Middle cerebral artery peak systolic velocity to predict fetal hemoglobin levels in twin anemia polycythemia sequence

Transcription

1 Ultrasound Obstet Gynecol 2015; 46: Published online 7 September 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: /uog Middle cerebral artery peak systolic velocity to predict fetal hemoglobin levels in twin anemia polycythemia sequence F. SLAGHEKKE*, S. PASMAN, M. VEUJOZ, J. M. MIDDELDORP*, L. LEWI, R. DEVLIEGER, R. FAVRE, E. LOPRIORE and D. OEPKES* *Department of Obstetrics, Leiden University Medical Centre, Leiden, The Netherlands; Department of Obstetrics and Gynecology, University Hospitals KU Leuven, Leuven, Belgium; Department of Obstetrics, University Hospital CMCO-HUS, Strasbourg, France; Department of Neonatology, Leiden University Medical Centre, Leiden, The Netherlands KEYWORDS: Doppler; fetal anemia; fetal polycythemia; hemoglobin; MCA-PSV; monochorionic twins ABSTRACT Objective To evaluate the diagnostic accuracy of middle cerebral artery peak systolic velocity (MCA-PSV) Doppler measurements in prediction of hemoglobin levels in twin anemia polycythemia sequence (TAPS). Methods This study involved a consecutive cohort comprising monochorionic twin pregnancies complicated by TAPS managed at three European fetal medicine centers between 2005 and The accuracy of MCA-PSV, measured immediately prior to fetal hemoglobin (Hb) measurement by fetal or cord blood sampling, for prediction of anemia and polycythemia was assessed using 2 2 tables. Results A total of 116 measurements (74 recorded in donors and 42 in recipients) from 43 twin pregnancies complicated by TAPS were available for analysis. MCA-PSV multiples of the median (MoM) values correlated well with Hb levels (r = 0.86; P < 0.001). The sensitivity of MCA-PSV 1.5 MoM to predict severe anemia (Hb deficit > 5 SD below the mean) in TAPS donors was 94% (95% CI, 85 98%); specificity was 74% (95% CI, 62 83%); positive and negative predictive values were 76% (95% CI, 65 85%) and 94% (95% CI, 83 98%), respectively. The sensitivity of MCA-PSV 1.0 MoM to predict polycythemia (Hb level > 5 SD above the mean) in TAPS recipients was 97% (95% CI, 87 99%); specificity was 96% (95% CI, 89 99%); positive and negative predictive values were 93% (95% CI, 81 97%) and 99% (95% CI, %), respectively. Conclusion MCA-PSV measurement has high diagnostic accuracy for predicting abnormal Hb levels in fetuses with TAPS. Copyright 2015 ISUOG. Published by John Wiley & Sons Ltd. INTRODUCTION Twin anemia polycythemia sequence (TAPS) is caused by chronic intertwin blood transfusion through minuscule placental anastomoses, leading to a large intertwin hemoglobin (Hb) difference without signs of twin oligopolyhydramnios sequence (TOPS) 1. TAPS may occur spontaneously or following laser treatment for twin twin transfusion syndrome (TTTS), defined as postlaser TAPS. Antenatal diagnosis is based on Doppler ultrasound-detected abnormalities of an increased peak systolic velocity (PSV) in the middle cerebral artery (MCA) of the donor twin, suggestive of fetal anemia and decreased MCA velocities in the recipient twin, suggestive of polycythemia, without signs of TOPS. These findings are often accompanied by a distinct difference in placental echodensity 1. Postnatal diagnosis is based on an intertwin Hb difference of 8 g/dl, with at least one of the following: reticulocytosis in the donor with a reticulocyte count ratio 1.7; small anastomoses (< 1 mm) at the placental surface 2. Middle cerebral artery peak systolic velocity (MCA-PSV) measurement, a non-invasive test, has become the standard assessment for diagnosis of fetal anemia in singletons in a variety of fetal diseases 3. Normal ranges for MCA-PSV in monochorionic diamniotic twins have been reported previously by Klaritsch et al. 4. However, MCA-PSV measurements in monochorionic twin pregnancies complicated by TAPS have not been reported before. The aim of this study was to evaluate the diagnostic accuracy of MCA-PSV measurements in prediction of hemoglobin levels in pregnancies complicated by TAPS. METHODS All monochorionic twin pregnancies complicated by TAPS, in which MCA-PSV measurements were Correspondence to: Dr F. Slaghekke, Department of Obstetrics, Leiden University Medical Center, PO Box 9600, 2300 RC, Leiden, The Netherlands ( f.slaghekke@lumc.nl) Accepted: 5 June 2015 Copyright 2015 ISUOG. Published by John Wiley & Sons Ltd. ORIGINAL PAPER

