Materials and Methods

We compiled available literature including all studies reporting long-term (longer than 1 year) neurological outcome after common fetal interventions from 1990 through 2014. Institutional review board approval was not obtained because this study was a review of previously published data.

Procedures were divided by singletons or multiples. Singleton procedures included amnioinfusion for preterm premature rupture of membranes, intrauterine transfusion for red cell alloimmunization–associated anemia, intrauterine transfusion for parvovirus-associated anemia, vesicoamniotic shunts, thoracoamniotic shunts, ventriculoamniotic shunts, fetal endoscopic tracheal occlusion for congenital diaphragmatic hernia, and open fetal cases for myelomeningocele (MMC) and others, such as fetal lung lesions with hydrops. Procedures done on multiples included those done for complicated monochorionic pregnancies: serial amnioreduction, selective fetoscopic laser photocoagulation of anastamotic vessels, and selective termination.

Both PubMed and Cochrane databases were searched for articles published in English from January 1990 through April 2014 using the search terms listed in Table 1 . Additionally, citations were reviewed from all included papers, and relevant studies were also included. Because available data are limited, we included all published studies if they had neurological outcome information for a cohort of more than 5 patients and if the follow-up period was 1 year or longer. In studies that combined patients from multiple cohorts, we attempted to include all patients who were not included from other reports. We excluded studies with fewer than 5 patients, those in which the neurological outcome was unclear, and those in which it was impossible to discern which of a cohort had been previously included.

Study results were combined to give an overall impression of outcome over multiple experiences and centers.

Results

Of 1341 studies identified by the search engines, 54 potentially relevant publications were reviewed, and 28 met our inclusion criteria ( Table 1 ). The length of follow-up in each category and methods used to evaluate the neurological outcome are detailed in Table 2 . Most studies utilized a standardized approach, but 2 were based on questionnaires alone, and 2 did not detail how follow-up was done.

The most commonly used assessment methods were basic neurological examination, Bayley Scales of Infant Development, and the Weschler Intelligence Scale for Children. Eleven of the 28 studies documented that they excluded fetuses with additional severe anomalies or genetic syndromes, whereas the remaining studies did not specifically mention this as part of their exclusion criteria.

Although the available studies used multiple different methods for assessing long-term neurological status ( Table 2 ), outcomes were generally divided into normal, mildly impaired, or severely impaired. Levels of impairment were not always clearly defined but when defined were generally as follows: (1) normal, normal physical and neurological development with no developmental delay; (2) mild impairment, neurological deficiencies as compared with normal such as mildly retarded motor and/or speech development or sensorineural deafness requiring hearing aids; and (3) severe impairment, cerebral palsy with hemiparesis or spastic quadriplegia, cognitive impairment with a developmental score less than 2 SD below the mean, blindness, or complete deafness.

Combined data from the available studies including the number of studies and patients included, the perinatal mortality rates, and the percentage from each study with normal, mildly impaired, or severely impaired neurological outcome are detailed in Table 3 and Figures 1 and 2 .

Neurological outcomes after fetal intervention

Gebb. Neuro outcome after fetal intervention. Am J Obstet Gynecol 2015 .

Neurological outcomes in survivors after fetal intervention

Gebb. Neuro outcome after fetal intervention. Am J Obstet Gynecol 2015 .

For the singleton procedures, studies were available for the following:

• 1.
  

  Amnioinfusion for preterm premature rupture of membranes. Three studies with 109 patients with long-term outcome between 12 and 106 months were included. Overall perinatal mortality was 55%, and long-term neurological data were available for 96% of survivors. Of the 47 survivors, 14 (30%) were reported to have severe impairment. Only 2 studies with 21 survivors reported on normal or mildly impaired outcome. Of those survivors, 13 of 21 (62%) were reported as neurologically normal and 3 of 21 (14%) were reported to have minor impairment.

  
  
• 2.
  

  Intrauterine transfusion.

  
  
  
  
  • a.
    

    Alloimmunization-associated anemia: 4 studies with 564 patients with long-term outcome between 24 and 204 months were included. Overall perinatal mortality was 12%, and long-term neurological data were available for 76% of the survivors. Of the 375 survivors, 307 (82%) were reported as neurologically normal, 50 (13%) were reported to have minor impairment, and 18 (5%) were reported to have severe impairment.

    
    
  • b.
    

    Parvovirus B19-associated anemia: 2 studies with 81 patients with long-term outcome between 13 and 156 months were included. Overall perinatal mortality was 22%, and long-term neurological data were available for 79% of survivors. Of the 48 available survivors, 40 (83%) were reported as neurologically normal, 5 (10%) were reported to have minor impairment, and 3 (6%) were reported to have severe impairment.

  
  
  
  
• 3.
  

  Vesicoamniotic shunt: 2 studies with 65 patients with follow-up from 12 to 168 months were included. The perinatal mortality rate was 23%, and long-term neurological follow-up was available in 64% of survivors. Of those, 22 (69%) were reported as normal, 9 (28%) were reported as mildly impaired, and 1 (3%) was reported as severely impaired.

  
  
• 4.
  

  Thoracoamniotic shunt: 1 study with 11 patients with follow-up from 12 to 42 months was included. The perinatal mortality rate was 9%, and long-term neurological follow-up was available in 60% of the survivors. Of those, 4 (67%) were reported as normal, 2 (33%) were reported as mildly impaired, and none were reported as severely impaired.

  
  
• 5.
  

  Cephalocentesis/ventriculoamniotic shunt: 1 study with 39 patients with follow-up from 12 to 60 months was included. The perinatal mortality rate was reported as 0, and long-term follow-up was available in all 39 patients. Of those, 26 (67%) were reported as normal, 6 (15%) were reported as mildly impaired, and 7 (18%) were reported as severely impaired.

  
  
• 6.
  

  Fetal endoscopic tracheal occlusion: only 1 study with 11 patients with follow-up from 23 to 26 months was available for analysis. The perinatal mortality rate was 36%, and long-term neurological follow-up was available in 5 survivors (71%). Of those, 2 (40%) were reported as normal, 1 (20%) was reported as mildly impaired, and 2 (40%) were reported as severely impaired.

  
  
• 7.
  

  MMC open fetal surgery: 2 studies with 64 patients with follow-up from 45 to 71 months were included. The perinatal mortality rate was 6%, and long-term neurological follow-up was available in 60% of the survivors. Of those, 23 (64%) were reported as normal, 6 (17%) were reported as mildly impaired, and 7 (19%) were reported as severely impaired. Of note, the physical impairment in these cases varied by the level of the lesion, and the neurological outcome reported refers to cognitive outcome.

  
  
• 8.
  

  Non-MMC open fetal surgery: 1 study with 36 patients with follow-up from 25 to 69 months was included. The perinatal mortality rate was 67%, and long-term neurological follow-up was available in 7 survivors (58%). Of those, 5 (71%) were reported as normal, 1 (14%) was reported as mildly impaired, and 1 (14%) was reported as severely impaired.

For the multiples procedures specific to monochorionic pregnancies, studies were available for three different kinds of procedures. The first was amnioreduction. Three studies with 182 fetuses with follow-up from 36 to 144 months were included. The perinatal mortality rate was 55% of fetuses, and the long-term neurological follow-up was available in 73 survivors (90%). Only 1 study reported severe impairment, so the number with normal or mild impairment for that study was unknown. Of survivors with known outcome, 43 of 54 (80%) were reported as normal, 3 of 54 (6%) were reported as mildly impaired, and 13 of 73 (18%) were reported as severely impaired.

The second procedure was selective fetoscopic laser photocoagulation. Seven studies with 1428 fetuses with follow-up from 14 to 72 months were included. The perinatal mortality rate for the available studies was 31%, and the long-term neurological follow-up was available in 969 survivors. Several studies reported only those patients with severe neurological outcome, and therefore, the number with a normal outcome or mild impairment was available for only a smaller cohort. Of those reported, 274 of 325 (84%) were deemed normal, 22 of 325 (7%) were mildly impaired, and 115 of 969 (12%) were severely impaired.

The third procedure was selective termination. One study with 87 fetuses with follow-up from 12 to 72 months was included. The perinatal mortality rate of the fetus selected for survival was 17%, and the long-term neurological follow-up was available in 93% of the survivors. Of those, 62 (93%) were reported as normal, 4 (6%) were reported as mildly impaired, and 1 (2%) was reported as severely impaired.