Selective intrauterine growth restriction in monochorionic twins : changing patterns in umbilical artery Doppler flow and outcomes

Objectives: To describe changes in umbilical artery (UA) Doppler flow in monochorionic diamniotic (MCDA) twins affected by selective intrauterine growth restriction (sIUGR), to correlate Doppler findings with pregnancy course and perinatal outcome, and to report postnatal follow-up. Methods: This was a retrospective study of 140 MCDA twins with sIUGR. UA end-diastolic flow, defined as Doppler waveform pattern Type I (persistently positive), Type II (persistently absent or persistently reversed) or Type III (intermittently absent or intermittently reversed), was recorded at first examination and monitored weekly until double or single intrauterine fetal death (IUFD), bipolar cord coagulation or delivery. All neonates had an early neonatal brain scan, magnetic resonance imaging, when indicated, and neurological assessment during infancy. Rates (per 100 person-weeks) and hazard ratios (HR) of IUFD in the IUGR twin in each pregnancy were calculated considering UA Doppler pattern as a time-dependent variable. Results: At first examination, there were 65 cases with UA Doppler waveform pattern Type I, 62 with Type II and 13 with Type III. Of the 65 Type-I cases, 48 (74%) remained stable, while 17 (26%) changed to either Type II absent (14%), Type II reversed (9%) or Type III (3%). Of 62 Type-II cases (47 with absent and 15 with reversed flow), 33 (53%) remained stable (18 with absent and all 15 with reversed flow). The 29 Type-II absent cases which changed became Type II reversed (24/47, 51%) or Type III (5/47, 11%). All 13 Type-III cases remained stable. Compared with Type I, the risk of IUFD (adjusted for estimated fetal weight discordance and amniotic fluid deepest vertical pocket) was highest when the pregnancy was or became Type II reversed (HR, 9.5; 95% CI, 2.7\u201332.7) or Type II absent (HR, 4.3; 95% CI, 1.3\u201314.3). Mild neurological impairment was more prevalent in the IUGR twin than in the large cotwin (7% vs 1%, P = 0.02). Conclusions: Risk stratification based on UA Doppler is useful for planning ultrasound surveillance. However, patterns can change over time, with important consequences for management and outcome