2 MCA-PSV in TAPS 433 performed, followed by fetal or neonatal blood sampling for Hb measurement within 24 h, were included in this retrospective study. Fetal blood sampling was performed routinely in cases of TAPS managed with intrauterine blood transfusion or partial exchange transfusion and postnatal blood sampling for Hb measurements was performed routinely at birth in all cases of TAPS. Postnatal blood sampling was carried out after cord clamping and delivery of the placenta. Cord blood samples, obtained from the umbilical vein, were not compared to neonatal blood samples. The study cohort consisted of consecutive cases of TAPS managed between 2005 and 2013 at three European fetal medicine centers (Leiden University Medical Center, The Netherlands; University Hospitals KU Leuven, Belgium; and University Hospital, Centre Medico Chirurgical Obstetrical (CMCO), Strasbourg, France). TAPS was identified using criteria published previously 1. Briefly, antenatal TAPS was diagnosed when Doppler ultrasound examination revealed an increase in MCA-PSV of > 1.5 multiples of the median (MoM) in one fetus that coincided with a decreased MCA-PSV of < 1.0 MoM in the cotwin, in the absence of TOPS. Postnatal diagnosis of TAPS was based on an intertwin Hb difference of 8 g/dl, with at least one of the following: reticulocytosis in the donor with a reticulocyte count ratio 1.7; small anastomoses (< 1 mm) at the placental surface 1. Intrauterine treatment was offered in cases of Stage 3 and 4 TAPS. In cases of Stage 1 or 2 TAPS, intrauterine treatment was offered in the event of quickly progressing TAPS (within days) or other signs of severe anemia not meeting criteria for Stage 3 (e.g. increasing heart size or prehydropic signs). Hb and MCA-PSV values were obtained retrospectively from medical records. MCA-PSV was measured according to a technique previously described by Mari et al. 5. MCA-PSV values were recalculated to MoM using the reference ranges for monochorionic diamniotic twin pregnancies described by Klaritsch et al. 4. In normal singleton pregnancies, Hb levels increase with gestation. Reference values for fetal Hb according to gestational age have been published by Nicolaides et al. 6. Hb deficit was calculated as the difference between the measured Hb value and the mean value for the corresponding gestational age. The definitions of severe anemia and polycythemia were those from previous studies on fetal anemia based on Rhesus alloimmunization 3. Severe anemia in TAPS donors was defined as Hb levels > 5 SD below the mean. Based on this study, severe polycythemia in the recipient twin was defined as Hb levels > 5 SD above the mean. Statistical analysis Statistical analyses were performed using SPSS version 20.0 (IBM, Armonk, NY, USA). Correlations between continuous variables were analyzed using Pearson s linear regression analysis. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio and negative likelihood ratio were calculated using 2 2 tables and standard formulae for binominal proportions. To calculate 95% CI values, the Wilson interval method was used 7. RESULTS During the study period, 43 twin pairs were diagnosed with TAPS and met the inclusion criteria (Figure 1a). In 31/43 (72%) cases TAPS occurred after laser surgery for TTTS (postlaser TAPS) and in 12/43 (28%) cases TAPS developed spontaneously. A total of 116 complete sets of both MCA-PSV and Hb measurements were available for analysis. In 55/116 (47%) sets, the correlation between MCA-PSV measurements and Hb levels from fetal blood sampling was determined at the first fetal intervention (intrauterine transfusion (IUT) and/or partial exchange transfusion (PET)). In the other 61/116 (53%) sets, the correlation between MCA-PSV measurements and Hb levels was tested during repeat fetal intervention or around the time of delivery; MCA-PSV measurements were obtained within 24 h before birth and Hb measurements were determined from cord blood samples taken at birth (Figure 1b). Of the 116 complete sets, 74 (64%) were obtained from donors and 42 (36%) from recipients. In 30 twin pairs, TAPS was diagnosed prenatally and measurements (MCA-PSV and Hb levels) were obtained during pregnancy. In 10 twin pairs, TAPS was diagnosed prenatally and measurements (MCA-PSV and Hb levels) were obtained around the time of delivery. In three twin pairs, TAPS was diagnosed only after delivery and confirmed using our predetermined postnatal criteria (Hb difference > 8 g/dl, reticulocyte count ratio 1.7 and placental anastomoses < 1 mm). Patient characteristics according to fetal and postnatal blood sampling are reported in Table 