We wish to congratulate the authors on their important study on long-term outcome in the Eurofoetus trial. The authors found a similar rate of cerebral palsy (CP) between the fetoscopic selective laser coagulation (FSLC) and the amniodrainage (AD) group, 13% (9/69) and 15% (6/41), respectively (rates of CP are higher than the reported rates because we excluded children lost to follow-up, n = 10).

Neurodevelopmental outcome scores using Ages and Stages Questionnaires at various time points were similar in both groups (except at 5 years of age). When assessed at 6 years of age with a more objective test (Wechsler Intelligence Scale for Children), again no differences were found between both groups.

These findings may seem controversial considering a lower incidence of neurologic sequelae at 6 months of age in the FSLC group. In addition, most large case series (albeit not from randomized controlled trials) report lower neurodevelopmental impairment (NDI) rates after FSLC treatment than after AD.

Several confounders may explain why FSLC was not associated with a reduced risk of NDI. First, although the total IQ was similar in both groups, the SD was much larger in the AD group (91 ± 33.1, ie, a –1 SD score of 57.9). Presuming that IQ scores are normally distributed, this suggests that a significant number of infants in the AD group had severe developmental delay (IQ < 70 [–2 SD]). If this is true, the rate of NDI (defined as CP, IQ < –2 SD, blindness or deafness) could be higher in the AD group. Unfortunately, this composite outcome (NDI) was not reported. We think it is of paramount importance to supply fetal surgeons and parents with this valuable information.

Second, the most important reason for absence of a difference in neurodevelopmental outcome is probably the extremely high rate of neonatal deaths in the AD group (36% [26/73] of all live-born babies). Detailed information on the causes of death is not provided, but the high mortality rate could be due to withdrawal of intensive care treatment in very preterm infants with severe cerebral lesions. This information is crucial to allow readers to put the results of the present study into perspective. As reported in the Eurofoetus trial, 22% (20/93) of live-born infants in the AD group had very severe cystic periventricular leukomalacia (grade 3). A positive correlation between high neonatal death rate and withdrawal of intensive care treatment in infants with severe cerebral injury strongly influences the interpretation of the long-term outcome results. Had these severely damaged children survived, the differences in long-term neurodevelopmental outcome between both groups would have been much more evident.

Finally, we fully agree with the authors that the objective of twin-to-twin transfusion syndrome treatment should be a high rate of intact survival and this goal is better achieved with FSLC than with AD. We think that the risk of adverse long-term neurodevelopmental outcome in twin-to-twin transfusion syndrome treated with AD may be even higher than suggested, considering the arguments reported here above.