks of gestation. Radiology 1996;199(2):447–50. 3. Hertzberg BS, Kliewer MA, Bowie JD. Sonographic evaluation of fetal CNS: technical and interpretive pitfalls. Ajr 1999;172(2):523–7. 14. Pierre-Kahn A, Hanlo P, Sonigo P, et al. The contribution of prenatal diagnosis to the understanding of malformative intracranial cysts: state of 4. Monteagudo A, Timor-Tritsch IE. Normal sonographic development of the the art. Childs Nerv Syst 2000;16(10–11):619–26. central nervous system from the second trimester onwards using 2D, 3D and transvaginal sonography. Prenat Diagn 2009;29(4):326–39. 15. Pierre-Kahn A, Sonigo P. Malformative intracranial cysts: diagnosis and outcome. Childs Nerv Syst 2003;19(7–8):477–83. 5. Korenromp MJ, Page-Christiaens GC, van den Bout J, et al. A prospective study on parental coping 4 months after termination of pregnancy for fetal 16. Malinger G, Lerman-Sagie T, Watemberg N, et al. A normal second-trimester anomalies. Prenat Diagn 2007;27(8):709–16. ultrasound does not exclude intracranial structural pathology. Ultrasound Obstet Gynecol 2002;20(1):51–6. 6. Boyd PA, Devigan C, Khoshnood B, et al. EUROCAT Working Group. Survey of prenatal screening policies in Europe for structural malformations and 17. Filly RA, Cardoza JD, Goldstein RB, Barkovich AJ. Detection of fetal central chromosome anomalies, and their impact on detection and termination rates nervous system anomalies: a practical level of effort for a routine sonogram. for neural tube defects and Down’s syndrome. BJOG 2008;115(6): 689–96. Radiology 1989;172(2):403–8. 7. Dommergues M, Benachi A, Benifla JL, et al. The reasons for termination of 18. Manganaro L, Savelli S, Francioso M, et al. Role of fetal MRI in the diagnosis pregnancy in the third trimester. Br J Obstet Gynaecol 1999;106(4):297–303. of cerebral ventriculomegaly assessed by ultrasonography. Radiol Med 2009;114(7):1013–23. 8. Barel O, Vaknin Z, Smorgick N, et al. Fetal abnormalities leading to third trimester abortion: nine-year experience from a single medical center. 19. Verrotti C, Caforio E, Gramellini D, Nardelli GB. Ultrasound Prenat Diagn 2009;29(3):223–8. screening in second and third trimester of pregnancy: an update. Acta Biomed 2007;78(3):229–32. 9. Pilu G, Falco P, Gabrielli S, et al. The clinical significance of fetal isolated cerebral borderline ventriculomegaly: report of 31 cases and review of the 20. Malinger G, Lev D, Lerman-Sagie T. Normal and abnormal fetal brain literature. Ultrasound Obstet Gynecol 1999;14(5):320–6. development during the third trimester as demonstrated by neurosonography. Eur J Radiol 2006;57(2):226–32. 10. Garel C. Imaging the fetus: when does MRI really help? Pediatr Radiol 2008; 38 Suppl 3:S467–70. Prenatal Diagnosis 2013, 33, 929–934 © 2013 John Wiley & Sons, Ltd. 107



Effective crizotinib schedule for brain metastases in ALK rearrangement metastatic non-small-cell lung cancer Case Reports | 2013 ‫ פרופ׳ ניר פלד‬:‫מנחה‬ ‫אונקולוגיה‬ [email protected] ‫אורי לירן‬ ‫אונ' תל אביב‬ ‫השתתף כסטודנט בפרויקט ח״ץ‬ 2012-2015 ‫בין השנים‬ [email protected] 109

CASE REPORT Effective Crizotinib Schedule for Brain Metastases in ALK  Rearrangement Metastatic Non–Small-Cell Lung Cancer Nir Peled, MD, PhD,* Leor Zach, MD,† Ori Liran, MSc,* Maya Ilouze, PhD,* Paul A. Bunn Jr., MD,‡ and Fred R. Hirsch, MD, PhD‡ The echinoderm microtubule–associated protein-like 4-ana- tomography computed tomography showed no evidence of plastic lymphoma kinase (EML4-ALK) gene fusion occurs active disease elsewhere. in 2% to 7% of non–small-cell lung cancer cases.1 Tumors expressing this fusion respond to treatment with crizotinib, DISCUSSION an ALK tyrosine kinase inhibitor. However, brain metastases Brain metastases are the Achilles’ heel of non–small-cell frequently occur, even in the presence of systemic response lung cancer with EML4-ALK translocation.2 However, there to therapy.2 is growing data supporting a therapeutic benefit of crizotinib also in cases of brain metastases.4 In this case report, the brain CASE PRESENTATION metastases were developed under crizotinib therapy and have A 45-year-old never-smoker man who was previously further shown resistance to WBRT, whereas it responded well reported in this journal,3 is presented here again for a dramatic to a different treatment schedule of crizotinib. Interestingly response to brain metastases. Previously, he was presented enough, Kaneda et al. have also shown a response to crizotinib for having abnormality in RNA editing for ALK gene and a therapy (standard dose) after a resistance to radiation.4 Other complete response to crizotinib 250 mg twice daily therapy. reports did not show a response to standard dose crizotinib He was diagnosed as harboring EML4-ALK rearrangement specifically after WBRT.5 Therefore we hypothesize that the through abnormal ALK immunohistochemistry staining and drug penetration might not increase sufficiently on standard breakpoints in ALK intron 19, detected by next-generation crizotinib schedule after WBRT and that increasing in the sequence. ALK fluorescence in situ hybridization result was Cmax, without increasing the daily dose of 500 mg/day could falsely negative. be beneficial. A previous report demonstrated brain response After a follow-up of 20 months, the patient developed after dose escalation up to 1000 mg/day, however, producing asymptomatic milliary spread of brain metastases without any asymptomatic bradycardia of 39 beats per minute.5 other evidence for active disease and he was treated by whole- It is important to emphasize that the response might brain radiotherapy (WBRT; 30 Gy). Crizotinib was stopped be a late benefit of WBRT and/or related to the increased for the radiation period and readministered thereafter in the previous standard schedule of 250 mg twice daily. A follow- up magnetic resonance imaging 3 months later showed lack of response and new brain lesions. Aiming to increase crizotinib brain penetration, we rescheduled crizotinib therapy to a single-day administration of 500 mg every day (morning). Brain magnetic resonance imaging in 2 months showed a dramatic response and elimi- nation of 90% of the brain lesions (Fig. 1). Positron emission *The Thoracic Cancer Research and Detection Center, †The Radiation FIGURE 1.  Brain MRIs (axial t1 with contrast images) in  Institute, Sheba Medical Center, Tel Aviv University, Tel-Aviv, Israel; and February, May, and July 2013. Brain metastases were diag- ‡University of Colorado Cancer Center, Division of Medical Oncology, nosed after 20 months of crizotinib therapy (250 mg twice  University of Colorado Denver, Aurora, Colorado. daily). Whole-brain radiotherapy (3000 cGy/10fr) was  administered while crizotinib was continued in standard dose  Disclosure: Dr. Hirsch is a (compensated) consultant (on the advisory board) (except for the radiation period). Follow-up MRI (May 2013)  for Pfizer and has a research agreement through University of Colorado showed lack of response. Then, crizotinib was rescheduled  with Ventana/Roche. The remaining authors declare no conflict of interest. for 500 mg X1/day, with a dramatic response 2 months later.  MRI, magnetic resonance imaging. Address for correspondence: Nir Peled, MD, PhD, FCCP, Head, The Thoracic Cancer Research and Detection Center, Sheba Medical Center, Tel Aviv University, Ramat-Gan 52621, Israel. E-mail: [email protected] Copyright © 2013 by the International Association for the Study of Lung Cancer ISSN: 1556-0864/13/0812-e112 e112 Journal of Thoracic Oncology®  •  Volume 8, Number 12, December 2013 110

Journal of Thoracic Oncology®  •  Volume 8, Number 12, December 2013CrizotinibScheduleforBrainMetastasesinALKRearrangementMetastaticNSCLC permeability of the blood–brain barrier to crizotinib second- REFERENCES ary to the WBRT. Further studies measuring crizotinib level in the cerebrospinal fluid are required to investigate this effect. 1. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N Engl J Med In summary, this case report suggests increasing crizo- 2010;363:1693–1703. tinib schedule to a single administration of the standard daily dose (500 mg/day) for brain metastases may be an option 2. Chun SG, Choe KS, Iyengar P, Yordy JS, Timmerman RD. Isolated cen- when they have developed under crizotinib therapy, after con- tral nervous system progression on Crizotinib: an Achilles heel of non- ventional radiotherapy. This approach would warrant further small cell lung cancer with EML4-ALK translocation? Cancer Biol Ther evaluation, potentially delaying the need for second-genera- 2012;13:1376–1383. tion inhibitors or other new therapies. 3. Peled N, Palmer G, Hirsch FR, et al. Next-generation sequencing iden- ACKNOWLEDGMENT tifies and immunohistochemistry confirms a novel crizotinib-sensitive The authors thank Dr. Shani Shilo for her assistance ALK rearrangement in a patient with metastatic non-small-cell lung can- with this case report. cer. J Thorac Oncol 2012;7:e14–e16. 4. Kaneda H, Okamoto I, Nakagawa K. Rapid response of brain metastasis to crizotinib in a patient with ALK rearrangement-positive non-small-cell lung cancer. J Thorac Oncol 2013;8:e32–e33. 5. Kim YH, Ozasa H, Nagai H, et al. High-dose crizotinib for brain metastases refractory to standard-dose crizotinib. J Thorac Oncol 2013;8:e85–e86. Copyright © 2013 by the International Association for the Study of Lung Cancer e113 111



‫‪Interleukin-12p40 in the spinal fluid as a biomarker‬‬ ‫‪for clinically isolated syndrome‬‬ ‫‪Multiple Sclerosis Journal | 2014‬‬ ‫מנחה‪ :‬דר' מיכאל גורביץ‬ ‫מנחה‪ :‬פרופ' ענת אחירון‬ ‫מנהל המעבדה הנוירואימונולוגית‪,‬‬ ‫מייסדת פרויקט ח\"ץ‪ ,‬מנהלת המרכז‬ ‫המרכז לטרשת נפוצה‬ ‫לטרשת נפוצה ואחראית הקתדרה‬ ‫למחלות אוטואימוניות אוניברסיטת ת\"א‬ ‫‪[email protected]‬‬ ‫‪[email protected]‬‬ ‫רותם אורבך‬ ‫אונ' תל אביב‬ ‫השתתפה כסטודנטית בפרויקט ח״ץ‬ ‫בין השנים ‪2007-2009‬‬ ‫‪[email protected]‬‬ ‫‪113‬‬

491166 MSJ0010.1177/1352458513491166Multiple Sclerosis JournalOrbach et al. 2013 Research Paper MULTIPLE SCLEROSIS MSJ Interleukin-12p40 in the spinal fluid as a JOURNAL biomarker for clinically isolated syndrome Multiple Sclerosis Journal 0(0) 1–8 © The Author(s) 2013 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1352458513491166 msj.sagepub.com Rotem Orbach1,2, Michael Gurevich1 and Anat Achiron1,2 Abstract Background: The cerebrospinal (CSF) constitutes a specific immune micro-environment to the central nervous system, thus it may contain specific biomarkers involved in the pathogenesis of multiple sclerosis (MS). Objectives: We aimed to study a large array of inflammatory CSF biomarkers in patients with suspected MS to recognize potential early diagnostic markers. Methods: CSF samples were obtained from 115 patients who presented with neurological symptomatology suggestive of MS as follows: clinically isolated syndrome (CIS) = 49, relapsing–remitting multiple sclerosis (RRMS) = 29, and other neurologic disorders (OND) = 37. Protein expression profiles of 30 inflammatory biomarkers were measured by multiplex Luminex bead assay and further analyzed by group comparison statistics, correlation studies and receiver- operating characteristic (ROC) analysis. Results: Interleukin-12 subunit p40 (IL12p40) demonstrated a significant differential expression pattern between the groups (CIS vs OND: p = 1.17*10–7; RRMS vs OND: p = 4.98*10–5), with higher levels in CIS and RRMS patients. ROC analysis demonstrated excellent diagnostic performance of IL12p40 for discrimination between CIS and OND patients (area under the curve = 0.87 (95% CI 0.78–0.93), p = 0.0001). No associations were found with disease activity or severity measures. Conclusions: An increased IL12p40 level characterizes the CSF of MS patients and appears to be helpful in identifying CIS and OND patients early in the process of clinical diagnostic assessment. Keywords Multiple sclerosis, immunology, clinically isolated syndrome, cerebrospinal fluid, biological markers, interleukin-12 subunit p40 Date received: 4 September 2012; accepted: 30 May 2013 Introduction into consideration the fact that brain MRI lesions could be nonspecific and mimic MS lesions despite meeting the cri- Making a firm diagnosis of multiple sclerosis (MS) could teria for DIT and DIS,3,4 CSF analysis remains a powerful be a difficult challenge at the early stage of the disease. The tool to evaluate specific central nervous system (CNS) revised 2010 McDonald diagnostic criteria for MS were inflammatory markers for diagnostic needs and to better simplified to allow diagnosis based on magnetic resonance understand disease-related mechanisms. imaging (MRI) evidence of dissemination in time (DIT) and space (DIS).1 Nevertheless, while the revised criteria Because of its physical proximity, the CSF offers an declare that cerebrospinal fluid (CSF) analysis is no longer immense opportunity to reveal specific biological path- needed to support the MRI requirement for DIS in relaps- ways and markers involved in MS pathogenesis, specifi- ing–remitting MS (RRMS), one must remember that an cally at the early stage of the disease. The commonly used exclusion of alternative diagnoses remains a cornerstone of test for the presence of CSF immunoglobulin G (IgG) diagnosis consolidation.2 The spectrum of diseases that must be excluded is wide and disease specific, and sensitive 1Multiple Sclerosis Center, Sheba Medical Center, Israel. biomarkers could aid enormously in the differential diagno- 2Sackler School of Medicine, Tel Aviv University, Israel. sis process.2 Respectively, the 2010 McDonald criteria authors acknowledge the importance of CSF findings and Corresponding author: emphasize that in the absence of specific biomarkers to Rotem Orbach, Neurogenomics Laboratory, Multiple Sclerosis Center, confirm the diagnosis there is a need to validate and test Sheba Medical Center, Tel- Hashomer, Ramat-Gan, 52621, Israel. such biomarkers prospectively.1 In accordance and taking Email: [email protected] 114Downloaded from msj.sagepub.com at Tel Aviv University on June 1, 2013

2 Multiple Sclerosis Journal 0(0) Table 1. Brain MRI data at onset. OND, n = 37 CIS = 49 RRMS = 29 0 25 22 Number of patients fulfilling MRI McDonald criteriaa 27 21 7 Number of patients with MRI lesions not fulfilling MRI McDonald criteriaa 10 3 0 Number of patients with no MRI lesions MRI: magnetic resonance imaging; OND: other neurologic disorders; CIS: clinically isolated syndrome; RRMS: relapsing–remitting multiple sclerosis. aPlease see Polman et al.6 for detailed description of MRI McDonald criteria. oligoclonal bands offers a high level of sensitivity, but its Table 2. Diagnoses among patients with other neurologic low positive predictive value (PPV) disables its usage as a disorders. single diagnostic tool in MS.5 Number of patients Diagnosis In the current work, we studied a wide array of biomark- ers in the CSF of patients with clinical presentation sus- 20 Nonspecific white matter lesions due to pected of MS. We aimed to identify CSF biomarkers that vascular etiology will be specific for the diagnosis and will enable us to dif- Ischemic lesions, history of vascular risk ferentiate other potential diagnostic possibilities. Based on factors, n = 12 a comprehensive neurological evaluation and prolonged aPL, n = 3 clinical follow-up together with the xMap technology for Vasculitis, n = 1 proteomic measurements, we characterized biomarkers that Small blood vessel disease due to differentiated between patients with clinically isolated syn- possible ischemic/vasculitis etiology, drome (CIS), RRMS, and those with other neurologic dis- n=4 orders (OND). 8 Degenerative disc disease, disc Materials and methods myelopathy Patients, consent and samples 2 Vascular malformation 2 Brain tumor The study population included a total of 115 patients (67 1 Communicating hydrocephalus females, 48 males; mean age ± SE, range (years): 37.2±1.2, 1 Migraine 16–68); all were referred to the Multiple Sclerosis Center in 1 Temporal cavernoma Sheba Medical Center, Israel, for evaluation with a diag- 1 Diabetic neuropathy nostic question of MS. The study was approved by the ethi- 1 Spinal stenosis cal committee of Sheba Medical Center. Written informed consent was obtained from each individual. CSF samples aPL: anti-phospholipid antibody; total: n = 37. were collected at the time of diagnostic procedure, between the period of 2007 and 2010. remaining 86 patients who were only suspected to have MS, a solid classification as CIS (n = 49) and other neuro- Inclusion criteria logic disorders (OND) (n = 37) was established after con- sidering the results of the thorough evaluation together with Inclusion criteria were: (1) onset of neurologic symptoma- a follow-up period of 2.3±0.2 (mean±SE (years)). Final tology suggestive of MS: (a) visual symptomatology with diagnoses among the OND group are presented in Table 2. the presentation of unilateral optic neuritis (n = 11); (b) long-tracts involvement with motor and/or sensory and/or For the purpose of the CSF biomarkers analyses, the urinary symptoms (n = 69); (c) brainstem and/or cerebellar complete cohort was divided into an experimental group (n symptoms (n = 35); (2) older than 16 years; (3) no steroids = 75; 27 CIS, 29 RRMS, 19 OND) and a validation group or immunomodulatory treatments received prior to CSF (n = 40; 22 CIS, 18 OND). The study design is presented in sample collection. Figure 1. Among 37 CIS patients, 24 patients received a solid diagnosis of RRMS during the follow-up period Following an initial comprehensive evaluation that (experimental group, n = 20/27 (74%); validation group, included medical history, physical and neurological exami- four of 22 (18%)). nation, brain MRI (Table 1), CSF oligoclonal bands evalu- ation (Hydrogel 3 and Hydrogel 6 CSF kits, Sebia Inc, Sample collection and preprocessing Norcross, GA, USA), and using the McDonald criteria,6 29 out of 115 patients were diagnosed with RRMS. For the Lumbar puncture (LP) was performed using standard guidelines;7 samples were immediately centrifuged, divided into aliquots, frozen (–196◦C), and stored until usage in polypropylene tubes (–80◦C). 115Downloaded from msj.sagepub.com at Tel Aviv University on June 1, 2013

Orbach et al. 3 Figure 1. Study flowchart. The study cohort was divided into two study groups: an experimental group for measuring the expression pattern of 30 inflammatory biomarkers and a validation group for validating the main result independently. CIS: clinically isolated syndrome; RRMS: relapsing–remitting multiple sclerosis; OND: other neurologic disorders. CSF biomarkers assessment by Luminex MCP-1, macrophage inflammatory protein (MIP)-1a, MIP-1b, nerve growth factor (NGF), platelet-derived growth Proteins documented in the literature as playing key roles in factor (PDGF)-BB, Rantes, tumor necrosis factor (TNF)a inflammation pathways were selected for the study. For this and TNFb, and vascular endothelial growth factor (VEGF). purpose, Human 30 Plex (Procarta, Affymetrix/Panomics, For validation, an IL12p40 single-plex assay (Procarta) was Fremont, CA, USA) was used. The kit encompasses a wide used. Protein expression profiles were obtained using the range of proteins, including: eotaxin, fibroblast growth fac- multiplex Luminex200 bead assay, based on the xMap tech- tor basic (FGF-b), granulocyte colony-stimulating factor nology (Linco Research, St. Charles, MO, USA). Preparation (G-CSF), granulocyte-monocyte colony-stimulating factor of samples, reagents and immunoassay procedure were per- (GM-CSF), growth-regulated protein (GRO)-α, interferon formed according to the manufacturer’s instructions. (IFN)-g, IL-1a, IL-1b, IL-2, IL-4, IL-5, IL-6, IL-7, IL-8, Converting median fluorescent intensity (MFI) to pg/ml was IL-10, IL-12(p40), IL-12(p70), IL-13, IL-17A, inducible carried out by Milliplex analyst software (VigeneTech, protein (IP)10, monocyte chemotactic protein (MCP)-3, Carlisle, MA, USA), using a five-parameter logistic. 116Downloaded from msj.sagepub.com at Tel Aviv University on June 1, 2013

4 Multiple Sclerosis Journal 0(0) Enzyme-linked immunosorbent 10,000 random comparisons was used to deal with the mul- assay (ELISA) tiple (30) comparisons and gives the probability of obtain- ing a particular p value by chance in 10,000 random ELISA for IL12p40 was conducted on a total of 40 samples assignments. The statistical thresholds for significance (14 CIS, 12 RRMS, 14 OND) that were available from were set to p < 0.05 and to p < 0.05 bootstrapping. The exploratory and validation groups (nine exploratory, 31 ROC analysis was used to determine the relationship validation). IL12p40 levels were measured using Human between sensitivity and specificity for the groups and to IL-12/IL-23 p40 Dimer Quantikine ELISA Kit (R&D quantify the overall ability of the test to discriminate indi- Systems Inc, Minneapolis, MN, USA) and Sunrise micro- viduals with CIS and those with OND. For the combined plate reader (TECAN, Männedorf, Switzerland). Results ROC analysis of experimental and validation samples were processed by Magellan data analysis software together, a pre-analysis linear scaling was performed by (TECAN, Männedorf, Switzerland). Partek software. Analyses Results CSF biomarker characterization in the experimental cohort Demographic and clinical characteristics of the experimen- was performed in three-step analyses. First, CIS to OND tal cohort are presented in detail in Table 3. As expected, comparison was performed to characterize early disease the following observations were noted: CIS and RRMS markers that have the potential of diagnostic utility accom- patients were younger as compared to the OND patients, panied by a receiver-operating characteristic (ROC) analy- disease duration was shorter in CIS and OND compared to sis to assess the diagnostic power of a significant biomarker. RRMS patients and oligoclonal bands were negative in the Second, RRMS to OND comparison was conducted to OND group. The mean CSF cell counts, total protein and characterize markers with different expression level that glucose, were within the normal range in all comparison persist throughout the disease course. Third, using the groups. expanded MS group (CIS and RRMS, n = 56), we analyzed markers’ associations with disease activity/severity meas- Early disease markers and their ures. For this purpose, disability was measured by the diagnostic utility Expanded Disability Status Scale (EDSS). Radiological disease activity was assessed by brain MRI exams (3.0 T A total of 30 proteins were measured in the CSF of the scanner, GE, Signa HDX) that were performed as part of experimental cohort; data on detection levels and assay pre- the diagnostic evaluation; only patients with MRI and LP cision is given in Supplementary Table 1. A single bio- time interval of less than one month were included (n = 27; marker, IL12p40, demonstrated highly significant 14 CIS, 13 RRMS). T2 lesions were recognized and counted differential expression pattern between CIS and OND by an experienced neuroradiologist particularly for the cor- patients. IL12p40 levels were above the detection threshold relation analysis. Disease phase at time of CSF collection in fewer than half of OND patients (n = 8; 42%) compared was determined for each patient; disease exacerbation was with the majority of the CIS group (n = 26; 96%) (two- defined as new and acute onset of neurologic deficits, tailed Fisher’s exact test, p = 0.006) (Figure 2(a)). IL12p40 reflected objectively by EDSS score worsening, and lasting levels comparison by Mann-Whitney test showed a highly for 48 hours or more (n = 23/56, 41.1%).8 Patients in clini- significant difference between the groups (p = 1.17*10–7, cal remission were defined by stable clinical evaluation and bootstrap for 104 randomization experiments p = 0; median a stable EDSS score (n = 33/56, 58.9%).9 (pg/ml): CIS 1.53, OND 0.4); results for all 30 proteins comparison by Mann-Whitney test are given in Statistics Supplementary Table 2. Descriptive and analytic statistics were performed using Age-matched groups were constructed in order to rule GraphPad (GraphPad Software 5.0, San Diego, CA), Partek out a possible age effect (16 CIS, mean age±SE: 40.3±1.6; 6.5 software (Partek Inc, St. Louis, MO, USA) and MedCalc 16 OND, mean age±SE: 39.8±2.2); the significance of the (MedCalc Software, Ostend, Belgium). A two-tailed results was re-established in the age-matched cohort and Fisher’s exact test was used for categorical variables com- IL12p40 levels were not correlated with age. parison. A two-sample t test was used for studying demo- graphic and clinical characteristics differences among the The IL12p40 ROC analysis exhibited excellent diagnos- comparison groups. Due to non-Gaussian distribution tic performance of IL12p40 with the area under the curve (D’Agostino & Pearson normality test), differences in lev- (AUC) of 0.96 (95% confidence interval (CI) 0.9–1.0), p < els of proteins were analyzed using the nonparametric 0.0001. At a cutoff of 0.7 pg/ml, the optimal sensitivity and Mann-Whitney and Spearman rank tests. Bootstrapping of specificity were obtained (88.9% and 94.7% respectively) (Figure 3). 117Downloaded from msj.sagepub.com at Tel Aviv University on June 1, 2013

Orbach et al. 5 Table 3. Demographic and clinical features of the cohorta. Total cohort, Experimental group, n = 75 Validation group, n = 40 n = 115 CIS, n = 27 RRMS, n = 29 OND, n = 19 p valueb CIS, n = 22 OND, n = 18 p valueb Age, y 37.2±1.2 33.1±2.0 35.1±2.4 43.4±2.7 pc = 0.003 33.2±2.3 45.28±3.2 pc = 0.003 0.6±0.4 4.3±0.9 pd = 0.03 pc = 0.009 Disease durationf, y 1.8±0.3 1.7±0.2 2.0±0.2 1.0±0.3 pe = n/s 0.6±0.1 1.7±0.4 pc = n/a pc = n/s EDSS/abnormal 1.8±0.1 18 (67) 17 (59) 15 (79) pc = n/s 1.6±0.3 11 (61) pc ≤ 0.001 neurologic 9 (33) 12 (41) pd = 0.009 pc = n/a examination, n (%) 67 (58) 17 (63) 15 (52) 9 (47) pe = 0.009 11 (50) 12 (67) 48 (42) 10 (53) 11 (50) 6 (33) Gender 43 (37) 0 (0) pc = n/a 13 (59) 0 (0) pd = n/a Female, n (%) n/a pe = n/s 1 (5) n/a 21 (95) Male, n (%) pc = n/s Oligoclonal bands pd = n/s positive, n (%) pe = n/s Relapse, n (%) 24 (31) 11 (41) 12 (41) Remission, n (%) 54 (69) 16 (59) 17 (59) pc ≤ 0.001 pd ≤ 0.001 pe = n/s pc =n/a pd =n/a pe = n/s y: years; CIS: clinically isolated syndrome; RRMS: relapsing–remitting multiple sclerosis; OND: other neurologic disorders; EDSS: Expanded Disability Status Scale; n/a: not applicable; n/s: not significant. aData are presented as mean±SE. bT test or two-sided Fisher’s exact test. cBetween CIS and OND. dBetween RRMS and OND. eBetween CIS and RRMS. fDisease duration = time interval between onset of symptomatology suggesting multiple sclero- sis (MS) to cerebrospinal (CSF) sampling. Markers profile in RRMS patients patients who are at MS risk and those who are not, a longer term follow-up and a larger cohort are required, hence these RRMS and OND group comparison revealed that IL12p40 findings should be considered only as preliminary. was also significantly elevated in the CSF of RRMS patients (Figure 2(a)), (p = 4.98*10–5, bootstrap for 104 ran- Independent cohort validation domization experiments p = 0.0003; median (pg/ml): RRMS 0.95, OND 0.4), hence suggesting an increased sig- IL12p40 levels were measured in the CSF of an inde- nificant level of IL12p40 not only at the early stage of the pendent validation cohort (n = 40; 24 females, 16 males; disease but also at later disease stages; results of all 30 pro- mean age±SE: 38.6±2.1; 22 CIS, 18 OND); data on teins comparison by Mann-Whitney test are given in detection levels and assay precision is given in Supplementary Table 2. Supplementary Table 1. This analysis validated a signifi- cant distinguished pattern of elevated IL12p40 levels in Interestingly, RRMS and CIS group comparison was not the CSF between CIS patients as compared to OND (p = significant for any of the markers; this is in line with the 0.035, bootstrap p = 0.032, median (pg/ml): CIS 5.64, fact that the classification of CIS and RRMS is essentially OND 3.61) (Figure 2(b)). The IL12p40 ROC analysis clinical and not an immunopathological-based approach. within the validation cohort returned an AUC of 0.69 (95% CI 0.53–1.00), p = 0.036. IL12p40 associations with disease activity measures Combined analysis of experimental and validation groups Levels of IL12p40 were not significantly different among patients experiencing acute disease exacerbation (n = A combined analysis of experimental and validation groups 23/56, 41%) and clinical remission (n = 33/56, 59%). In (n = 86; 49 CIS, 37 OND) was performed. The Mann- addition, Spearman rank correlation analyses of IL12p40 Whitney analysis repeated a significant elevated expression levels with EDSS and number of T2 lesions were not sig- among CIS patients (p = 8.5*10–9). The combined ROC nificant. Finally, there was no significant difference in analysis returned a significant AUC of 0.87 (95% CI 0.78– IL12p40 concentrations between CIS patients who con- 0.93), p = 0.0001. verted to MS and CIS patients who did not. However, in order to discriminate levels of the cytokine between CIS 118Downloaded from msj.sagepub.com at Tel Aviv University on June 1, 2013

6 Multiple Sclerosis Journal 0(0) Figure 2. IL12p40 distribution among the comparison groups. (a) Experimental group.The majority of MS patients had IL12p40 concentration above the detection threshold (CIS 96%, RRMS 86%) as compared to the OND group (42%). (b) Validation group.All samples were within the detectable range. Figures display median with interquartile range. IL12p40: interleu- kin-12 subunit p40; MS: multiple sclerosis; CIS: clinically isolated syndrome; RRMS: relapsing–remitting multiple sclerosis; OND: other neurologic disorders. ELISA validation MS was found to be a strong predictor for the diagnosis of CIS. Our complementary findings in the RRMS group fur- The ELISA validation study returned the IL12p40 concen- ther suggest that IL12p40 is a marker that characterizes the trations among 40 samples taken from both exploratory and ongoing disease process in MS, being elevated in both the validation patients. The Mann-Whitney analysis of the initial and later phases of the disease. ELISA results confirmed the significance of increased IL12p40 concentrations among CIS and RRMS patients IL-12 belongs to a family of cytokines produced primar- compared to OND patients (CIS vs OND: p < 0.0001, ily by monocytes, macrophages, dendritic cells and B cells. RRMS vs OND: p = 0.003) (Figure 4). The most distinctive activity of IL-12 is the ability to regu- late the balance between type 1 T helper cells (Th1) and Expression levels of IL12p40 were detected in all CIS type 2 T helper cells (Th2), promoting Th1 responses in patients (n = 14) (range: 3.0–92.7 pg/ml; median: 9.8 pg/ reply to antigenic stimulation.10 Exaggerated production of ml) in nine RRMS patients (75%) (range: 0.7–38.2 pg/ml; IL-12 was shown to play a key role in CNS inflammation median 12.3 pg/ml), and only five OND patients (36%) and glial activation during the animal model of experimen- (range: 0.2–1.8 pg/ml; median: 0.7 pg/ml). tal autoimmune encephalomyelitis (EAE) and MS.11 IL-12 is formed by the heterodimerization of IL12p40 subunit Discussion and ILp35 subunit.12 Becher et al. evaluated the role of IL-12 and its subunits in the EAE model of IL-12-deficient In the present study high levels of IL12p40 in the CSF of mice. They demonstrated that p40–/– EAE mice did not patients with neurologic symptomatology suggestive of show any clinical disease activity when immunized with 119Downloaded from msj.sagepub.com at Tel Aviv University on June 1, 2013

Orbach et al. 7 Figure 3. Receiver operating characteristic (ROC) analysis of the development of EAE.13 Moreover, there are clear IL12p40. associations between IL12p40 and MS as increased expres- Differentiating clinically isolated syndrome patients and patients with sion of IL12p40 was observed in acute MS plaque,14 and as other neurologic disorders based on IL12p40 concentration in the spinal it was shown that the mRNA of IL12p40 in peripheral fluid. (a) ROC curve, the “optimum” cut-off value, 0.7 pg/ml, is marked blood mononuclear cells of both RRMS and SPMS patients (enlarged dot) and determined so that the sum of sensitivity and speci- were increased compared with healthy controls.15 ficity is maximal. (b) Best calculated sensitivity and specificity with their 95% confidence intervals and cut-off values. Our present findings indicate the same inclination IL12p40: interleukin-12 subunit p40. regarding the distinct role of the p40 subunit. While IL12p40 level was increased significantly among MS Figure 4. ELISA validation. patients, IL-12 (IL12p70), the assembled protein, was IL12p40 concentrations among 40 samples taken from exploratory and under the detectable threshold in the CSF of the majority of validation patients. Note that the majority of OND patients had non- the experimental study participants (n = 72/75, 96%; detectable level of expression of IL12p40.All CIS patients had IL12p40 CIS=25, RRMS=29, OND=18). These findings confirm concentrations higher than the highest OND measured concentration studies reporting the superior potency of IL12p40 as com- (1.8 pg/ml). Figure displays median with interquartile range. pared to the heterodimeric assembled cytokine IL-12.16,17 ELISA: enzyme-linked immunosorbent assay; IL12p40: interleukin-12 Though we have no data in our study regarding IL23p19 subunit p40; CIS: clinically isolated syndrome; RRMS: relapsing–remitting and IL23 expression in the CSF, it is of interest to note that multiple sclerosis; OND: other neurologic disorders. that IL12p40 molecule also heterodimerizes with IL23p19 to form the bioactive IL-2318 and that IL12p40 homodimers myelin olygodendrocyte glycoprotein (MOG), while p35– are also known to have biological effects.16,17 /– mice were highly susceptible to EAE, suggesting that p40 containing cytokine distinct from IL-12 is essential for We did not find meaningful associations between the CSF expression of IL12p40 and disease activity or severity variables. This is in line with the results of the ustekinumab clinical trial in RRMS that failed to show efficacy in reduc- ing the cumulative number of gadolinium-enhancing T1-weighted lesions or disease exacerbations despite being a neutralizing monoclonal antibody against the p40 subu- nit.19 Studies with larger cohorts and longer follow-ups are required to further establish the associations between IL12p40 and disease activity/severity measures. The added value of this work is brought by the ROC results that highlight IL12p40 not just as an immunologic phenomenon in MS, but also as a possible diagnostic marker for MS-related CIS. While an ideal diagnostic test has an AUC of 1.00, the IL12p40 established a significant AUC of 0.96 and 0.69 in the experimental and validation groups, respectively, and a significant AUC of 0.87 in the common ROC analysis. These findings indicate the poten- tial ability of the marker to discriminate between CIS and OND patients at time of initial clinical diagnostic evalua- tion. In addition, using the optimal cut-off given by this analysis, a specificity of 95% and 94% was reached (exploratory group ROC analysis). Though CSF analysis is no longer a prerequisite needed for the diagnosis of MS, many patients in our cohort were referred for diagnosis work-up following initial clinical and radiological evalua- tions that did not lead to a final diagnosis. Many of these patients had clinical symptoms and/or neurological find- ings and/or brain MRI with nonspecific white matter lesions that could not discriminate MS from other diagnostic pos- sibilities. For these patients, CSF diagnostic biomarkers signify a potential tool to aid in establishing the diagnosis during the early stage of the evaluation. However, further studies with larger numbers of patients are required to determine the optimal concentration cut-off value for 120Downloaded from msj.sagepub.com at Tel Aviv University on June 1, 2013

8 Multiple Sclerosis Journal 0(0) IL12p40 as a diagnostic marker. As displayed in Table 3, 5. Awad A, Hemmer B, Hartung HP, et al. Analyses of cere- the OCB positivity was between 52% and 63% among CIS brospinal fluid in the diagnosis and monitoring of multiple and RRMS patients. These relatively low rates, as com- sclerosis. J Neuroimmunol 2010; 219: 1–7. pared to the reported percentage in the literature, reflect the standard values of our laboratory. 6. Polman CH, Reingold SC, Edan G, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald Crite- As all samples were taken from patients with suspected ria”. Ann Neurol 2005; 58: 840–846. MS, the established diagnostic accuracy of IL12p40 relates only to this study population. Other neuro-inflammation 7. Johnson KS. Lumbar puncture: Technique; indications; con- conditions such as infections are not represented in the traindications; and complications in adults. In:Basow DS (ed.) study cohort, mostly because they are easily differentiated UpToDate. MA: UpToDate in Waltham, 2012. from MS by other simple laboratory findings, hence they were not referred to the MS center for diagnosis work-up. 8. Vollmer T. The natural history of relapses in multiple sclero- sis. J Neurol Sci 2007; 256 (Suppl 1): S5-S13. We conclude, based on its differential expression pat- terns among CIS, RRMS and OND subjects and the sig- 9. Iuliano G, Napoletano R and Esposito A. Multiple sclerosis: nificant results of the ROC analysis, that IL12p40 could be Relapses and timing of remissions. Eur Neurol 2008; 59: applied as a biomarker to be used during the diagnostic pro- 44–48. cess of MS for recognizing MS-related CIS patients early in the disease course. 10. Gately MK, Renzetti LM, Magram J, et al. The interleu- kin-12/interleukin-12-receptor system: Role in normal and Acknowledgement pathologic immune responses. Annu Rev Immunol 1998; 16: 495–521. The authors thank Prof Joab Chapman and the Neurology Department physicians for their efforts in collecting CSF samples (Sheba 11. Brahmachari S and Pahan K. Role of cytokine p40 family in Medical Center, Israel), Ms Polina Sonis and Dr Anna Feldman multiple sclerosis. Minerva Med 2008; 99: 105–118. for their significant technical assistance. 12. Gee K, Guzzo C, Che Mat NF, et al. The IL-12 family of cyto- Conflict of interest statement kines in infection, inflammation and autoimmune disorders. Inflamm Allergy Drug Targets 2009; 8: 40–52. The authors declare that there are no conflicts of interest. 13. Becher B, Durell BG and Noelle RJ. Experimental autoim- Funding mune encephalitis and inflammation in the absence of inter- leukin-12. J Clin Invest 2002; 110: 493–497. This work was supported in part by the Israeli Ministry of Health, Chief Scientist Office (grant number 3-00000-6632).This work 14. Windhagen A, Newcombe J, Dangond F, et al. Expression was performed in partial fulfillment of the M.D. thesis require- of costimulatory molecules B7-1 (CD80), B7-2 (CD86), and ments of the Sackler Faculty of Medicine, Tel Aviv University, for interleukin 12 cytokine in multiple sclerosis lesions. J Exp R. Orbach. Med 1995; 182: 1985–1996. References 15. van Boxel-Dezaire AH, Hoff SC, van Oosten BW, et al. Decreased interleukin-10 and increased interleukin-12p40 1. Polman CH, Reingold SC, Banwell B, et al. Diagnostic cri- mRNA are associated with disease activity and character- teria for multiple sclerosis: 2010 revisions to the McDonald ize different disease stages in multiple sclerosis. Ann Neurol criteria. Ann Neurol 2011; 69: 292–302. 1999; 45: 695–703. 2. Miller DH, Weinshenker BG, Filippi M, et al. Differen- 16. Jana M and Pahan K. Induction of lymphotoxin-alpha by tial diagnosis of suspected multiple sclerosis: A consensus interleukin-12 p40 homodimer, the so-called biologically approach. Mult Scler 2008; 14: 1157–1174. inactive molecule, but not IL-12 p70. Immunology 2009; 127: 312–325. 3. Solomon AJ, Klein EP and Bourdette D. “Undiagnosing” multiple sclerosis: The challenge of misdiagnosis in MS. Neu- 17. Jana M and Pahan K. IL-12 p40 homodimer, but not IL-12 rology 2012; 78: 1986–1991. p70, induces the expression of IL-16 in microglia and macro- phages. Mol Immunol 2009; 46: 773–783. 4. Spain R and Bourdette D. The radiologically isolated syn- drome: Look (again) before you treat. Curr Neurol Neurosci 18. Oppmann B, Lesley R, Blom B, et al. Novel p19 protein Rep 2011; 11: 498–506. engages IL-12p40 to form a cytokine, IL-23, with biologi- cal activities similar as well as distinct from IL-12. Immunity 2000; 13: 715–725. 19. Segal BM, Constantinescu CS, Raychaudhuri A, et al. Repeated subcutaneous injections of IL12/23 p40 neutralising antibody, ustekinumab, in patients with relapsing–remitting multiple sclerosis: A phase II, double-blind, placebo-con- trolled, randomised, dose-ranging study. Lancet Neurol 2008; 7: 796–804. 121Downloaded from msj.sagepub.com at Tel Aviv University on June 1, 2013



Circulating angiogenic factors in monochorionic twin pregnancies complicated by twin-to-twin transfusion syndrome and selective intrauterine growth restriction American Journal of Obstetrics & Gynecology | 2014 ‫ פרופ' שלמה ליפיץ‬:‫מנחה‬ ‫ פרופ' יואב ינון‬:‫מנחה‬ ‫אולטרסאונד גינקולוגי‬ ‫אולטרסאונד גינקולוגי‬ [email protected] [email protected] ‫אלעד בן מאיר‬ ‫אונ' תל אביב‬ ‫השתתף כסטודנט בפרויקט ח“ץ‬ 2011-2012 ‫בין השנים‬ [email protected] 123

Research www.AJOG.org OBSTETRICS Circulating angiogenic factors in monochorionic twin pregnancies complicated by twin-to-twin transfusion syndrome and selective intrauterine growth restriction Yoav Yinon, MD; Elad Ben Meir, BSc; Alexandra Berezowsky, BSc; Boaz Weisz, MD; Eyal Schiff, MD; Shali Mazaki-Tovi, MD; Shlomo Lipitz, MD OBJECTIVE: To determine maternal plasma levels of soluble vascular (P < .01). In contrast, in the sIUGR group, sVEGFR-1 and sEng levels endothelial growth factor receptor-1 (sVEGFR-1), placental growth were significantly higher only at the late second trimester (P < .05). factor (PLGF), and soluble endoglin (sEng) in monochorionic diamniotic PLGF levels were significantly lower at the early and late second (MC/DA) twin pregnancies complicated by twin-to-twin transfusion trimester in both TTTS and sIUGR compared with controls (P < .01). syndrome (TTTS) or selective intrauterine growth restriction (sIUGR). Plasma concentrations of sVEGFR-1 were significantly higher among TTTS pregnancies compared with sIUGR at the late second trimester STUDY DESIGN: A longitudinal cohort study of pregnant women with MC/ (P ¼ .027). Cord blood levels of sVEGFR-1 were significantly higher in DA twins who were classified into 3 groups: (1) uncomplicated MC/DA the smaller intrauterine growth restricted twin compared with the twins (n ¼ 22), (2) TTTS (n ¼ 23), and (3) sIUGR (n ¼ 15). Maternal normal cotwin. plasma samples were obtained between 13-20 and 21-28 weeks of gestation and cord blood samples were collected at delivery. Maternal CONCLUSION: Monochorionic pregnancies complicated by TTTS and plasma concentrations of sVEGFR-1, PLGF, and sEng, as well as cord sIUGR are characterized by decreased angiogenic activity. The blood levels of sVEGFR-1 were measured by enzyme-linked immunoassay. disparity in severity of the antiangiogenic state between TTTS and sIUGR suggests that these 2 conditions may represent a continuum. RESULTS: Maternal plasma levels of sVEGFR-1 and sEng were significantly higher in patients with TTTS at the early and late second Key words: angiogenic factors, monochorionic twins, PLGF, sEng, trimester compared with normal monochorionic pregnancies sIUGR, sVEGFR-1, TTTS Cite this article as: Yinon Y, Ben Meir E, Berezowsky A, et al. Circulating angiogenic factors in monochorionic twin pregnancies complicated by twin to twin transfusion syndrome and selective intrauterine growth restriction. Am J Obstet Gynecol 2014;210:141.e1-7. M onochorionic twin pregnancies polycythemia sequence (TAPS), and ne- development of TTTS, it’s mere presence are associated with increased urologic injury to the surviving twin in is not sufficient and other additional fac- perinatal morbidity and mortality com- the event of intrauterine demise of its tors may play a role in the pathophysi- pared with their dichorionic counterparts cotwin.3-5 TTTS is the result of an un- ology of this condition and singleton pregnancies.1,2 The in- balanced transfusion of blood across creased risk of monochorionic pregnan- placental vascular anastomoses from one In addition, the single monochorionic cies is largely attributable to the presence twin (donor) to the other (recipient).6,7 placenta has to nourish 2 fetuses and is of vascular anastomoses connecting the However, even though almost all mono- often not equally shared, which may lead 2 fetal circulations. The intertwin an- chorionic twin placentas have vascular to growth restriction and severe discor- astomoses are nearly always present and anastomoses, TTTS develops only in dant birthweight.9 Selective intrauterine account for a range of pregnancy com- 10-15% of these pregnancies.8 There- growth restriction (sIUGR) affects about plications, including twin-to-twin trans- fore, although the presence of vascular 12-25% of monochorionic twin preg- fusion syndrome (TTTS), twins anemia anastomoses is mandatory for the nancies4 and is increasingly recognized as an important complication of mo- From the Department of Obstetrics and Gynecology, Chaim Sheba Medical Center, Tel Hashomer, nochorionic twin pregnancies. Indeed, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel. sIUGR is associated with an increased fetal and neonatal mortality and morbidity.10,11 Received July 8, 2013; revised Aug. 29, 2013; accepted Sept. 13, 2013. The ability to predict monochorionicity- associated complications of pregnancy The authors report no conflict of interest. is quite limited and is mainly based on first trimester sonographic markers.12,13 Presented at the 33rd annual meeting of the Society for Maternal-Fetal Medicine, San Francisco, CA, Moreover, the differentiation between Feb. 11-16, 2013. TTTS and sIUGR could be challenging as both might be manifested by discordant Reprints: Yoav Yinon, MD, Department of Obstetrics and Gynecology, Shelba Medical Center, growth, oligohydramnios in one sac and Tel-Hashomer, Israel 52621. [email protected]. 0002-9378/$36.00 � ª 2014 Mosby, Inc. All rights reserved. � http://dx.doi.org/10.1016/j.ajog.2013.09.022 FEBRUARY 2014 American Journal of Obstetrics & Gynecology 141.e1 124

Research Obstetrics www.AJOG.org absent or reverse end-diastolic velocity in linking abnormal placentation with preg- estimated fetal weight below the 10th the umbilical artery of the small twin. The nancy complications of monochorionic percentile in 1 twin and estimated fetal importance of distinctions between these twins. weight discordance of 25% or greater. 2 conditions stems from the fact that the Patients with a combined pathology of management of these complications is The aim of this study was to deter- TTTS and sIUGR were classified in the different: laser coagulation of the vascular mine maternal and fetal plasma levels of TTTS group. The inclusion criteria for anastomoses is the treatment of choice sVEGFR-1, PLGF, and sEng in mono- uncomplicated monochorionic preg- for TTTS,14 whereas the current man- chorionic diamniotic (MC/DA) twin nancies included: (1) appropriately agement options in monochorionic twin pregnancies complicated by TTTS or grown fetuses; (2) estimated fetal weight pregnancies with sIUGR include expec- sIUGR compared with uncomplicated difference of less than 25%; (3) normal tant management with close surveillance monochorionic twins. amniotic fluid volumes; (4) normal or selective umbilical cord occlusion in Doppler velocimetry in the umbilical cases of high risk of in-utero fetal demise, MATERIALS AND METHODS arteries; and (5) similar measurements although laser coagulation of anastomoses of the middle cerebral artery-peak sys- has also been used in certain cases.15 This was a prospective cohort study of tolic velocity among both twins. Patients patients with MC/DA twin pregnancies with chronic hypertension and pre- Angiogenesis plays a central role in who were enrolled between November gestational diabetes, as well as pregnan- normal placental development.16,17 It has 2010 and May 2012 at a single tertiary cies complicated with congenital been shown that angiogenic factors such as center. The study was approved by the anomalies or chromosomal abnormal- vascular endothelial growth factor (VEGF) local institutional ethics committee, and ities or intrauterine fetal death of 1 of the and placental growth factor (PLGF) and all patients provided a written informed twins at presentation were excluded. 2 antiangiogenic peptides produced by consent. The patients were classified into Patients whose pregnancies were the placenta, soluble vascular endothelial 3 groups: (1) uncomplicated mono- complicated by severe TTTS (defined as growth factor receptor-1 (sVEGFR-1) and chorionic pregnancies (n ¼ 22), (2) TTTS stage �2) were treated by feto- soluble endoglin (sEng), contribute to the TTTS (n ¼ 23), and (3) sIUGR (n ¼ 15). scopic laser ablation (n ¼ 20). Patient pathogenesis of preeclampsia18-21 as well All patients had a first trimester ultra- with TTTS stage 1 (n ¼ 3) were managed as IUGR,22-24 both associated with aber- sound, which confirmed the diagnosis of conservatively. Demographic data, ul- rant placentation. monochorionicity and established an trasound findings, and perinatal out- accurate gestational age. The diagnosis of comes were entered into a computerized Because there is no clear explanation TTTS was based on the internation- database. Serial samples of peripheral why only certain monochorionic twins are ally accepted ultrasound criteria: an MC blood were obtained throughout preg- complicated by TTTS or sIUGR, new twin pregnancy with polyhydramnios of nancy starting at the patient’s first visit to mechanisms, which play a role in the �8 cm deepest vertical pocket in the our fetal medicine clinic and thereafter pathophysiology of these conditions, must recipient (or �10 cm from 20 weeks of every 6 weeks. All the maternal samples be explored. We hypothesized that altered gestation onwards) and oligohydramnios used for analysis were taken before laser angiogenesis mediated by sVEGFR-1, of �2 cm deepest vertical pocket in the sEng and PLGF might be the mechanism donor. Selective IUGR was defined as an TABLE Characteristics and pregnancy outcome of study patients Characteristic Normal (n [ 22) TTTS (n [ 23) sIUGR (n [ 15) P value 28 (22e35) NS Maternal age, y 32 (21e41) 28 (21e44) 33.3 NS 22.5 (20e32) NS Primigravida, % 22.7 26.1 0 NS Prepregnancy BMI, kg/m2 22.4 (19e27) 22.8 (18e30) 19 (16e20) NS Smoking, % 4.5 13 24.2 (21e27) NS Gestational age at blood draw 17.7 (13e20) 17.5 (15e20) 32.5 (24e38) .009 (early second trimester, wk) 10 NS Gestational age at blood draw 23.5 (21e27) 22 (21e28) (late second trimester, wk) Gestational age at delivery, wk 36.2 (36e37) 32.2 (24e36) Sample storage time, mos 11 8 Values are expressed as median (range) or as percentage BMI, body mass index; NS, not significant; sIUGR, selective intrauterine growth restriction; TTTS, twin-to-twin transfusion syndrome. Yinon. Angiogenic factors in monochorionic twin pregnancies. Am J Obstet Gynecol 2014. 141.e2 American Journal of Obstetrics & Gynecology FEBRUARY 2014 125

www.AJOG.org Obstetrics Research treatment or selective termination. Cord FIGURE 1 blood samples were obtained at delivery. Blood samples were collected in tubes Maternal plasma levels of sVEGFR-1 containing EDTA, centrifuged at 4�C for 10 minutes and stored at �70�C until A, Box plot with whiskers of sVEGFR-1 levels at 13-20 weeks of gestation in normal MC twins (n ¼ further analysis. Maternal plasma levels 22), TTTS (n ¼ 23), and sIUGR (n ¼ 15) pregnancies. B, Box plot with whiskers of sVEGFR-1 levels at of sVEGFR-1, sEng, and PLGF and cord 21-28 weeks of gestation in normal MC twins (n ¼ 22), TTTS (n ¼ 23), and sIUGR (n ¼ 15) blood levels of sVEGFR-1 were deter- pregnancies. The bold line represents the median, and data in the box represents the interquartile mined by enzyme-linked immunoassays range (25e75 percentile). (R&D Systems, Minneapolis, MN). All samples were assayed in duplicate at the MC, monochorionic; sIUGR, selective intrauterine growth restriction; sVEGFR-1, soluble vascular endothelial growth factor receptor-1; same time using the same standard curve TTTS, twin-twin transfusion syndrome. to minimize interassay variation. The calculated interassay coefficients of Yinon. Angiogenic factors in monochorionic twin pregnancies. Am J Obstet Gynecol 2014. variation for sVEGFR-1, sEng, and PLGF were 7.4%, 7.1%, and 8.3%, respectively. radiofrequency ablation. All 3 groups Figure1, B), and higher median plasma The calculated intraassay coefficients of were comparable in terms of maternal levels of sEng (22.8 vs 7.1 ng/mL, variation for sVEGFR-1, sEng, and PLGF age, parity, maternal pregestational body respectively, P < .01, Figure 2, B) at 21- were 2.4%, 2.6%, and 4.8%, respectively. mass index, and the rate of smokers. 28 weeks of gestation. In contrast, median None of the pregnancies in our cohort plasma levels of sVEGFR-1 (1671.4 vs Normality of the data was tested using were complicated by preeclampsia. As 1359.9 pg/mL, respectively, P ¼ .13 Kolmogorov-Smirnov test. Because the expected, patients with TTTS and sIUGR Figure 1, A), and sEng (9.6 vs 6.6 ng/mL, data did not fit a normal distribution, delivered earlier compared with the un- respectively, P ¼ .27, Figure 2, A) did values of the factors tested are presented as complicated MC twins (32.2, 32.5, and not differ significantly between patients median and interquartile range. The 3 36.2 weeks of gestation respectively, with sIUGR and controls at 13-20 weeks groups were compared using Kruskal- Kruskal-Wallis P ¼ .009). of gestation. However, at 21-28 weeks Wallis test, and comparison of contin- of gestation patients with sIUGR had uous variables between every 2 groups was The gestational age at blood sampling significantly higher median plasma levels conducted using Mann-Whitney U test. did not differ among the 3 groups of sVEGFR-1 and sEng compared with (Table 1). normal MC twins (3237.2 pg/mL vs The c2 test was used for comparison of 1927 pg/mL, P ¼ .035; 18.8 ng/mL vs We have divided the time frame in 7.1 ng/mL, P < .01, respectively, categorical variables. Significance was which maternal blood samples were Figures 1, B, and 2, B). accepted at P < .05. Statistical analyses analyzed into 2: early second trimester were conducted using the IBM Statistical (13-20 weeks of gestation) and late sec- Both patients with TTTS and sIUGR Package for the Social Sciences (IBM SPSS ond trimester (21-28 weeks of gestation). had significant lower median plasma v.19; IBM Corporation Inc, Armonk, NY). Patients with TTTS had significantly levels of PLGF compared with normal higher median plasma levels of sVEGFR-1 MC twins at the early second trimester RESULTS (3835.2 pg/mL vs 1359.9 pg/mL, respec- (140.3 and 281.9 vs 514.8 pg/mL, P < tively, P < .01, Figure 1, A), and higher .01, respectively, Figure 3, A) as well as at Sixty patients were included in the study; median plasma levels sEng (13 ng/mL vs the late second trimester (263.7 and 22 were uncomplicated monochorionic 6.6 ng/mL, respectively,P < .01, Figure 2, 350.1 pg/mL vs 956.8 pg/mL, P < .01, twins, 23 were diagnosed with TTTS, and A) compared with normal MC twins at respectively, Figure 3, B). 15 with sIUGR. The demographic and 13-20 weeks. Similarly, patients with clinical characteristics of the study pop- TTTS had significantly higher median Comparison between TTTS and ulation are shown in Table 1. Of the 23 plasma levels of sVEGFR-1 (6068.7 sIUGR: the median plasma levels of patients with TTTS included in this vs 1927 pg/mL, respectively, P < .01 sVEGFR-1 were higher among patients study, 3 had a stage 1 disease and were followed conservatively, 12 had a stage 2 disease, and 8 presented with TTTS stage 3-19 of them underwent fetoscopic laser ablation and 1 patient had selective termination. Four of the patients in the TTTS group had a combined pathology of TTTS and sIUGR. Twelve of the 15 patients with sIUGR had abnormal Doppler in the umbilical artery; 10 had persistent AEDV, and 2 had intermittent AEDV. Five of the patients in this group underwent selective termination with FEBRUARY 2014 American Journal of Obstetrics & Gynecology 141.e3 126

Research Obstetrics www.AJOG.org FIGURE 2 TTTS. In the first patient, blood was initially sampled at 15 weeks of gestation Maternal plasma levels of sEng when only amniotic fluid discordance appeared on ultrasound and subse- A, Box plot with whiskers of sEng levels at 13-20 weeks of gestation in normal MC twins (n ¼ 22), quently developed severe TTTS at 20 TTTS (n ¼ 23), and sIUGR (n ¼ 15) pregnancies. B, Box plot with whiskers of sEng levels at 21-28 weeks of gestation. In the second patient, weeks of gestation in normal MC twins (n ¼ 22), TTTS (n ¼ 23), and sIUGR (n ¼ 15) pregnancies. blood samples were initially collected The bold line represents the median, and data in the box represents the interquartile range (25e75 at 18 weeks of gestation followed by percentiles). development of severe TTTS at 21 weeks of gestation. In both patients, sVEGFR-1 MC, monochorionic; sEng, soluble endoglin; sIUGR, selective intrauterine growth restriction; TTTS, twin-twin transfusion syndrome. levels were higher and PLGF levels were lower at the first blood sample 5 Yinon. Angiogenic factors in monochorionic twin pregnancies. Am J Obstet Gynecol 2014. and 3 weeks before the development of TTTS. with TTTS compared with patients with Maternal plasma levels of sVEGFR-1, sIUGR although this difference was sig- sEng, and PLGF as a function of gesta- Among newborns, cord blood analysis nificant only at the late second trimester tional age are shown in Figure 4. revealed significantly higher median (6068.7 vs 3237.2 pg/mL, P ¼ .027, plasma levels of sVEGFR-1 among the Figure 1, B). In 2 of our TTTS cases, maternal blood IUGR twins compared with the normal was sampled before the development of cotwins (458.2 pg/mL vs 100.7 pg/mL, P < .01, Figure 5). In contrast, no dif- FIGURE 3 ferences were found between the twins of normal MC pregnancies or between Maternal plasma levels of PLGF the former donors and former recipients of TTTS pregnancies. Six of the 8 TTTS A, Box plot with whiskers of PLGF levels at 13-20 weeks of gestation in normal MC twins (n ¼ 22), pregnancies whose cord blood was TTTS (n ¼ 23), and sIUGR (n ¼ 15) pregnancies. B, Box plot with whiskers of PLGF levels at 21-28 analyzed for sVEGFR-1 level had laser weeks of gestation in normal MC twins (n ¼ 22), TTTS (n ¼ 23), and sIUGR (n ¼ 15) pregnancies. ablation during pregnancy. The bold line represents the median, and data in the box represents the interquartile range (25e75 percentiles). Angiogenic factors in MC twin pregnancies COMMENT MC, monochorionic; PLGF, placental growth factor; sIUGR, selective intrauterine growth restriction; TTTS, twin-twin transfusion Our study have demonstrated that syndrome. TTTS pregnancies are characterized by increased maternal plasma levels of Yinon. Angiogenic factors in monochorionic twin pregnancies. Am J Obstet Gynecol 2014. sVEGFR-1 and sEng and decreased maternal plasma levels of PLGF both at the early and late second trimester compared with normal MC twins. In contrast, in pregnancies complicated by sIUGR, sVEGFR-1, and sEng were significantly increased only at the late second trimester. Moreover, maternal plasma levels of sVEGFR-1 were signifi- cantly increased among TTTS pregnan- cies compared with sIUGR. Finally, we have shown elevated cord blood levels of sVEGFR-1 among the small IUGR twins compared with their normal cotwins but no differences between the donor and recipient following pregnancies compli- cated by TTTS. TTTS is thought to result from an unbalanced intertwin blood flow th- rough placental vascular anastomoses. At least 1 unidirectional arteriovenous anastomosis is required for the devel- opment of TTTS, whereas the presence of an arterioarterial anastomosis seems 141.e4 American Journal of Obstetrics & Gynecology FEBRUARY 2014 127

www.AJOG.org Obstetrics Research FIGURE 4 concentrations of antiangiogetic factor in the TTTS group support this hypothesis. Maternal plasma levels of sVEGFR-1, sEng, and PLGF as a function Even though TTTS and sIUGR have strict of gestational age definitions and are regarded as 2 distinct pathologies, the findings reported herein A, Maternal plasma levels of sVEGFR-1 as a function of gestational age. B, Maternal plasma levels of suggest that these 2 conditions may sEng as a function of gestational age. C, Maternal plasma levels of PLGF as a function of gestational represent a continuum. In other words, age. assuming that antiangiogenic state plays an important role in the pathophysiology PLGF, placental growth factor; sEng, soluble endoglin; sVEGFR-1, soluble vascular endothelial growth factor receptor-1. of TTTS and sIUGR, sIUGR may be regarded as a forme fruste of TTTS. Yinon. Angiogenic factors in monochorionic twin pregnancies. Am J Obstet Gynecol 2014. However, more evidence is needed to establish this hypothesis. to be protective.6,7 However, a large body between TTTS pregnancies and uncom- of evidence strongly suggests that the plicated MC twins. Fox et al26 have Growth in monochorionic twins is pathophysiology of TTTS cannot be showed a transient increase in sVEGFR-1/ determined by the division of the single attributed solely to vascular anastomoses PLGF ratio after laser ablation followed placenta between the twins as well as because almost all monochorionic pla- by reduction to below basal levels by by the vascular anastomoses.9,27 All the centas have vascular anastomoses but 1 week. Despite this consistent ob- sIUGR cases included in our study were only 10-15% develop TTTS.8 Therefore, servation of decreased angiogenesis in characterized by early-onset discordant other pathologic processes may play a pregnancies complicated by TTTS, the growth, which typically have an un- role in the pathophysiology of this dis- initiating event leading to increased equally shared placenta with large ease.4 Our findings indicate that TTTS placental secretion of sVEGFR-1 and anastomoses.27 Because vascular anas- is an anti-angiogenic state, supporting sEng is still unknown. tomoses also influence growth in mon- previous reports, which have showed ochorionic twins, unbalanced net maternal angiogenic activity to be Our study is the first to show differ- arteriovenous transfusion as occurs in decreased in TTTS.25,26 Kusanovic et al25 ences in the levels of angiogenic factors TTTS may result in growth restriction of have collected blood samples from 16 between TTTS and sIUGR with increased the donor twin. Therefore, each of these patients with TTTS between 16-26 weeks maternal plasma levels of sVEGFR-1 2 conditions could evolve into the other of gestation and found elevated levels of in pregnancies complicated by TTTS resulting in a combined pathology of sVEGFR-1 and sEng and decreased levels compared with sIUGR. The anti- TTTS and sIUGR.27,28 Moreover, the of PLGF compared with women with angiogenic state became apparent in clinical distinction between these pa- monochorionic twins without TTTS. sIUGR pregnancies only at the late second thologies is not always clear. Therefore, it Similarly, Fox et al26 reported on trimester. The results of this study is possible that maternal plasma levels of increased plasma levels of sVEGFR-1 as strongly suggest that TTTS is associated angiogenic factors may assist in differ- well as increased sVEGFR-1/PLGF ratio with intense severity of antiangiogenic entiating between sIUGR and TTTS among patients with TTTS, but no dif- state compared with sIUGR. Both pregnancies or even predict later devel- ferences were found in PLGF levels temporal differences, ie, earlier evidence opment of TTTS in pregnancies com- of antiangiogenic state, and higher plicated by sIUGR. In contrast to our findings, Kusanovic et al25 did not ob- serve any differences in maternal plasma levels of sVEGFR-1, sEng, and PLGF between MC twins with and without IUGR. The fact that most of our sIUGR patients had AEDV in the umbilical ar- tery as well as early onset of growth re- striction might explain the differences between these 2 studies. Finally, cord blood analysis revealed increased sVEGFR-1 levels in the IUGR twin compared with the normal cotwin. This observation is in accordance with previous studies on MC twins with sIUGR showing increased expression of sVEGFR-1 as well as endoglin in the IUGR twin placenta compared with the normal cotwin’s placenta.22,23 However, FEBRUARY 2014 American Journal of Obstetrics & Gynecology 141.e5 128

Research Obstetrics www.AJOG.org FIGURE 5 To conclude, our study demonstrates Cord blood levels of sVEGFR-1 that TTTS is characterized by an anti- angiogenic state, in accordance with pre- vious reports. Although antiangiogenic state becomes apparent only at the late second trimester in pregnancies compli- cated by sIUGR, TTTS is characterized by antiangiogenic state that is earlier and more intense in comparison. The dis- parity in severity of the antiangiogenic state between TTTS and sIUGR suggests that these 2 conditions may represent a continuum. Further studies are required to determine the predictive value of angiogenic and antiangiogenic factors in MC pregnancies as well as their impor- tance in early distinction between sIUGR and TTTS. - Box plot with whiskers of cord blood levels of sVEGFR-1 in normal MC twins (n ¼ 8 pairs of twins), REFERENCES TTTS (n ¼ 7 pairs of twins) and sIUGR (n ¼ 6 pairs of twins). The bold line represents the median, and data in the box represents the interquartile range (25e75 percentiles). 1. Dube J, Dodds L, Armson BA. Does chorio- nicity or zygosity predict adverse perinatal out- MC, monochorionic; sIUGR, selective intrauterine growth restriction; sVEGFR-1, soluble vascular endothelial growth factor receptor-1; comes in twins? Am J Obstet Gynecol TTTS, twin-twin transfusion syndrome. 2002;186:579-83. 2. Sebire NJ, Snijders RJ, Hughes K, Yinon. Angiogenic factors in monochorionic twin pregnancies. Am J Obstet Gynecol 2014. Sepulveda W, Nicolaides KH. The hidden mor- tality of monochorionic twin pregnancies. Br J unbalanced net arteriovenous trans- and whether the antiangiogenic state is a Obstet Gynaecol 1997;104:1203-7. fusion may have also contributed to the cause or consequence of TTTS. Interest- 3. Denbow ML, Cox P, Taylor M, Hammal DM, difference in cord blood levels of ingly, in 2 of our cases the alteration in the Fisk NM. Placental angioarchitecture in mono- sVEGFR-1 between the IUGR twin and maternal plasma levels of sVEGFR-1 and chorionic twin pregnancies: relationship to fetal the normal cotwin. Despite previous PLGF preceded the onset of TTTS by 3-5 growth, fetofetal transfusion syndrome, and report showing that VEFGR-1 mRNA is weeks. Therefore, further studies are pregnancy outcome. Am J Obstet Gynecol overexpressed in the villi of the donor in needed to evaluate the predictive value of 2000;182:417-26. some cases of TTTS,29 we did not find these factors in MC pregnancies. It is 4. Lewi L, Van Schoubroeck D, Gratacos E, any differences in cord blood levels of possible that by measuring maternal cir- Witters I, Timmerman D, Deprest J. Mono- sVEGFR-1 between the donor and the culatory levels of sVEGFR-1, sEng, and chorionic diamniotic twins: complications and recipient, perhaps because most of these PLGF early in gestation, we will be able to management options. Curr Opin Obstet Gyne- cases were treated by laser ablation dur- detect which of the MC pregnancies are col 2003;15:177-94. ing pregnancy with resolution of the prone to develop TTTS or sIUGR later in 5. Gucciardo L, Lewi L, Vaast P, et al. Twin condition. gestation. Furthermore, measurement of anemia polycythemia sequence from a prenatal these factors might be proven useful perspective. Prenat Diagn 2010;30:438-42. As a tertiary fetal therapy center, many clinically in differentiating between TTTS 6. Diehl W, Hecher K, Zikulnig L, Vetter M, of the patients were referred to us only and sIUGR. Hackeloer BJ. Placental vascular anastomoses when TTTS or sIUGR were suspected and visualized during fetoscopic laser surgery in se- therefore, in most cases, the samples were The strength of this study lies in its vere mid-trimester twin-twin transfusion syn- collected when the diagnosis was already cohort of 3 well-defined subgroups of drome. Placenta 2001;22:876-81. established. Hence, the current study monochorionic twin pregnancies and 7. Bermudez C, Becerra CH, Bornick PW, cannot determine whether these factors being the first study to compare the levels Allen MH, Arroyo J, Quintero RA. Placental types can predict development of TTTS or of angiogenic factors between TTTS and and twin-twin transfusion syndrome. Am J sIUGR in MC twins early in pregnancy, sIUGR. Obstet Gynecol 2002;187:489-94. 8. Lutfi S, Allen VM, Fahey J, O’Connell CM, Vincer MJ. Twin-twin transfusion syndrome: a population-based study. Obstet Gynecol 2004;104:1289-97. 9. Lewi L, Cannie M, Blickstein I, et al. Placental sharing, birthweight discordance, and vascular anastomoses in monochorionic diamniotic twin placentas. Am J Obstet Gynecol 2007;197:587. e1-8. 10. Gratacos E, Carreras E, Becker J, et al. Prevalence of neurological damage in 141.e6 American Journal of Obstetrics & Gynecology FEBRUARY 2014 129

www.AJOG.org Obstetrics Research monochorionic twins with selective intrauterine vascular endothelial cells. Placenta 2002;23: 24. Boutsikou T, Malamitsi-Puchner A, growth restriction and intermittent absent or 742-50. Economou E, Boutsikou M, Puchner KP, reversed end-diastolic umbilical artery flow. Ul- 17. Kaufmann P, Mayhew TM, Charnock- Hassiakos D. Soluble vascular endothelial trasound Obstet Gynecol 2004;24:159-63. Jones DS. Aspects of human fetoplacental vas- growth factor receptor-1 in intrauterine growth 11. Adegbite AL, Castille S, Ward S, Bajoria R. culogenesis and angiogenesis. II. Changes during restricted fetuses and neonates. Early Hum Dev Neuromorbidity in preterm twins in relation to normal pregnancy. Placenta 2004;25:114-26. 2006;82:235-9. chorionicity and discordant birth weight. Am J 18. Maynard SE, Min JY, Merchan J, et al. 25. Kusanovic JP, Romero R, Espinoza J, et al. Obstet Gynecol 2004;190:156-63. Excess placental soluble fms-like tyrosine kinase Twin to twin transfusion syndrome: an anti- 12. Lewi L, Lewi P, Diemert A, et al. The role of 1 (sFlt1) may contribute to endothelial dysfunc- angiogenic state? Am J Obstet Gynecol ultrasound examination in the first trimester and tion, hypertension, and proteinuria in pre- 2008;198:382.e1-8. at 16 weeks’ gestation to predict fetal compli- eclampsia. J Clin Invest 2003;111:649-58. 26. Fox CE, Lash GE, Pretlove SJ, Chan BC, cations in monochorionic diamniotic twin preg- 19. Levine RJ, Maynard SE, Qian C, et al. Holder R, Kilby MD. Maternal plasma and nancies. Am J Obstet Gynecol 2008;199:493. Circulating angiogenic factors and the risk of amniotic fluid angiogenic factors and their re- e1-7. preeclampsia. N Engl J Med 2004;350:672-83. ceptors in monochorionic twin pregnancies 13. Memmo A, Dias T, Mahsud-Dornan S, 20. Venkatesha S, Toporsian M, Lam C, et al. complicated by twin to twin transfusion syn- Papageorghiou AT, Bhide A, Thilaganathan B. Soluble endoglin contributes to the patho- drome. Ultrasound Obstet Gynecol 2010;35: Prediction of selective fetal growth restriction genesis of preeclampsia. Nat Med 2006;12: 695-701. and twin-to-twin transfusion syndrome in mon- 642-9. 27. Lewi L, Deprest J, Hecher K. The vascular ochorionic twins. BJOG 2012;119:417-21. 21. Levine RJ, Lam C, Qian C, et al. Soluble anastomoses in monochorionic twin pregnan- 14. Senat MV, Deprest J, Boulvain M, Paupe A, endoglin and other circulating antiangiogenic cies and their clinical consequences. Am J Winer N, Ville Y. Endoscopic laser surgery factors in preeclampsia. N Engl J Med Obste Gynecol 2013;208:19-30. versus serial amnioreduction for severe twin- 2006;355:992-1005. 28. Lopriore E, Deprest J, Slaghekke F, et al. to-twin transfusion syndrome. N Eng J Med 22. Yinon Y, Nevo O, Xu J, et al. Severe intra- Placental characteristics in monochorionic 2004;351:136-44. uterine growth restriction pregnancies have twins with and without twin anemia- 15. Lewi L, Gratacos E, Ortibus E, et al. Preg- increased placental endoglin levels: hypoxic polycythemia sequence. Obstet Gynecol nancy and infant outcome of 80 consecutive regulation via transforming growth factor-beta 3. 2008;112:753-8. cord coagulations in complicated mono- Am J Pathol 2008;172:77-85. 29. Kumazaki K, Nakayama M, Suehara N, chorionic multiple pregnancies. Am J Obstet 23. Nevo O, Many A, Xu J, et al. Placental Wada Y. Expression of vascular endothelial Gynecol 2006;194:782-9. expression of soluble fms-like tyrosine kinase 1 growth factor, placental growth factor, and their 16. Mayhew TM. Fetoplacental angiogenesis is increased in singletons and twin pregnancies receptors Flt-1 and KDR in human placenta during gestation is biphasic, longitudinal and with intrauterine growth restriction. J Clin under pathologic conditions. Hum Pathol occurs by proliferation and remodelling of Endocrinol Metab 2008;93:285-92. 2002;33:1069-77. FEBRUARY 2014 American Journal of Obstetrics & Gynecology 141.e7 130

Pituitary function in children following infectious diseases of the central nervous system Pituitary | 2014 ‫ פרופ' אורית חמיאל‬:‫מנחה‬ ‫מנהלת יחידה לאנדוקרינולוגיה ילדים‬ [email protected] ‫ענבל גזית‬ ‫אונ' תל אביב‬ ‫השתתפה כסטודנטית בפרויקט ח“ץ‬ 2010-2011 ‫בין השנים‬ [email protected] 131

Pituitary (2014) 17:118–124 DOI 10.1007/s11102-013-0476-2 Pituitary function in children following infectious diseases of the central nervous system Yael Levy-Shraga • Inbal Gazit • Dalit Modan-Moses • Orit Pinhas-Hamiel Published online: 8 March 2013 Ó Springer Science+Business Media New York 2013 Abstract Recent studies in adults suggest that pituitary meningitis by the primary care physician is probably suffi- deficiencies develop in a considerable proportion of patients cient. Invasive assessments should be reserved for selected who recover from infectious meningitis. The aim of this cases where there is slow growth or other clinical suspicion study was to evaluate pituitary function of children with a of hypopituitarism. history of meningitis. Seventy-nine children were admitted to the Safra Children’s Hospital due to meningitis between Keywords Hypopituitarism Á Meningitis Á 2007 and 2010. Twenty-four families were lost for follow- Pituitary Á Growth up, 55 were interviewed by phone and 14 (9 males) partici- pated in the study. Evaluation included medical history, Abbreviations physical examination, auxological measurements and basal CNS Central nervous system levels of TSH, fT4, cortisol and IGF1. Children with TBI Traumatic brain injury abnormal results were followed for a year and dynamic CSF Cerebrospinal fluid testing was performed when indicated. Mean age at time of TSH Thyrotropin infectious meningitis was 3.8 ± 5.4 years (range 0.03– fT4 Free thyroxin 15.8), and at clinical evaluation 6.4 ± 6.4 (range 1.2–20). IGF1 Insulin-like growth factor-1 The interval between the acute event and evaluation was GH Growth hormone 2.7 ± 1.2 years. Thyroid function tests and basal cortisol FSH Follicle-stimulating hormone levels were normal for all children. Three children had low LH Luteinizing hormone IGF1 levels; however over a year of follow-up two of them APA Anti-pituitary antibodies had normal height and growth velocity, making growth AHA Anti-hypothalamus antibodies hormone deficiency unlikely. One child had low height SDS, but exhibited a normal response to a growth hormone stim- Introduction ulation test. Pituitary dysfunction with overt clinical symp- toms is not a frequent consequence of acute meningitis in children. Follow-up of growth and puberty of children post- Y. Levy-Shraga (&) Á I. Gazit Á D. Modan-Moses Á Anterior pituitary hormone dysfunction was found to be an O. Pinhas-Hamiel important feature of long term morbidity in subjects of Pediatric Endocrine and Diabetes Unit, Safra Children’s traumatic brain injury (TBI) [1]. A recent meta-analysis of Hospital, Sheba Medical Center, 52621 Tel Hashomer, Israel adult TBI data reported the rate of long-term hypopituita- e-mail: [email protected] rism as 27.5 %, most commonly affecting growth hormone (GH) or gonadotropins [2]. Pituitary dysfunction may be Y. Levy-Shraga Á I. Gazit Á D. Modan-Moses Á detected months and even years after the injury. Cerebral O. Pinhas-Hamiel edema and the release of inflammatory mediators have Sackler School of Medicine, Tel-Aviv University, Tel Aviv, been implicated in causing injury to the pituitary gland and Israel 123 132

Pituitary (2014) 17:118–124 119 hypothalamus [3]. Therefore, routine pituitary assessment confirmed by blood and CSF results. According to these after moderate to severe TBI has been recommended for criteria, 69 were diagnosed with viral meningitis and 7 with adult population [4] and extrapolated to children as well bacterial meningitis (Streptococcus Pneumonia was identi- [5]. fied in 3 patients, Neisseria Meningitis in 2 patients, Hae- mophilus Influenza and Staphylococcus coagulase negative Recently, studies among adults who recovered from in 1 patient each). Three children received antibiotic treat- acute meningitis demonstrated a similar phenomenon ment prior to arrival to the hospital and were labeled as [6–9]. About one-third of adults developed hypopituitarism partially treated meningitis with negative culture CSF. both during the acute phase and a year or more after the infection. Among the pediatric population, tuberculous Out of the 79 files that were screened, 24 families were meningitis has long been known as a cause of hypotha- lost to follow-up. Fifty-five families were interviewed by lamic-pituitary dysfunction [10]. Partial hypopituitarism, phone regarding their child’s growth, development and particularly GH deficiency, has been documented in 20 % of medical problems and were invited to participate in the patients several years after recovery from tuberculous study. Forty-one declined. The most common reason given meningitis in childhood [11]. In addition, several case was having no concern about the child’s health or devel- reports described hypopituitarism following bacterial men- opment. Fourteen families agreed to participate in the study ingitis [12, 13]. However, to our knowledge, no systematic and signed informed consent. study has yet been conducted in children. Methods Although many western countries have a mandatory reporting system regarding meningitis, the exact incidence All participants included in the study underwent: rate is unknown. In Greece, for example, the estimated mean annual incidence rate was 16.9/100,000 for bacterial 1. During the first study visit, written informed consent meningitis [14]. Population-wide studies have shown that was obtained from the parents or the patient, as viral meningitis is even more common [15]. The implica- needed. Detailed medical history was obtained regard- tions of hypopituitarism in children, relating to growth and ing clinical features suggestive of hypopituitarism puberty, are substantial. Therefore, the aim of this pre- (Table 1) and physical examination, including height liminary study was to assess pituitary function in children and weight measurements, was conducted. Pubertal with a history of infectious disease of the CNS. To address staging was assessed as per Marshall and Tanner [16]. this issue, the children underwent an evaluation that Basal hormone levels were obtained. included clinical, auxological and hormonal assessments. 2. Participants with height or IGF1 level below -2 SDS for Subjects their age were invited for follow-up every 4–6 months for a year. The study was approved by the Institutional Review Board of Sheba Medical Center. 3. Patients with signs, symptoms or basal hormonal levels suggestive for pituitary dysfunction underwent dynamic Medical records of all children admitted to the Edmond testing for evaluation of pituitary function. Only one and Lily Safra Children’s Hospital, a tertiary medical participant had an indication for GH stimulation test center, due to infectious diseases of the CNS (meningitis or according to these criteria. Glucagon test was performed encephalitis) between 2007 and 2010 were reviewed. (0.03 mg/kg intramuscular) with blood sampling for GH Patients of both sexes aged 0–16 years with infectious and cortisol at baseline, 30, 60, 90, 120 and 180 min. GH disease of the CNS and an interval of at least 6 months level below 7.5 mcg/l was considered as suggestive of between the acute disease and the evaluation qualified for deficiency. inclusion in the study. The diagnosis of meningitis was confirmed by clinical findings, cerebrospinal fluid (CSF) Growth assessment analysis, and blood and CSF cultures. Exclusion criteria were any prior severe chronic disorder, or any traumatic Patients were measured using a wall mounted stadiometer. brain injury. Height standard deviation scores (SDS) were calculated using age and gender-specific growth data (based on the A total of 79 records were reviewed. Seventy-six patients Centers for Disease Control and Prevention’s Year 2000 were diagnosed with meningitis, and 3 with meningoen- Growth Charts) (www.cdc.gov/growthcharts). These data cephalitis. In all patients, an elevation of protein concen- have been found adequate for assessing Israeli children trations and pleocytosis of CSF confirmed the diagnosis of [17]. meningitis. The diagnosis of bacterial or viral meningitis was 123 133

120 Pituitary (2014) 17:118–124 Table 1 Clinical features Hormone Clinical features suggestive of hypopituitarism that were inquired at the first Corticotrophin deficiency Fatigue, pallor, anorexia, weight loss, failure to thrive study visit Thyrotropin deficiency Tiredness, cold intolerance, constipation, dry skin, retarded development, Gonadotropin deficiency growth retardation GH deficiency Delayed puberty Antidiuretic hormone Growth retardation, decreased muscle mass and strength, obesity, fatigue Polyuria, polydipsia deficiency Target height using one-way analysis of variance followed by multiple comparisons using the Bonferroni correction method. A two- Genetic target heights were calculated by averaging the sided p value\\0.05 was considered statistically significant. heights of the parents, as reported by them, and adjusting by adding 6.5 cm for male patients and subtracting 6.5 cm Results for females. Target heights were expressed as the HSDS (z-score) using the same reference values [18]. Target Fourteen persons (9 males, 5 females) were included in the height was considered as the child’s ‘‘genetic potential’’ study. Their age at diagnosis of meningitis was between that would have been achieved if the disease and/or the 2 weeks and 15.8 years (mean 3.8 ± 5.4 years). Three therapy had not influenced growth. patients had bacterial meningitis. Five patients needed treatment in intensive care unit (ICU). Time interval from Growth velocity patterns diagnosis until the study ranged from 0.7 to 4.3 years (mean 2.6 ± 0.3 years). Comparing the study group to the Growth velocity was determined by linear growth starting 41 children who were only interviewed by phone and the at the first evaluation and continuing until the last recorded 24 who were lost to follow-up revealed no significant visit. differences regarding the following criteria: male to female ratio, age at meningitis, ICU admission, duration of hos- Basal hormonal evaluation included: thyrotropin (TSH, pitalization and time interval from meningitis until the normal range 0.4–5 mIU/l), free thyroxin (fT4, normal range study. Bacterial meningitis was more frequent in the study 7–16 pmol/l) and cortisol (normal range 138–690 pmol/l). group. Insulin-like growth factor-1 (IGF1) levels were transformed to natural logarithm (ln) in order to achieve normal distri- The characteristics of the study group during the first bution. Standard deviation scores for each subject were study visit, including: auxologic parameters, Tanner stage calculated as explained elsewhere [19]. Blood samples were and basal hormone levels are presented in Tables 2 and 3. obtained between 08:00 and 09:00 am after an overnight fast. Detailed medical history revealed no symptoms suspected for corticotrophin, thyrotropin, gonadotropin or antidiuretic TSH and fT4 were measured by UniCelTM DxI 800 hormone deficiency. All of them had normal levels of basal Access Immunoassay System (Beckman Coulter Inc., Brea, cortisol, TSH and fT4. Regarding the gonadotrophic axis, CA, USA). IGF-I and cortisol levels were measured by a 12 children were pre-pubertal at the time of the study (age chemiluminescent method (Immulite 2000, Diagnostic 1.1–9.6 years). No participant had precocious puberty. Products Corporation, Los Angeles, CA, USA). Therefore, measurement of FSH and LH levels was unwarranted. Two male patients were aged 14–15 years at Follicle-stimulating hormone (FSH) and luteinizing meningitis diagnosis and 19–20 years during the study. hormone (LH) levels were not included since 12 of the Physical examination showed Tanner 5 in both of them children were pre-pubertal and the other two participants suggesting normal gonadotropin function. have successfully completed puberty and were at Tanner stage 5 at the time of evaluation. Three participants had IGF1 level below -2 SDS (patients’ no. 7, 9 and 11 in Table 3). Therefore, they Statistical analyses were invited for follow-up visits every 4–6 months dur- ing a year. Two of them (patients no. 7 and 11) had Data were analyzed using SAS software, version 9.1 (SAS normal height z-scores and a normal growth velocity. Institute, Cary, NC, USA). Continuous variables were Only one child had height SDS below -2, as well as low expressed as the mean value ± standard deviation (SD). Comparisons between the study groups were performed 123 134

Pituitary (2014) 17:118–124 121 Table 2 Baseline characteristics and basal hormonal level at the time of first study visit Subjects Age at meningitis Gender Diagnosis Duration after TSH (0.4–5 FT4 Cortisol (years) meningitis (years) mIU/l) (7–16 pmol/l) (138–690 nmol/l) 1 5.96 Female ABM 3.0 1.6 12.3 292 2 0.06 Female AVM 1.7 1.0 9.1 265 3 0.03 Male AVM 2.8 2.9 12.8 566 4 6.30 Female AVM 2.9 3.5 15.3 648 5 0.26 Male AVM 1.5 2.1 11.3 150 6 2.14 Male ABM 4.1 1.0 13.2 227 7 0.11 Male AVM 2.0 2.6 10.8 212 8 14.86 Male Part. 4.3 4.4 12.8 458 9 0.72 Male Part. 2.7 3.4 12.8 268 10 5.55 Female AVM 4.1 1.6 13.4 364 11 0.46 Male Part. 0.7 2.3 11.6 439 12 15.83 Male ABM 4.3 2.1 11.3 287 13 0.03 Female AVM 1.9 2.4 10.4 251 14 0.42 Male AVM 1.4 1.7 10.1 232 ABM acute bacterial meningitis, AVM acute viral meningitis, Part. partially treated meningitis Table 3 Growth parameters and Tanner stage at the time of first study visit MPH SDS Delta SDSa IGF1 SDS Subjects Age (years) Gender Tanner stage Height SDS Weight SDS 0.29 0.45 0.17 1 9.0 Female 1 0.74 1.31 0.75 -1.43 -1.95 2 1.8 Female 1 -0.68 -0.15 1.25 3 2.8 Male 1 -1.1 0.65 0.09 4 9.2 Female 1 1.9 2.68 1.39 0.71 -0.93 5 1.7 Male 1 -0.39 1.01 -1.13 -1.0 6 6.3 Male 1 1.42 0.26 1.04 0.88 7 2.1 Male 1 0.39 0.04 -0.16 -0.3 -1.25 8 19.1 Male 5 -0.09 -0.66 0.54 0 -4.01 9 3.5 Male 1 -0.04 0.49 -0.48 -2.73 10 9.6 Female 1 -0.16 -3.62 -0.16 -0.54 2.44 11 1.1 Male 1 -2.19 -2.01 1.25 -0.42 -2.80 12 20.1 Male 5 -1.02 -2.26 1.22 -0.40 -1.00 13 1.9 Female 1 -0.58 -0.39 -1.13 -1.96 -4.17 14 1.8 Male 1 0.97 -0.14 0.85 -1.06 3.50 -0.74 -0.87 -1.27 -1.90 MPH mid parental height a Delta SDS = Height SDS - MPH SDS levels of IGF1 and decreased growth velocity and was Discussion considerably shorter than expected from his midparental height (patient no. 9, Table 3). GH stimulation test Childhood and adolescence are life periods in which (glucagon test) revealed peak GH level of 12.3 mcg/l, hormones regulate the important functions of growth and excluding GH deficiency (GH levels of 4.2, 1.1, 12.3, pubertal development. Therefore, pituitary dysfunction 6.2, 2.7 and 2.3 mcg/l at baseline, 60, 90, 120, 150 and after meningitis might have serious implications. In the 180 min, respectively). It also revealed normal peak current study, 14 children were evaluated for hypopitu- cortisol level of 646 nmol/l (cortisol levels of 218, 240, itarism after recovering from acute meningitis. None of 309, 574, 590 and 646 nmol/l at baseline, 60, 90, 120, them developed pituitary hormone deficiencies. 150 and 180 min, respectively). This patient was born small for gestational age (birth weight 550 g at gesta- Corticotropic axis: Corticotrophin and cortisol secretion tional week 27). follow a diurnal rhythm, with the highest levels in the early morning and the lowest concentrations around midnight. 123 135

122 Pituitary (2014) 17:118–124 Corticotropin deficiency can be excluded at morning cor- assessment. However, since water deprivation tests were tisol concentrations greater than 500 nmol/l and is diag- not performed, partial posterior insufficiency might have nosed at levels of less than 100 nmol/l [20]. Levels been missed. Nevertheless, our findings are in accordance between these values warrant a stimulation test if there are with data from recent studies, none of which implement clinical features suspicious for the condition. None of our CNS infections in the development of diabetes insipidus patients had a cortisol level less than 100 nmol/l, or any [6–9]. clinical features suggestive of adrenal insufficiency. To our knowledge, no previous study on pituitary dys- Thyrotropic axis: Central hypothyroidism is diagnosed function following acute infectious meningitis has been when concentrations of free thyroxin are decreased and conducted in children. However, 4 systematic studies have TSH amounts are low to slightly elevated. All patients had been performed in adults. In the retrospective study by both free thyroxin and TSH levels within the normal range. Schaefer et al. [6], 19 patients with a previous viral or bacterial meningitis/encephalitis have been evaluated at a Gonadotropic axis: Hypogonadism in childhood causes mean of 26 months (range 10–56 months) after the acute no clinical symptoms until the onset of puberty, at which event. In that group, 4 patients had isolated corticotropin time it usually presents with delayed or missing onset of deficiency (peak cortisol \\181.25 mcg/l during insulin puberty. Twelve children were prepubertal at the time of tolerance test) and two patients showed borderline gonad- the study (age 1.1–9.6 years), and measuring FSH and LH otropic insufficiency (basal testosterone between 2.4 and 3 levels was thus not indicated. Two male patients were aged mcg/l). No patient had somatotropic or thyrotropic insuf- 14–15 years at meningitis diagnosis and 19–20 years dur- ficiency or evidence for diabetes insipidus. In the other ing the study. Both completed normal puberty, and at the retrospective preliminary study, pituitary functions of 14 time of study their Tanner stage was 5. patients with acute bacterial or viral meningitis were investigated [7]. The mean evaluation time was 20 months Somatotropic axis: In the evaluation of growth in chil- after acute meningitis (range 6–48 months) and 4 of 14 dren there are basically three parameters that can be patients had isolated GH deficiency proven by assessed. First, height can be compared with age references GHRH ? arginine stimulation test. The other pituitary and expressed as standard deviation score [21]. Second, axes were considered normal based on basal hormone height SDS can be compared with the sex corrected mid- levels. parental height (MPH). Third, a longitudinal analysis of growth should be conducted and expressed as height In a recent study, Tsiakalos et al. [8] demonstrated velocity. After the auxological evaluation, the most com- substantial risk of hypopituitarism during the acute phase monly used screening test to evaluate growth hormone (first 24 h) and at 12 months after acute viral and bacterial deficiency is IGF1 levels [22, 23]. In our study only one meningitis. At 12 months, based on basal hormone levels subject had height SDS below -2, as well as a considerable and insulin tolerance test, 5 of 16 patients had at least one difference between his height and the MPH, and low levels pituitary hormone deficiency (2 had isolated ACTH defi- of IGF1 (no. 9 in Table 3). Therefore, growth hormone ciency, 1 had isolated GH deficiency and 2 combined stimulation test was performed, giving normal results. His GH ? ACTH deficiency). In another prospective study, short stature can probably be attributed to the fact he was Tanriverdi et al. [9] checked basal hormone levels and GH born small for gestational age. Two additional patients had and cortisol levels after glucagon stimulation test during low IGF1 levels but normal height z-scores (no. 7 and 11 in the acute phase, at 6 and at 12 months in 16 patients. They Table 3) that were compatible with their genetic target found that at 12 months after meningitis 4 had isolated GH height z-scores, and their growth velocity during 1 year of deficiency, 1 combined GH ? ACTH deficiency and 1 follow-up was normal, making GH deficiency unlikely. combined GH ? FSH/LH deficiency. IGF1 has limited sensitivity because of significant overlap with normal values, mainly in children younger than Tanriverdi and Tsiakalos in their prospective studies 5 years of age [24]. About 50 % of low levels of IGF1 are evaluated pituitary function both during the acute phase and not associated with GH deficiency, but with low energy 12 months after. They found that pituitary deficiencies in the intake and less commonly GH receptor and post receptor acute phase may recover in a proportion of patients, and defects [25]. In a recent study measurement of height deficiencies may also develop during the subsequent follow- velocity was found to be a useful marker for monitoring up period. Two prospective studies focused on evaluating the pituitary function in children who had traumatic brain pituitary function at the time of presentation [31, 32]. Both injury [26]. studies found abnormalities in the pituitary hormone profile during the acute infection. Hyperprolactinemia was the most ADH deficiency: Posterior pituitary deficiency after CNS common, followed by cortisol deficiency. infection has been reported in a limited number of case reports [27–30]. In our study, we found no evidence of The mechanisms of pituitary dysfunction after acute posterior pituitary deficiencies based on clinical meningitis are not clearly defined. The pituitary hormone 123 136

Pituitary (2014) 17:118–124 123 deficiency pattern is suggestive of a dynamic ongoing prospective investigations in children need to be carried process. Tanriverdi et al. [9] demonstrated the occurrence out. Meanwhile, routine follow-up of growth and puberty of anti-pituitary antibodies (APA) and anti-hypothalamus of children post-meningitis is required by the primary care antibodies (AHA) for the first time during the acute phase physician. Invasive assessments should be reserved for of the meningitis and 6 and 12 months later. The frequency selected cases where there is slow growth or other clinical of antibodies positivity ranged from 35 to 50 % of the suspicion of hypopituitarism. patients throughout the 12 month period. The presence of a substantially high frequency of AHA and APA after acute Conflict of interest Nothing to declare. meningitis suggests that an autoimmune hypothalamic- pituitary process could be triggered by acute meningitis. References However, when the relation between pituitary dysfunction and the presence of the antibodies at any time during 1. Rose SR, Auble BA (2012) Endocrine changes after pediatric follow-up period was analyzed, no significant correlation traumatic brain injury. Pituitary 15:267–275 was found. The lack of correlation was assumed to be due to the short duration of the prospective follow-up. 2. Schneider HJ, Kreitschmann-Andermahr I, Ghigo E, Stalla GK, Agha A (2007) Hypothalamopituitary dysfunction following One limitation of our study is the small group of children traumatic brain injury and aneurysmal subarachnoid hemorrhage: that underwent the endocrine evaluation. However, during a systematic review. JAMA 298:1429–1438 the recruitment, 55 families were approached by letters and telephone calls and were asked specifically about signs and 3. Einaudi S, Bondone C (2007) The effects of head trauma on symptoms of pituitary dysfunction. Many parents declined to hypothalamic-pituitary function in children and adolescents. Curr take part in our study as they felt that their children are Opin Pediatr 19:465–470 completely healthy. The probable selection bias is towards children who would be expected to be at greater risk of 4. Ghigo E, Masel B, Aimaretti G, Leon-Carrion J, Casanueva FF, hypopituitarism. Another limitation was the fact that we did Dominguez-Morales MR, Elovic E, Perrone K, Stalla G, Thomp- not use stimulation tests to evaluate pituitary function in the son C, Urban R (2005) Consensus guidelines on screening for entire group but only when it was indicated in contrast to the hypopituitarism following traumatic brain injury. Brain Inj studies that were done in the adult population. However, in 19:711–724 pediatric patients, the diagnosis of growth hormone defi- ciency relies on comprehensive clinical and auxological 5. Acerini CL, Tasker RC (2007) Traumatic brain injury induced assessment, followed by biochemical tests of the GH-IGF-1 hypothalamic-pituitary dysfunction: a paediatric perspective. axis. GH stimulation tests are typically performed only in Pituitary 10:373–380 children with slow growth, whose history and auxology suggest GH deficiency [33]. According to the guidelines, 6. Schaefer S, Boegershausen N, Meyer S, Ivan D, Schepelmann K, only 1 child required stimulation test for investigation of the Kann PH (2008) Hypothalamic-pituitary insufficiency following GH axis. None required other stimulation test after evalua- infectious diseases of the central nervous system. Eur J Endo- tion that included detailed history, physical examination and crinol 158:3–9 basal hormones levels. Moreover, reliance only on stimula- tion tests also has limitations. It is common for normally 7. Tanriverdi F, Alp E, Demiraslan H, Dokmetas HS, Unluhizarci growing children to display low GH responses that falsely K, Doganay M, Casanueva FF, Kelestimur F (2008) Investigation suggest deficiency [34]. In addition, GH tests are poorly of pituitary functions in patients with acute meningitis: a pilot reproducible; individuals who undergo repeated tests have study. J Endocrinol Invest 31:489–491 highly variable results [35]. Consistent with our approach, Heather et al. [36] recently concluded that relying on pro- 8. Tsiakalos A, Xynos ID, Sipsas NV, Kaltsas G (2010) Pituitary vocative GH tests alone might have incorrectly diagnosed insufficiency after infectious meningitis: a prospective study. with GH deficiency up to 33 % of their pediatric patients J Clin Endocrinol Metab 95:3277–3281 with TBI. Similar problems are evident also in Synacten test evaluating ACTH deficiency [36]. 9. Tanriverdi F, De Bellis A, Teksahin H, Alp E, Bizzarro A, Sinisi AA, Bellastella G, Paglionico VA, Bellastella A, Unluhizarci K, In conclusion, based on the present study, in contrast to Doganay M, Kelestimur F (2012) Prospective investigation of reports in adult patients, pituitary dysfunction with overt pituitary functions in patients with acute infectious meningitis: is clinical symptoms is not a frequent consequence of acute acute meningitis induced pituitary dysfunction associated with meningitis in children. It is possible that partial hypopitu- autoimmunity? Pituitary 15(4):579–588 itarism can be found to some extent after stimulation test, as has been found in adult studies. Further long-term 10. Haslam RH, Winternitz WW, Howieson J (1969) Selective hypopituitarism following tuberculous meningitis. Am J Dis Child 118:903–908 11. Lam KS, Sham MM, Tam SC, Ng MM, Ma HT (1993) Hypo- pituitarism after tuberculous meningitis in childhood. Ann Intern Med 118:701–706 12. Pai KG, Rubin HM, Wedemeyer PP, Linarelli LG (1976) Hypo- thalamic-pituitary dysfunction following group b beta hemolytic streptococcal meningitis in a neonate. J Pediatr 88:289–291 13. Hanna CE, LaFranchi SH (1983) Evolving hypopituitarism in children with central nervous system lesions. Pediatrics 72:65–70 14. Theodoridou MN, Vasilopoulou VA, Atsali EE, Pangalis AM, Mostrou GJ, Syriopoulou VP, Hadjichristodoulou CS (2007) Meningitis registry of hospitalized cases in children: epidemio- logical patterns of acute bacterial meningitis throughout a 32-year period. BMC Infect Dis 7:101 15. Logan SA, MacMahon E (2008) Viral meningitis. BMJ 336:36–40 123 137

124 Pituitary (2014) 17:118–124 16. Tanner JM, Whitehouse RH (1976) Clinical longitudinal stan- 28. Lee YJ, Yang D, Shyur SD, Chiu NC (1995) Neurogenic diabetes dards for height, weight, height velocity, weight velocity, and insipidus in a child with fatal coxsackie virus b1 encephalitis. stages of puberty. Arch Dis Child 51:170–179 JPEM 8:301–304 17. Goldstein A, Haelyon U, Krolik E, Sack J (2001) Comparison of 29. Madhoun ZT, DuBois DB, Rosenthal J, Findlay JC, Aron DC body weight and height of israeli schoolchildren with the tanner (1991) Central diabetes insipidus: a complication of herpes and centers for disease control and prevention growth charts. simplex type 2 encephalitis in a patient with aids. Am J Med Pediatrics 108:E108 90:658–659 18. Lee MM (2006) Clinical practice. Idiopathic short stature. N Engl 30. Woredekal Y (1998) Recurrent central diabetes insipidus sec- J Med 354:2576–2582 ondary to cryptococcal meningitis. J Assoc Acad Minor Phys 9:22–24 19. Ranke MB, Schweizer R, Elmlinger MW, Weber K, Binder G, Schwarze CP, Wollmann HA (2000) Significance of basal igf-i, 31. Dhanwal DK, Vyas A, Sharma A, Saxena A (2010) Hypotha- igfbp-3 and igfbp-2 measurements in the diagnostics of short lamic pituitary abnormalities in tubercular meningitis at the time stature in children. Horm Res 54:60–68 of diagnosis. Pituitary 13:304–310 20. Schneider HJ, Aimaretti G, Kreitschmann-Andermahr I, Stalla 32. Dhanwal DK, Kumar S, Vyas A, Saxena A (2011) Hypothalamic GK, Ghigo E (2007) Hypopituitarism. Lancet 369:1461–1470 pituitary dysfunction in acute nonmycobacterial infections of central nervous system. Indian J Endocrinol Metab 15(Suppl 21. Oostdijk W, Grote FK, de Muinck Keizer-Schrama SM, Wit JM 3):S233–S237 (2009) Diagnostic approach in children with short stature. Horm Res 72:206–217 33. Consensus guidelines for the diagnosis and treatment of growth hormone (gh) deficiency in childhood and adolescence: Summary 22. Ranke MB, Schweizer R, Lindberg A, Price DA, Reiter EO, statement of the gh research society. Gh research society. J Clin Albertsson-Wikland K, Darendeliler F (2004) Insulin-like growth Endocrinol Metab 2000 85:3990–3993 factors as diagnostic tools in growth hormone deficiency during childhood and adolescence: the kigs experience. Horm Res 34. Ghigo E, Bellone J, Aimaretti G, Bellone S, Loche S, Cappa M, 62(Suppl 1):17–25 Bartolotta E, Dammacco F, Camanni F (1996) Reliability of provocative tests to assess growth hormone secretory status. 23. Jensen RB, Jeppesen KA, Vielwerth S, Michaelsen KF, Main Study in 472 normally growing children. J Clin Endocrinol Metab KM, Skakkebaek NE, Juul A (2005) Insulin-like growth factor i 81:3323–3327 (igf-i) and igf-binding protein 3 as diagnostic markers of growth hormone deficiency in infancy. Horm Res 63:15–21 35. Tassoni P, Cacciari E, Cau M, Colli C, Tosi M, Zucchini S, Cicognani A, Pirazzoli P, Salardi S, Balsamo A et al (1990) 24. Strasburger CJ, Bidlingmaier M (2005) How robust are labora- Variability of growth hormone response to pharmacological and tory measures of growth hormone status? Horm Res 64(Suppl sleep tests performed twice in short children. J Clin Endocrinol 2):1–5 Metab 71:230–234 25. Richmond EJ, Rogol AD (2008) Growth hormone deficiency in 36. Heather NL, Jefferies C, Hofman PL, Derraik JG, Brennan C, children. Pituitary 11:115–120 Kelly P, Hamill JK, Jones RG, Rowe DL, Cutfield WS (2012) Permanent hypopituitarism is rare after structural traumatic brain 26. Bellone S, Einaudi S, Caputo M, Prodam F, Busti A, Belcastro S, injury in early childhood. J Clin Endocrinol Metab 97:599–604 Parlamento S, Zavattaro M, Verna F, Bondone C, Tessaris D, Gasco V, Bona G, Aimaretti G (2012) Measurement of height velocity is an useful marker for monitoring pituitary function in patients who had traumatic brain injury. Pituitary 2012 Nov 23 27. Mena W, Royal S, Pass RF, Whitley RJ, Philips JB III (1993) Diabetes insipidus associated with symptomatic congenital cytomegalovirus infection. J Pediatr 122:911–913 123 138

‫‪Regional apparent diffusion coefficient values‬‬ ‫‪in 3rd trimester fetal brain‬‬ ‫‪Neuroradiology | 2014‬‬ ‫מנחה‪ :‬פרופ' אלדד קטורזה‬ ‫מנחה‪ :‬דר' חן הופמן‬ ‫דימות העובר‪ ,‬יילוד וגינקולוגיה‬ ‫דימות‬ ‫מנהל פרויקט ח\"ץ‬ ‫מנהל מכון גרטנר‬ ‫‪[email protected]‬‬ ‫דפי ברגמן‬ ‫אונ' תל אביב‬ ‫השתתפה כסטודנטית בפרויקט ח“ץ‬ ‫בין השנים ‪2011-2014‬‬ ‫‪[email protected]‬‬ ‫‪139‬‬

Neuroradiology (2014) 56:561–567 DOI 10.1007/s00234-014-1359-6 PAEDIATRIC NEURORADIOLOGY Regional apparent diffusion coefficient values in 3rd trimester fetal brain Chen Hoffmann & Boaz Weisz & Shlomo Lipitz & Gal Yaniv & Eldad Katorza & Dafi Bergman & Anat Biegon Received: 25 August 2013 / Accepted: 30 March 2014 / Published online: 20 April 2014 # Springer-Verlag Berlin Heidelberg 2014 Abstract did not significantly change with gestational age in any of the Introduction Apparent diffusion coefficient (ADC) values in the regions tested. In the four cases with fetal CMV infection, developing fetus can be used in the diagnosis and prognosis of ADC value was associated with a global decrease. prenatal brain pathologies. To this end, we measured regional Conclusion ADC values in normal fetal brain are relatively ADC in a relatively large cohort of normal fetal brains in utero. stable during the third trimester, show consistent regional Methods Diffusion-weighted imaging (DWI) was performed variation, and can make an important contribution to the early in 48 non-sedated 3rd trimester fetuses with normal structural diagnosis and possibly prognosis of fetal brain pathologies. MR imaging results. ADC was measured in white matter (frontal, parietal, temporal, and occipital lobes), basal ganglia, Keywords Fetus . Brain . MRI . DWI . ADC thalamus, pons, and cerebellum. Regional ADC values were compared by one-way ANOVA with gestational age as covar- Introduction iate. Regression analysis was used to examine gestational age- related changes in regional ADC. Four other cases of CMV Diffusion-weighted imaging (DWI) is gaining ground in devel- infection were also examined. opmental studies of the human brain in vivo, specifically with Results Median gestational age was 32 weeks (range, 26– regard to myelination [1]. Maturational changes in diffusion 33 weeks). There was a highly significant effect of region on indices in healthy fetuses are thought to be related to the devel- ADC, whereby ADC values were highest in white matter, opment of the myelin sheath [2–7], which proceeds from primary with significantly lower values in basal ganglia and cerebel- cortices to secondary cortices and to associative cortices. These lum and the lowest values in thalamus and pons. ADC processes have been intensely studied in postmortem tissue [8]. There is a growing need for a comprehensive in vivo database of Chen Hoffmann and Boaz Weisz contributed equally to the manuscript. apparent diffusion coefficient (ADC) values in different brain regions of the developing fetus, which can be used in the C. Hoffmann : G. Yaniv : D. Bergman diagnosis and prognosis of prenatal brain pathologies. Department of Radiology, Sheba Medical Center, Tel Hashomer In the present study, we measured ADC in several brain (affiliated to the Sackler School of Medicine), Tel Aviv University, regions of a relatively large cohort (N=48) of non-sedated Tel Aviv, Israel healthy fetuses and compared them to four cases of congenital CMV infection scanned at similar gestational ages. B. Weisz : S. Lipitz : E. Katorza Subjects and methods Department of Obstetrics and Gynecology, Sheba Medical Center, Tel Hashomer (affiliated to the Sackler School of Medicine), Tel Aviv This is a prospective cohort study performed to establish University, Tel Aviv, Israel normal values of the ADC of normal fetuses in relation to gestational age (GA). A. Biegon Department of Neurology, Stony Brook University School of Institutional review board approval was given. Medicine, Stony Brook, NY, USA C. Hoffmann (*) Diagnostic Imaging, Sheba Medical Center, 52621, Tel Hashomer, Israel e-mail: [email protected] 140

562 Neuroradiology (2014) 56:561–567 Table 1 Subject population demographics: gestational age and grounds for referral MRI findings Serial no. GA (weeks) Reason for referral Normal MRI 1 28 Follow-up scan Normal MRI 2 32 Follow-up scan Normal pituitary gland, normal MRI 3 31 Pituitary agenesis in previous pregnancy (US normal) Normal pituitary gland, normal MRI 4 32 Pituitary agenesis in previous pregnancy (US normal) Normal pituitary gland, normal MRI 5 32 Pituitary agenesis in previous pregnancy (US normal) Normal MRI 6 28 Previous pregnancy terminated due to brain malformations (US normal) Normal MRI 7 30 Previous pregnancy twins with germinal matrix bleeding (US normal) Normal MRI 8 32 Previous pregnancy terminated due to malformations (US normal) Normal MRI 9 33 Coloboma in previous pregnancy (US normal) Normal MRI 10 33 Two previous pregnancies with brain malformations (US normal) Normal MRI 11 28 Sibling with schizencephaly (US normal) Normal MRI 12 29 Sibling with hemangiomatosis (US normal) Normal MRI 13 29 Sibling with CP (US normal) Normal MRI 14 30 Sibling with tuberous sclerosis (US normal) Normal MRI 15 31 Sibling with melanosis (US normal) Normal MRI 16 31 Sibling with neurological deficit (US normal) Normal MRI 17 32 Sibling with holoprosencephaly (US normal) Normal MRI 18 32 Sibling with neurological deficits (US normal) Normal MRI 19 32 Sibling with white matter disease (US normal) Normal MRI 20 32 Sibling with neurological retardation (US normal) Normal MRI 21 32 Sibling with brain malformations (US normal) Normal MRI 22 33 Sibling with AVM (US normal) Normal MRI 23 33 Sibling with brain malformation (US normal) Normal MRI 24 33 Sibling with white matter disease (US normal) Normal MRI 25 37 Sibling with holoprosencephaly (US normal) Normal MRI 26 33 Brain malformation in the family (US normal) Normal MRI 27 32 Father with holoprosencephaly (US normal) Normal vermis, normal MRI 28 30 Suspected small vermis Normal vermis, normal MRI 29 32 Suspected vermis/choroid plexus cyst? Normal vermis, normal MRI 30 32 Suspected small vermis? Choroid plexus cyst? Normal vermis, normal MRI 31 33 Suspected vermian agenesis Normal vermis, normal MRI 32 32 Suspected vermian agenesis Normal MRI 33 33 Suspected cystic structure in the posterior fossa Normal posterior fossa, normal MRI 34 26 Enlarged posterior fossa Normal posterior fossa, normal MRI 35 32 Suspected cystic structure in the posterior fossa Normal posterior fossa, normal MRI 36 30 Enlarged posterior fossa Normal corpus callosum, normal MRI 37 29 Difficult in visualization of corpus callosum Normal corpus callosum, normal MRI 38 29 Difficult in visualization of corpus callosum Normal corpus callosum, normal MRI 39 32 Difficult in visualization of corpus callosum Normal ventricular system, normal MRI 40 28 Narrow cavum septum pellucidum on US Normal ventricular system, normal MRI 41 31 Suspected abnormal cavum septum pellucidum Normal ventricular system, normal MRI 42 32 Suspected abnormal cavum septum pellucidum Normal ventricular system, normal MRI 43 33 Wide cavum septum pellucidum Normal MRI 44 30 Situs inversus? Normal sulcation, normal MRI 45 31 Unclear sulcation on third trimester US Normal MRI 46 27 US not diagnostic (technical) Normal MRI 47 32 Abdominal cyst Normal MRI 48 31 VSD Abnormal high T2 signal, temporal lobes 49 32 CMV infection abnormal MRI 141

Neuroradiology (2014) 56:561–567 563 Table 1 (continued) Reason for referral MRI findings Normal MRI Serial no. GA (weeks) CMV infection normal MRI Normal MRI CMV infection normal MRI Abnormal high T2 signal, temporal lobes 50 32 CMV infection abnormal MRI Periventricular cysts 51 33 52 30 Subjects Forty-eight normal fetuses, scanned in our diagnos- parietal, two temporal, two occipital), two on the basal gan- tic imaging department in the years 2006–2012, gestational glia, two on the thalami, and two on the cerebellar hemi- age (GA) from 26–33 weeks (mean GA 30 weeks, median spheres. A single ROI was placed over the pons. Bilateral 32 weeks) were included in the study. The indications for the ROIs were averaged, resulting in eight values/fetus, which MR scan were (1) normal ultrasound (US), but an abnormal were averaged for each gestational age. outcome in previous pregnancies in siblings (n=27); and (2) inconclusive findings in fetal neurosonography which were ADC measurements were also obtained in a subgroup of excluded by the MRI (N=21) and by subsequent US scans the CMV-infected fetuses previously described [9]. Four (Table 1). Fetuses were considered normal if they had normal CMV-infected fetuses (GA range 30–33 weeks) were com- prenatal ultrasonographic results (group 1) or suspected sono- pared to a subgroup (N=35) of the healthy fetuses in the same graphic findings, with subsequent normal fetal MR and US age range (i.e., fetuses <26 weeks old were not included in this imaging results (group 2). Maternal disease and pregnancy analysis) using two-way analysis of variance (diagnosis×re- complications as well as poor overall scan quality/motion gion). Two fetuses had a normal MR scan, one had an abnor- artifacts were used as exclusion criteria. In eight cases, not mal high signal in the white matter of the temporal lobes, in all the regional values were obtained due to technical the anterior aspect of the lobes, and another fetus presented difficulties. with multiple cystic periventricular lesions in the temporal and frontal lobes. The signal of the white matter of the temporal MRI According to our MRI protocol, the mothers refrained lobes was increased. There were no calcifications or other from eating or drinking fluids with sugar 4 h before the MRI findings in these brains on US. examination, which was performed without sedation. Scans were obtained using a 1.5 T MR system (GE Medial System, Results Milwaukee, WI). A single-shot fast spin echo (SSFSE) T2- weighted sequences in three orthogonal planes was employed The characteristics of the individual subjects are summarized with slice thickness of 3–4 mm, no gap, using a flexible coil in Table 1. Fetuses included in this study were in the third (eight channels cardiac coil). The field of view was deter- trimester, and the distribution across GA is shown in Fig. 1. mined by the size of the fetal head and was 24 cm for smaller ROI placement is illustrated in Fig. 2. As shown in Fig. 3, fetuses, up to 30 cm for the bigger fetuses. Other parameters changes in ADC were rather small over this relatively narrow were a matrix of 320/224, echo time (TE) of 90 ms, and gestational age range, and although a trend towards a decrease repetition time (TR) of 1,298 ms. A FSPGR T1 sequence in ADC with increasing GA was observed in most regions, it was performed only in the axial plane with a larger field of view of 40 cm, 4-mm slice thickness and .0.5-mm gap, and TR 160 and TE 2.3. Then, a DWI sequence in one to three orthogonal planes (axial coronal and sagittal planes) was performed with a 40-cm field of view, b value of 0 and 1,000 or 700 msec, slice thickness of 4 mm with no gap, and scan time of 56 s. The apparent diffusion coefficient (ADC) was calculated in each region of interest (ROI), using the functool software of the AD-HD work station (GE Medial System, Milwaukee, WI). A circular ROI was used, placed over the desired anatomical area. ADC measurements Fifteen circular regions of interest (ROIs) Fig. 1 Histogram of age distribution of normal fetuses included in the ranging in size from 62 to 104 mm2 (mean 83.22; SD 5.1) study were placed as follows: two on each lobe (two frontal, two 142

564 Neuroradiology (2014) 56:561–567 Fig. 2 Examples of ROI placement on diffusion-weighted images: 1, 2 frontal lobes; 3, 4 parietal lobes; 5, 6 basal ganglia; 7, 8 thalami; 9, 10 temporal lobes; 11, 12 occipital lobes; 13 brain stem; 14, 15 cerebellum was not statistically significant in any region. The only excep- two-way ANOVA. We found highly significant effects of tion was the frontal lobe, which exhibited a trend towards an diagnosis (F=10, p<0.003) and region (F=19, p<0.0001) increase, rather than decrease, in ADC with increasing GA, but no significant interaction term. The decreases ranged from but this trend also failed to reach statistical significance. 20 % in the parietal white matter to 10 % in the cerebellum. In view of this finding, mean ADC/region was calculated Finally, we have reanalyzed our data omitting the two and subjected to statistical analysis by one-way ANOVA normal cases that had siblings with white matter disease (cases (Fig. 4). ADC values were highest in the white matter, with #19 and 24, Table 1), since they may be at increased risk, but significantly (p<0.05) lower values in basal ganglia and cer- the MRI sequences used in our study are not diagnostic of ebellum and the lowest values in thalamus and pons. white matter disease in fetuses. The removal of these cases (one with GA of 32 weeks and one with GA of 33 weeks, Fetuses with confirmed CMV infection had significantly Table 1) had a negligible effect (less than 1 % change) on the decreased ADC throughout the brain (Fig. 5), as revealed by Fig. 3 ADC and gestational age in various brain regions 143

Neuroradiology (2014) 56:561–567 565 Fig. 4 Mean regional ADC in normal 3rd trimester fetuses. Bars repre- order based on the rate of maturation and myelination of these sent means±SEM of 40–48 cases/region. BG basal ganglia, Cb cerebel- regions and also related to the cellularity of the region, as lum, FWM frontal white matter, OccWM occipital white matter, ParWM documented by other techniques such as US and T2- and T1- parietal white matter, TempWM temporal white matter, Thal thalamus. weighted sequences, diffusion tensor imaging (DTI) fMRI, and *p<0.05 relative to all white matter regions, ANOVA followed by spectroscopy [10, 11]. Previous publications report decreases in Fisher’s PLSD post hoc analysis ADC in the majority of brain regions [1] (with the exception of the frontal lobe) during fetal and postnatal brain development, mean ADC values across regions and GA and did not affect while others [12] report significant GA-related decreases only in the statistical significance (or lack thereof) of the region and cerebellum, pons, and thalamus with no changes in other regions, GA dependence of ADC or the comparison between normal over the age range of 19–37 weeks. Our results show a weak and CMV-infected fetuses. trend for regional ADC decline in all regions which does not reach statistical significance. This apparent disagreement most Discussion likely reflects the relatively narrow gestational age range in our study relative to some previous publications and is related to the The results of this study, utilizing the largest cohort of 3rd fact that the white matter tracts are formed earlier, even before the trimester fetuses relative to the published literature, demonstrate 23rd week of gestation [13]. Thus, inspection of the data pre- that brain ADC is regionally variable but relatively stable sented by Schneider et al. [2, 14] reveals that the apparent between gestational weeks 26 and 33. Regional variation was decrease in ADC with gestational age is driven by the subgroup demonstrated by higher values in the white matter relative to the of fetuses scanned during the 2nd trimester (21–25 weeks GA), basal ganglia, cerebellum, thalamus, and pons. This regional which are not represented in our sample. This observation raises profile is similar to the one reported in other studies of fetal, the interesting possibility that regional myelination and other premature, and postnatal brains and follows the expected rank processes which reduce ADC are not a linear function of GA, but rather occur in “waves” with some developmental periods showing higher rates of ADC change than others. This interpre- tation is supported by data from histology and other methods for evaluation of brain maturation and myelination [15–17]. Petra et al. [18] shared the ADC values of term infants; the highest value was found in the anterior white matter. They also found that decrease in the ADC value allows detection of changes in response to brain injury. It is well-established that reduction in ADC values of various brain areas occurs following different Fig. 5 Effect of CMV infection on regional ADC in utero. Bars represent diagnosis (F=10, p<0.003) and region (F=19, p<0.0001) but no signif- means±SEM of 35 normal fetuses and 4 confirmed CMV cases matched icant interaction term for gestational age. ANOVA revealed highly significant effects of 144

566 Neuroradiology (2014) 56:561–567 neuropathologies such as ischemia [19, 20] and malignancies laid down in the 1964 Declaration of Helsinki and its later amendments. [21, 22]. Our small subgroup of CMV-infected fetuses also We declare that patient consent was waived in this study. presented with lowering of the ADC value, and the significance of this observation is still to be investigated. Conflict of interest We declare that we have no conflict of interest. In general, the absolute values of ADC we report are close to References those reported by others for fetal brain, suggesting that ADC measurements are relatively platform-independent. The coeffi- 1. Righini A, Bianchini E, Parazzini C, Gementi P, Ramenghi L, Baldoli cient of variance within 3rd trimester fetal brain regions did not C, Nicolini U, Mosca F, Triulzi F (2003) Apparent diffusion coeffi- exceed 15 %, which is conducive to using these data as a cient determination in normal fetal brain: a prenatal MR imaging normative data base. Viola et al. [23] compared the ADC value study. AJNR Am J Neuroradiol 24(5):799–804 of premature infants with fetuses with mean age of 37 weeks GA. They found differences between the two groups, indicating that 2. Schneider JF, Confort-Gouny S, Le Fur Y, Viout P, Bennathan M, MR data from the premature brain are not appropriate for the Chapon F, Fogliarini C, Cozzone P, Girard N (2007) Diffusion- assessment of the fetal brain maturation. weighted imaging in normal fetal brain maturation. Eur Radiol 17(9):2422–2429. doi:10.1007/s00330-007-0634-x We also report the first evaluation of ADC in CMV-infected fetuses in utero. Our finding of reduced ADC most likely reflects 3. Cannie M, De Keyzer F, Meersschaert J, Jani J, Lewi L, Deprest J, microgliosis which is the hallmark of neuroinflammation of all Dymarkowski S, Demaerel P (2007) A diffusion-weighted template causes [24]. Van der Voorn et al. [25] reported the ADC value for gestational age-related apparent diffusion coefficient values in the and the fractional anisotropy (FA) in the infected pediatric brain. developing fetal brain. Ultrasound Obstet Gynecol Off J Int Soc In their study, the ADC value was higher than the normal value, Ultrasound Obstet Gynecol 30(3):318–324. doi:10.1002/uog.4078 probably due to damage to the tissue. 4. Manganaro L, Perrone A, Savelli S, Di Maurizio M, Maggi C, The data presented here add to the previously published data Ballesio L, Porfiri LM, De Felice C, Marinoni E, Marini M (2007) on ADC values at various gestational ages, but represent the Evaluation of normal brain development by prenatal MR imaging. largest study of normal third trimester fetuses. Future research Radiol Med 112(3):444–455. doi:10.1007/s11547-007-0153-5 utilizing DWI in fetal neuroimaging should emphasize the early and accurate detection of perinatal brain abnormalities. 5. Huppi PS, Maier SE, Peled S, Zientara GP, Barnes PD, Jolesz FA, Volpe JJ (1998) Microstructural development of human newborn Limitations of the current study The normal fetuses included cerebral white matter assessed in vivo by diffusion tensor magnetic in the current study were considered healthy at the time of resonance imaging. Pediatr Res 44(4):584–590. doi:10.1203/ scanning based on several criteria including normal course of 00006450-199810000-00019 pregnancy, healthy mother, and no overt findings on T1-, T2-, and diffusion-weighted MRI. While these inclusion criteria 6. Neil JJ, Shiran SI, McKinstry RC, Schefft GL, Snyder AZ, Almli CR, are similar to those used in the majority of studies on normal Akbudak E, Aronovitz JA, Miller JP, Lee BC, Conturo TE (1998) fetal development of ADC (e.g., [12, 23]), they do not exclude Normal brain in human newborns: apparent diffusion coefficient and conditions for which these criteria are not diagnostic (e.g., diffusion anisotropy measured by using diffusion tensor MR imag- fetal white matter disease, incipient autism spectrum disorder). ing. Radiology 209(1):57–66. doi:10.1148/radiology.209.1.9769812 Subtle motion artifacts and occult diseases cannot be excluded in this study, so the adoption of the values from the “normal” 7. Tanner SF, Ramenghi LA, Ridgway JP, Berry E, Saysell MA, group as normative should be done with caution, and more Martinez D, Arthur RJ, Smith MA, Levene MI (2000) Quantitative studies need to be done with more extended postnatal follow- comparison of intrabrain diffusion in adults and preterm and term up to firm up a normative data base. neonates and infants. AJR Am J Roentgenol 174(6):1643–1649. doi: 10.2214/ajr.174.6.1741643 Conclusion Brain ADC is regionally variable but relatively stable between gestational weeks 26 and 33. Regional varia- 8. Miller SP, Vigneron DB, Henry RG, Bohland MA, Ceppi-Cozzio C, tion is demonstrated by higher values in the white matter Hoffman C, Newton N, Partridge JC, Ferriero DM, Barkovich AJ relative to the basal ganglia, cerebellum, thalamus, and pons. (2002) Serial quantitative diffusion tensor MRI of the premature Establishing the normal values of ADC enables us to evaluate brain: development in newborns with and without injury. J Magn and investigate cases with suspected pathology in the future. Reson Imaging JMRI 16(6):621–632. doi:10.1002/jmri.10205 Ethical standards and patient consent We declare that all human and 9. Hoffmann C, Grossman R, Bokov I, Lipitz S, Biegon A (2010) Effect animal studies have been approved by the Institutional Review Board and of cytomegalovirus infection on temporal lobe development in utero: have therefore been performed in accordance with the ethical standards quantitative MRI studies. Eur Neuropsychopharmacol J Eur Coll Neuropsychopharmacol 20(12):848–854. doi:10.1016/j.euroneuro. 2010.08.006 10. Mailath-Pokorny M, Kasprian G, Mitter C, Schopf V, Nemec U, Prayer D (2012) Magnetic resonance methods in fetal neurology. Semin Fetal Neonatal Med 17(5):278–284. doi:10.1016/j.siny.2012. 06.002 11. Maas LC, Mukherjee P, Carballido-Gamio J, Veeraraghavan S, Miller SP, Partridge SC, Henry RG, Barkovich AJ, Vigneron DB (2004) Early laminar organization of the human cerebrum demonstrated with diffusion tensor imaging in extremely premature infants. NeuroImage 22(3):1134–1140. doi:10.1016/j.neuroimage.2004.02.035 12. Boyer AC, Goncalves LF, Lee W, Shetty A, Holman A, Yeo L, Romero R (2013) Magnetic resonance diffusion-weighted imaging: reproducibility of regional apparent diffusion coefficients for the normal fetal brain. Ultrasound Obstet Gynecol Off J Int Soc Ultrasound Obstet Gynecol 41(2):190–197. doi:10.1002/uog.11219 145

Neuroradiology (2014) 56:561–567 567 13. Vasung L, Huang H, Jovanov-Milosevic N, Pletikos M, Mori S, Kostovic 20. Desai V, Shen Q, Duong TQ (2012) Incorporating ADC temporal I (2010) Development of axonal pathways in the human fetal fronto- profiles to predict ischemic tissue fate in acute stroke. Brain Res limbic brain: histochemical characterization and diffusion tensor imaging. 1458:86–92. doi:10.1016/j.brainres.2012.04.004 J Anat 217(4):400–417. doi:10.1111/j.1469-7580.2010.01260.x 21. Kralik SF, Taha A, Kamer AP, Cardinal JS, Seltman TA, Ho CY 14. Schneider MM, Berman JI, Baumer FM, Glass HC, Jeng S, Jeremy (2013) Diffusion imaging for tumor grading of supratentorial brain RJ, Esch M, Biran V, Barkovich AJ, Studholme C, Xu D, Glenn OA tumors in the first year of life. AJNR Am J Neuroradiol. doi:10.3174/ (2009) Normative apparent diffusion coefficient values in the devel- ajnr.A3757 oping fetal brain. AJNR Am J Neuroradiol 30(9):1799–1803. doi:10. 3174/ajnr.A1661 22. Barajas RF Jr, Rubenstein JL, Chang JS, Hwang J, Cha S (2010) Diffusion-weighted MR imaging derived apparent diffusion coeffi- 15. Ramenghi LA, Martinelli A, De Carli A, Brusati V, Mandia L, cient is predictive of clinical outcome in primary central nervous Fumagalli M, Triulzi F, Mosca F, Cetin I (2011) Cerebral maturation system lymphoma. AJNR Am J Neuroradiol 31(1):60–66. doi:10. in IUGR and appropriate for gestational age preterm babies. Reprod 3174/ajnr.A1750 Sci 18(5):469–475. doi:10.1177/1933719110388847 23. Viola A, Confort-Gouny S, Schneider JF, Le Fur Y, Viout P, 16. Weidenheim KM, Epshteyn I, Rashbaum WK, Lyman WD (1993) Chapon F, Pineau S, Cozzone PJ, Girard N (2011) Is brain Neuroanatomical localization of myelin basic protein in the late first maturation comparable in fetuses and premature neonates at and early second trimester human foetal spinal cord and brainstem. J term equivalent age? AJNR Am J Neuroradiol 32(8):1451– Neurocytol 22(7):507–516 1458. doi:10.3174/ajnr.A2555 17. Weidenheim KM, Bodhireddy SR, Rashbaum WK, Lyman WD 24. Carson MJ, Thrash JC, Walter B (2006) The cellular response in (1996) Temporal and spatial expression of major myelin proteins in neuroinflammation: the role of leukocytes, microglia and astrocytes the human fetal spinal cord during the second trimester. J in neuronal death and survival. Clin Neurosci Res 6(5):237–245. doi: Neuropathol Exp Neurol 55(6):734–745 10.1016/j.cnr.2006.09.004 18. Huppi PS, Dubois J (2006) Diffusion tensor imaging of brain devel- 25. van der Voorn JP, Pouwels PJ, Vermeulen RJ, Barkhof F, opment. Semin Fetal Neonatal Med 11(6):489–497. doi:10.1016/j. van der Knaap MS (2009) Quantitative MR imaging and siny.2006.07.006 spectroscopy in congenital cytomegalovirus infection and periventricular leukomalacia suggests a comparable neuro- 19. Marks MP, de Crespigny A, Lentz D, Enzmann DR, Albers GW, pathological substrate of the cerebral white matter lesions. Moseley ME (1996) Acute and chronic stroke: navigated spin-echo Neuropediatrics 40(4):168–173. doi:10.1055/s-0029- diffusion-weighted MR imaging. Radiology 199(2):403–408. doi:10. 1243228 1148/radiology.199.2.8668785 146

Anogenital distance in male and female fetuses at 20 to 35  weeks of gestation: centile charts and reference ranges Prenatal Diagnosis | 2014 ‫ פרופ' אלדד קטורזה‬:‫מנחה‬ ‫ פרופ' יואב ינון‬:‫מנחה‬ ‫ יילוד וגינקולוגיה‬,‫דימות העובר‬ ‫אולטרסאונד גינקולוגי‬ ‫מנהל פרויקט ח\"ץ‬ ‫מנהל מכון גרטנר‬ [email protected] [email protected] ‫פזית כהן‬-‫יסכה‬ ‫אונ' תל אביב‬ ‫השתתפה כסטודנטית בפרויקט ח“ץ‬ 2012-2013 ‫בין השנים‬ [email protected] 147

DOI: 10.1002/pd.4399 ORIGINAL ARTICLE Anogenital distance in male and female fetuses at 20 to 35 weeks of gestation: centile charts and reference ranges Yinon Gilboa1,2*, Zvi Kivilevitch1, Meri Oren1,2, Yisca Pazit Cohen1,2, Eldad Katorza1,2 and Reuven Achiron1,2 1Department of Obstetrics and Gynecology, Sheba Medical Center, Ramat Gan, Israel 2Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel *Correspondence to: Yinon Gilboa. E-mail: [email protected]; [email protected] ABSTRACT Objectives In pediatrics, anogenital distance (AGD) serves as a bioassay of normal phenotypic androgen action on the external genitalia. Recently, hypospadias and cryptorchidism were reported to be associated with reduced AGD. No normal reference ranges exist for fetuses throughout gestation. This study defines the normal centile chart of the AGD, in male and female fetuses, between 20 and 35 weeks of gestation. Methods Participants were low-risk pregnant women, 20 to 35 weeks of gestation. All fetuses were singleton, with normal anatomic scan and appropriate fetal biometry for gestational age. Fetal AGD was measured by transabdominal ultrasound. The perineum was assessed in the axial plane. The distance was measured from the center of the anus to the posterior convergence of the fourchette in female fetuses and to the posterior base of the scrotum in male fetuses. Statistical analysis included the construction of the normal, modeled, centile, and standard deviation range, for each gestational age. Intraobserver and interobserver variability was assessed by the interclass correlation and Bland–Altman plot. Results Adequate measurements were obtained for 218 female and 206 male fetuses. For 17 cases, measurements were not possible because of fetal lie. AGD increased linearly throughout gestational age (GA) (r2 = 0.808) for both sexes and was expressed by the following regression equation: for male fetuses, À12.348 + 1.075*GA, and for female fetuses, À3.179 + 0.513*GA, where GA is gestational age. The normal centiles, means, and standard deviations, per week, are presented. Conclusion AGD measurement in utero is feasible. These measurements assess the normality of the perineal region and may assist in the detection of genital anomalies. © 2014 John Wiley & Sons, Ltd. Funding sources: None Conflicts of interest: None declared INTRODUCTION reduced sperm production, and micropenis.7–11 Only few studies During fetal life, differentiation of the male external genitalia have evaluated AGD measurements in newborn human requires normal androgen production and adequate response infants.3,12 Eisenberg et al.13 showed that AGD is relatively of target tissues. The fetal anogenital distance (AGD), the constant during adult life. Although the fetal perineum region distance between the anus and the genitals, is androgen and external genitalia are the most sensitive regions for (dihydrotestosterone) dependent. Fetal AGD thus represents androgen influence, no data are available concerning the a sensitive marker for the effects of in utero exposure to feasibility of their assessment or the variability and normal range androgens and chemicals with antiandrogen effects.1,2 of the fetal AGD. This study was designed to address these issues. AGD is a sexually dimorphic trait; newborn male humans METHODS have a significantly greater AGD than do female humans.3 Study design In animal studies, measurement of AGD serves as a bioassay A cross-sectional, prospective study was conducted over a 1-year of fetal androgen activity.4,5 The Environmental Protection period. Eligibility criteria for the ultrasound assessment were Agency has designated AGD as a reproductive toxicity pregnant women at gestational weeks 20 to 35 with (1) a history endpoint in animal studies, according to which data are of regular menses and an accurate date of the beginning of the assessed for the human health risks incurred by endocrine last menstrual period, (2) gestational age ascertained by sono- disrupting chemicals.6 In humans and animals, a shortened graphic measurement of the crown-rump length in the first AGD was reported to be associated with a variety of male genital abnormalities including hypospadias, cryptorchidism, Prenatal Diagnosis 2014, 34, 946–951 © 2014 John Wiley & Sons, Ltd. 148