1 and neonatal outcome in donors and recipients in Table 2. Analyses of the correlation between MCA-PSV measurements and Hb levels were performed for all donors (Table 3) and also separately in a subgroup of donors, using only measurements obtained before the first IUT (Table 4). In recipients, these analyses were performed for all measurements (Table 5) and in a subgroup of recipients using only measurements obtained before the first PET (Table 6). MCA-PSV measurements and Hb levels were highly correlated (r = 0.86; P < 0.001; Figure 2). In 16 measurements, MCA-PSV was 1.5 MoM but Hb levels were 5 SD below the mean. The majority of these measurements (11/16; 69%) were performed in fetuses that had already received IUT. The sensitivity of MCA-PSV measurements in the TAPS donor as a test for Hb levels > 5 SD below the mean was 94% (95% CI, 85 98%); specificity was 74% (95% CI, 62 83%); PPV was 76% (95% CI, 65 85%); NPV was 94% (95% CI, 83 98%); positive likelihood ratio was 3.66; negative likelihood ratio was 0.07 (Table 3). Excluding measurements obtained after IUT, performance was higher, with a sensitivity of 97% (95% CI, 84 99%), specificity of 79% (95% CI, 59 91%), PPV of 86% (95% CI, 71 94%), NPV of 95% (95% CI, 76 99%),

3 434 Slaghekke et al. (a) Twin pairs (n = 43) Prenatal diagnosis, prenatal blood sample (n = 30) Prenatal diagnosis, postnatal blood sample (n = 10) Postnatal diagnosis, postnatal blood sample (n = 3) (b) Twin pairs (n = 43; 116 measurements) Prenatal blood sample (n = 55 (47%)) Postnatal blood sample (n = 61 (53%)) First measurement (n = 33) Repeat measurement (n = 22) First measurement (n = 22) Repeat measurement (n = 39) Figure 1 Flowchart of all monochorionic twin pairs with twin anemia polycythemia sequence included in the study, showing (a) time of diagnosis and collection of blood samples and (b) time of all measurements of middle cerebral artery peak systolic velocity and blood samples, with indication of whether it was the first or a repeat measurement after the (first) intrauterine intervention. Table 1 Characteristics of 43 monochorionic twin pregnancies with twin anemia polycythemia sequence, according to timing of blood sampling Characteristic Fetal blood sampling (30twinpairs) Postnatal blood sampling (13 twin pairs) GA at diagnosis (weeks) 25 (19 30) 26.5 (19 32) GA at cordocentesis (weeks) 26 ± 3 GA at birth (weeks) 31 (26 34)* 30 (27 36) Cesarean delivery 19/25 (76)* 11/13 (85) Diagnosis only after birth 3/13 (23) Data are given as median (range), mean ± SD or n/n (%). *n = 25 (one termination of pregnancy and four lost to follow-up, including cord occlusion in three). n = 10 (three diagnosed after birth). GA, gestational age. Table 2 Neonatal outcome in donors and recipients in 43 cases of monochorionic twin pregnancy complicated by twin anemia polycythemia sequence Neonatal outcome Donor Recipient P Overall survival 29/41 (71)* 40/41 (98)* Birth weight (g) 1370 ± ± Hemoglobin at birth (g/dl) 8.2 ± ± 2.7 < Data are given as n/n (%) or mean ± SD. *n = 41 (data not available for two). n = 27 (data not available for 16). n = 34 (data not available for nine). n = 28 (data not available for 15). n = 36 (data not available for seven). positive likelihood ratio of 4.65 and negative likelihood ratio of 0.04 (Table 4). In five recipients, Hb was measured before birth (as part of the PET procedure). In total, 11 measurements Table 3 Middle cerebral artery peak systolic velocity (MCA-PSV) Doppler measurements in donors of twin anemia polycythemia sequence for prediction of severe fetal anemia* MCA-PSV 1.5 MoM MCA-PSV < 1.5 MoM Data are given as n. *Severe fetal anemia defined as hemoglobin (Hb) deficit > 5 SD below the mean. MoM, multiples of the median. Table 4 Middle cerebral artery peak systolic velocity (MCA-PSV) Doppler measurements prior to first intrauterine intervention in donors of twin anemia polycythemia to predict severe fetal anemia* MCA-PSV 1.5 MoM MCA-PSV < 1.5 MoM Data are given as n. *Severe fetal anemia defined as hemoglobin (Hb) deficit > 5 SD below the mean. MoM, multiples of the median. were obtained prenatally; PET was performed once in two cases, twice in one case, three times in one case and four times in one case. In one recipient with TAPS detected postnatally, MCA-PSV before birth was not 1.0 MoM but the postnatal Hb level was > 5 SD above the mean, resulting in a sensitivity of 97% (95% CI, 87 99%). Three MCA-PSV measurements in recipients were 1.0 MoM although Hb levels were 5 SD above the mean, resulting in a specificity of MCA-PSV

4 MCA-PSV in TAPS 435 Table 5 Middle cerebral artery peak systolic velocity (MCA-PSV) Doppler measurements in recipients of twin anemia polycythemia sequence to predict severe fetal polycythemia* MCA-PSV 1.0 MoM MCA-PSV > 1.0 MoM Data are given as n. *Severe polycythemia defined as hemoglobin (Hb) deficit > 5 SD above the mean. MoM, multiples of the median. Table 6 Middle cerebral artery peak systolic velocity (MCA-PSV) Doppler measurements prior to first intrauterine intervention in recipients of twin anemia polycythemia sequence to predict severe fetal polycythemia* MCA-PSV 1.0 MoM MCA-PSV > 1.0 MoM Data are given as n. *Severe polycythemia defined as hemoglobin (Hb) deficit > 5 SD above the mean. MoM, multiples of the median. MCA-PSV MoM Hb deficit (SD of the mean) Figure 2 Scattergram of middle cerebral artery peak systolic velocity multiples of the median (MCA-PSV MoM) plotted against hemoglobin (Hb) deficit (r = 0.86; P < 0.001) in donors and recipients with twin anemia polycythemia sequence. Individuals with Hb levels > 5 SD below the mean (0) are severely anemic donors, and those with Hb levels > 5 SD above the mean are severely polycythemic recipients. in predicting polycythemia of 96% (95% CI, 89 99%). PPV was 93% (95% CI, 81 97%) and NPV was 99% (95% CI, %); positive likelihood ratio was 25 and negative likelihood ratio was 0.03 (Table 5). The mean MCA-PSV in recipients was 0.70 ± 0.17 cm/s. Table 6 shows the measurements obtained prior to any intrauterine intervention. Sensitivity was 94% (95% CI, 71 99%); specificity, 100% (95% CI, %); positive predictive value, 100% (95% CI, %); negative predictive value, 97% (95% CI, %); positive likelihood ratio, infinity; and negative likelihood ratio, DISCUSSION This is the first study reporting on the diagnostic accuracy of MCA-PSV measurements in predicting abnormal fetal Hb levels in TAPS. We found high sensitivities and specificities of MCA-PSV for both anemia and polycythemia, confirming the clinical usefulness of this non-invasive test in yet another pregnancy complication. The most common application of MCA-PSV Doppler measurements is in the management of pregnancies with red-cell alloimmunization 3. Obviously, the pathophysiology underlying development of fetal anemia in this disease differs completely from anemia in TAPS. Yet the predictive values of Doppler measurements are remarkably similar in both diseases. Interestingly, just as in alloimmune anemia, the accuracy of MCA-PSV measurements was lower following IUT Our study is the first to evaluate MCA-PSV measurements for prediction of fetal polycythemia. Recently, a study on the use of MCA-PSV in neonates showed a similar correlation between low PSV and polycythemia 11.The PPV was 76% for TAPS donors and 93% for TAPS recipients. The lower PPV in donors compared to recipients is due mainly to reduced accuracy in measurements obtained when anemia had redeveloped following IUT. Using only measurements performed prior to the first transfusion, the PPV increased to 86%. Our results strengthen our previous findings as well as our suggestion to use 1.0 MoM as a cut-off level for the TAPS recipient, as described previously in our proposed staging system 1. With the cut-off level of 0.8 MoM suggested previously, a significant number of cases with severe polycythemia would be missed 12. As shown in Figure 2, there were more outliers with a high MCA-PSV measurement compared to a low MCA-PSV measurement. Whether delta MCA-PSV MoM is a more accurate predictor of TAPS should be studied in future research. A limitation of this study is its retrospective nature. The highly selected patient group, i.e. those referred to specialized centers and predefined with suspected TAPS, influences the PPV and NPV in particular since these parameters depend strongly on prevalence. In a general population of uncomplicated monochorionic twin pregnancies, MCA-PSV measurements may not have a similar performance. To evaluate the use of MCA-PSV as a screening tool to detect TAPS in a timely manner, only large-scale prospective studies involving cohorts of monochorionic twins would provide adequate data. Such studies are time-consuming, expensive and impossible to perform in specialized fetal medicine centers, given the preselection of high-risk cases. In addition, fetal blood sampling is an invasive procedure with inherent risks, which cannot be undertaken on a large scale in pregnancies without strong suspicion of TAPS, thus limiting our ability to confirm NPVs.

5 436 Slaghekke et al. In conclusion, in selected pregnancies at increased risk of TAPS being managed in fetal medicine centers, MCA-PSV Doppler is a powerful tool for the prediction of anemia and polycythemia. REFERENCES 1. Slaghekke F, Kist WJ, Oepkes D, Pasman SA, Middeldorp JM, Klumper FJ, Walther FJ, Vandenbussche FP, Lopriore E. Twin anemia-polycythemia sequence: diagnostic criteria, classification, perinatal management and outcome. Fetal Diagn Ther 2010; 27: Lopriore E, Slaghekke F, Oepkes D, Middeldorp JM, Vandenbussche FP, Walther FJ. Hematological characteristics in neonates with twin anemia-polycythemia sequence (TAPS). Prenat Diagn 2010; 30: Oepkes D, Seaward PG, Vandenbussche FP, Windrim R, Kingdom J, Beyene J, Kanhai HH, Ohlsson A, Ryan G; DIAMOND Study Group. Doppler ultrasonography versus amniocentesis to predict fetal anemia. N Engl J Med 2006; 355: Klaritsch P, Deprest J, van MT, Gucciardo L, Done E, Jani J, Lewi P, Rasmussen S, Lewi L. Reference ranges for middle cerebral artery peak systolic velocity in monochorionic diamniotic twins: a longitudinal study. Ultrasound Obstet Gynecol 2009; 34: Mari G, Deter RL, Carpenter RL, Rahman F, Zimmerman R, Moise KJ, Jr., Dorman KF, Ludomirsky A, Gonzalez R, Gomez R, Oz U, Detti L, Copel JA, Bahado-Singh R, Berry S, Martinez-Poyer J, Blackwell SC. Noninvasive diagnosis by Doppler ultrasonography of fetal anemia due to maternal red-cell alloimmunization. Collaborative Group for Doppler Assessment of the Blood Velocity in Anemic Fetuses. NEnglJMed2000; 342: Nicolaides KH, Soothill PW, Clewell WH, Rodeck CH, Mibashan RS, Campbell S. Fetal haemoglobin measurement in the assessment of red cell isoimmunisation. Lancet 1988; 1(8594): Wilson EB. Probable inference, the law of succession, and statistical inference. Journal of the American Statistical Association 2014; 22: Mari G, Rahman F, Olofsson P, Ozcan T, Copel JA. Increase of fetal hematocrit decreases the middle cerebral artery peak systolic velocity in pregnancies complicated by rhesus alloimmunization. J Matern Fetal Med 1997; 6: Stefos T, Cosmi E, Detti L, Mari G. Correction of fetal anemia on the middle cerebral artery peak systolic velocity. Obstet Gynecol 2002; 99: Grubbs BH, Korst LM, Llanes A, Chmait RH. Middle cerebral artery Doppler and hemoglobin changes immediately following fetal transfusion. J Matern Fetal Neonatal Med 2013; 26: Weissman A, Olanovski I, Weiner Z, Blazer S. Doppler middle cerebral artery peak systolic velocity for diagnosis of neonatal anemia. J Ultrasound Med 2012; 31: Robyr R, Lewi L, Salomon LJ, Yamamoto M, Bernard JP, Deprest J, Ville Y. Prevalence and management of late fetal complications following successful selective laser coagulation of chorionic plate anastomoses in twin-to-twin transfusion syndrome. Am J Obstet Gynecol 2006; 194: