Objective
To determine preoperative predictive factors for donor and recipient death after laser ablation of placental vessels in twin-to-twin transfusion syndrome.
Study Design
Retrospective analysis of North American Fetal Therapy Network center laser procedures, 2002-2009. Factors associated with donor and recipient death were identified by regression analysis.
Results
There were 466 patients from 8 centers. Factors significantly associated with donor fetal death were low donor estimated fetal weight (odds ratio [OR], 0.69; 95% confidence interval [CI], 0.55–0.87) and reversed end diastolic velocity in the umbilical artery (OR, 4.0; 95% CI, 1.54–10.2); for recipient fetal death–low recipient estimated fetal weight (OR, 0.65; 95% CI, 0.44–0.95), recipient reversed “a” wave in the ductus venosus (OR, 2.39; 95% CI, 1.27–4.51) and hydrops (OR, 3.7; 95% CI, 1.1–12.7); for recipient neonatal death–low donor estimated fetal weight (OR, 0.54; 95% CI, 0.30–0.95), high recipient estimated fetal weight (OR, 1.55; 95% CI, 1.06–2.26) and recipient reversed end diastolic velocity in the umbilical artery (OR, 7.8; 95% CI, 1.03–59.3).
Conclusion
Preoperative findings predict fetal and neonatal demise in twin-to-twin transfusion syndrome treated with laser therapy.
Severe twin-to-twin transfusion syndrome (TTTS) is a devastating disease produced by a circulatory imbalance of blood and other vasoactive substances from 1 fetus to another through placental vascular connections in the setting of a monochorionic twin pregnancy. Management options for TTTS include observation, termination of the entire pregnancy, amnioreduction, septostomy, endoscopic laser ablation of placental vessels, and selective feticide of a single fetus. An ethics-based clinical decision tree was recently reported, to aid health care professionals in the counseling of patients with TTTS. Informed decisions, as well as directive and nondirective counseling, imply a reasonable understanding of the natural evolution of the disease and the expected outcome for the various therapeutic interventions. Selective feticide, in particular, requires accurate predictive indicators for each fetus. If this modality is chosen, the decision as to which fetus should be targeted for feticide should be made based on medical knowledge for the prediction of death. Based on a large study of TTTS patients treated with amnioreduction, the outcome for any single fetus (donor or recipient) can be predicted based on the presence or absence of specific ultrasound abnormalities. A clinical impression has developed in many experts that surgical intervention with laser may alter the natural evolution of TTTS, rendering preoperative factors no longer predictive of outcomes. The objective of the current study was to determine whether fetal and neonatal death of individual twins can be predicted preoperatively for pregnancies treated by endoscopic laser ablation of placental vessels in TTTS.
Materials and Methods
The North American Fetal Therapy Network (NAFTNet) is a collaborative initiative of 20 fetal therapy centers in the United States and Canada. TTTS pregnancies are commonly cared for at NAFTNet centers and clinical data and outcomes are routinely recorded prospectively as part of clinical care. This study is a retrospective analysis of prospectively collected data of TTTS pregnancies that underwent endoscopic laser ablation of placental vessels at 8 NAFTNet centers from January 2002 to June 2009. Cases that were included in the National Institutes of Health (NIH)-sponsored randomized clinical trial comparing amniocentesis vs laser therapy were excluded from this study. Twelve of the 20 NAFTNet centers perform laser for TTTS. Two of the centers participated in the NIH trial. One did not participate in this study because of their participation in the NIH trial. The other center that participated in the NIH trial did participate in this study, but cases were included from this center only after their completion of the NIH trial. The other centers were unable to comply with the data requirements of the study or had institutional review board restrictions preventing their participation.
The protocol was approved by the human institutional review board of Weill Cornell College of Medicine, which was the data coordinating center, and the institutional review boards at the respective centers. For each center, all consecutive cases where laser was performed for TTTS were included. Clinical and ultrasound data were used to identify possible predictive factors for fetal and neonatal death ( Table 1 ). Diagnostic criteria for TTTS included ultrasound evidence of monochorionicity, disparate amniotic fluid volumes with 1 sac having a deepest vertical pocket <2.0 cm and the other sac having >8 cm deepest vertical pocket when ≤20 weeks of gestation and >10 cm when >20 weeks of gestation, and disparate bladder sizes. Doppler abnormalities and cardiac functional abnormalities were used to confirm the diagnosis in several borderline cases where amniotic fluid levels were not quite diagnostic. The preoperative assessment included the ultrasound and Doppler studies necessary for diagnosis and those listed in Table 1 . Doppler flow abnormalities included either intermittent or persistent findings in 1 category—differentiation between intermittent and persistent was not possible in the analysis. All centers performed routine fetal echocardiography. Three of 8 participating centers upstaged based on echocardiographic findings. The technique for endoscopic laser ablation of placental vessels was a single-port fetoscopy under regional or general anesthesia, identification of the placental vascular equator, and anastomotic connections between the fetuses, and neodynium-yttrium aluminum garnet (Nd-YAG) or diode laser ablation of all anastomoses. All centers during the study period used a random (nonsequential) and selective ablation method. Termination of pregnancy and spontaneous fetal deaths were recorded after all procedures. Cases of selective feticide or termination of the entire pregnancy after laser therapy were included in the denominator of fetal deaths because it was believed that these cases represented untoward events after laser treatment. Fetal death was defined as death before birth. Neonatal death was defined as the death of a living neonate within 30 days of birth. Deidentified data were submitted from each center with a centralized website using a “point and click” data entry tool that used minimal free text entry.
| Factor |
|---|
| Gestational age |
| Amniotic fluid volume by deepest vertical pocket measurement, cm |
| Estimated fetal weight EFW, g |
| Percent difference in EFW between donor and recipient |
| EFW percent difference = 100 × (Recipient EFW – Donor EFW)/Recipient EFW |
| Presence of fetal bladder |
| Absent end diastolic velocity in the umbilical artery a |
| Reverse end diastolic velocity in the umbilical artery a |
| Pulsatile umbilical vein waveform |
| Absent “a” wave in the ductus venosus a |
| Reverse “a” wave in the ductus venosus a |
| Ascites |
| Pericardial effusion |
| Pleural effusion |
| Skin/cord edema |
| Hydrops fetalis |
| Presence of ≥2 body cavity effusions |
| Mitral or tricuspid cardiac valvular regurgitation |
| More than trace |
| Global cardiac dysfunction |
| Abnormal Tai index, ventricular dyskinesia, abnormal ejection fraction for either ventricle, abnormal cardiac “score” or other measure of cardiac dysfunction |
Four distinct endpoints were analyzed—fetal donor death, fetal recipient death, neonatal donor death, and neonatal recipient death. For each endpoint, 2 analyses were performed. First, to provide simple descriptive measures of the association between each individual preoperative factor and death, bivariate tests of association were carried out—Fisher’s exact test for categorical predictors; Mann-Whitney U test for continuous predictors. Second, a multivariate logistic model was developed to predict the probability of death, based on preoperative factors. To obtain a “parsimonious” model (one with as few factors as possible that was nearly as good a predictor as a model using all available factors), a stepwise algorithm was applied. All 32 available factors were provided to the algorithm, which then systematically removed factors which either were not effective predictors of death, or which were redundant with other factors that were even more strongly predictive. The stepwise algorithm was generally able to produce effective predictive models which, depending on the endpoint, required only 6 to 10 of the 32 available preoperative factors. For each of the 4 models, a receiver operator characteristic (ROC) curve was prepared, showing the sensitivity/specificity tradeoffs that can be achieved by judiciously choosing the cutoff value that dichotomizes the predicted probability of death into a “death or no death” prognosis. Each component of the Quintero staging score is included individually in the data analysis ( Table 1 ) and thus each is considered as part of the stepwise logistic regression analysis.
Sample size calculations are based on Martinez et al. In this recent report, fetal demise occurred in 10/15 (67%) donors with absent or reversed end diastolic velocity (AEDV, REDV) in the umbilical artery (UA) and survival occurred in 22/95 without this finding. Fetal demise occurred in 9/33 (27%) recipients with pulsatile umbilical venous flow and 7/74 (9.5%) without this finding. Thus, using α <0.05 and β = .80, the sample size necessary to avoid a type 2 error is either 51 donors or 229 recipients. A larger sample size allows greater precision in developing predictive models, so we attempted to include as many subjects as possible. All graphs and statistical analyses were generated using the “R” software package ( http://www.r-project.org ).
Results
There were a total of 466 cases of laser ablation of placental vessels for TTTS. All pregnancies were confirmed to be monochorionic by ultrasound criteria or pathologic examination of the placenta after birth. The participating centers and the number of procedures performed at each center are seen in Table 2 . The baseline data of the study population can be found in Table 3 . Complete termination of pregnancy occurred after laser therapy in 3 cases—2 for worsening TTTS and 1 for worsening condition of the recipient. Selective feticide occurred in 7 cases: 4 recipients (2 for persistent TTTS, 1 for worsening fetal condition, and 1 for fetal bradycardia) and 3 donors (2 for worsening fetal condition and 1 for intracranial hemorrhage seen on ultrasound). Thirty cases were lost to neonatal follow-up (6.4%). Overall neonatal survival was 596/872 (68.3%). Donor fetal and neonatal survival were 355/466 (76.2%) and 280/436 (64.2%), respectively. Recipient fetal and neonatal survival were 389/466 (83.5%) and 316/436 (72.5%), respectively. Survival of at least 1 twin was 357/436 (81.9%) and survival of both twins was 236/436 (54.1%). There were 49 subjects with stage 1 TTTS, 116 subjects with stage 2 TTTS, 271 subjects with stage 3 TTTS, and 30 subjects with stage 4 TTTS.
| Participating center | Procedures performed, n |
|---|---|
| Evergreen Hospital, Seattle, WA | 168 |
| Children’s Hospital of Philadelphia/University of Pennsylvania | 106 |
| Baylor University/Texas Children’s Fetal Center | 50 |
| University of Toronto | 44 |
| Alpert Medical College of Brown University | 42 |
| Ohio State University | 30 |
| University of North Carolina | 25 |
| Yale University | 1 |
| Total | 466 |
| Variable | Donor (n = 466) | Recipient (n = 466) |
|---|---|---|
| Gestational age at procedure, wk, median (range) (n) | 20 (12–26) (465) | |
| Percent EFW difference at procedure, median (range) (n) | 24.6 (−36.9 to 64.9) (416) | |
| EFW at procedure, g mean ± SD (n) | 289 ± 137 (418) | 387 ± 180 (417) |
| Maximum vertical pocket, cm mean ± SD (n) | 0.70 ± 0.88 (464) | 11.2 ± 3.1 (465) |
| Bladder present | 148/466 (31.8%) | 457/465 (98.3%) |
| Absent umbilical arterial end diastolic flow | 124/463 (26.8%) | 29/463 (6.3%) |
| Reversed umbilical arterial end diastolic flow | 24/465 (5.2%) | 6/458 (1.3%) |
| Pulsatile UV | 18/436 (4.1%) | 120/434 (27.6%) |
| Absent a wave | 26/393 (6.6%) | 29/393 (7.4%) |
| Reversed a wave | 24/401 (6.0%) | 134/399 (33.6%) |
| Ascites | 1/360 (0.3%) | 25/360 (6.9%) |
| Pericardial effusion | 5/358 (1.4%) | 37/356 (10.4%) |
| Pleural effusion | 0/354 (0%) | 9/356 (2.5%) |
| Subcutaneous edema | 1/353 (0.3%) | 17/353 (4.8%) |
| Hydrops fetalis | 0/466 (0%) | 32/466 (6.9%) |
| Tricuspid or mitral regurgitation | 8/465 (1.7%) | 196/462 (42.4%) |
| Abnormal global cardiac function | 12/464 (2.6%) | 165/459 (35.9%) |
Univariate analyses were performed to identify factors that were individually associated with donor fetal death, recipient fetal death, donor neonatal death, and recipient neonatal death ( Tables 4 and 5 ). For each of the 4 analyses, stepwise logistic regression identified a model that was a combination of preoperative factors that best predicted fetal and neonatal death ( Figure 1 and Table 6 ). The best achievable combination of sensitivity and specificity respectively is as follows: for donor fetal death 65% and 65%, for recipient fetal death 70% and 68%, for donor neonatal death 62% and 60%, and for recipient neonatal death 60% and 55%.
| Variable | Donor fetal demise | Recipient fetal demise | ||||||
|---|---|---|---|---|---|---|---|---|
| No fetal demise | Fetal demise | Univariate P value | Multivariate P value | No fetal demise | Fetal demise | Univariate P value | Multivariate P value | |
| n (%) | 355 (76.2) | 111 (23.8) | 389 (83.5) | 77 (16.5) | ||||
| Gestational age at procedure, wk median (range) | 20 (12–26) | 19 (14–26) | .007 | 20 (12–26) (388) | 19 (16–24) (77) | .115 | ||
| Maximum vertical pocket amniotic fluid donor, cm mean ± SD (n) | 0.69 ± 0.89 (354) | 0.73 ± 0.86 (111) | .384 | 0.68 ± 0.87 (388) | 0.78 ± 0.91 (77) | .325 | ||
| Maximum vertical pocket amniotic fluid recipient, cm mean ± SD (n) | 11.2 ± 3.0 (354) | 11.2 ± 3.4 (111) | .533 | 11.2 ± 3.1 (388) | 11.2 ± 2.9 (77) | .488 | ||
| EFW donor, g mean ± SD (n) | 299 ± 133 (354) | 247 ± 109 (111) | < .001 | .002 | 288 ± 134 (388) | 281 ± 107 (77) | .884 | .09 |
| EFW recipient, g mean ± SD (n) | 392 ± 170 (354) | 356 ± 170 (111) | .012 | 389 ± 177 (388) | 355 ± 131 (77) | .335 | .03 | |
| Growth discrepancy or percent difference in EFW, % median (range) (n) | 24.6 (−36.9 to 64.9) | 28.5 (−7.9 to 59.1) | < .001 | .18 | 24.6 (−36.9 to 64.9) (388) | 24.6 (−14.8 to 48.8) (77) | .049 | |
| Donor bladder present | 110/354 (31%) | 38/111 (34%) | .560 | 125/388 (32%) | 23/77 (30%) | .789 | ||
| AEDV UA donor | 83/351 (24%) | 40/111 (36%) | .013 | .15 | 103/385 (27%) | 20/77 (26%) | > .99 | |
| AEDV UA recipient | 25/352 (7.1%) | 4/110 (3.6%) | .261 | 22/386 (5.7%) | 7/76 (9.2%) | .297 | ||
| REDV UA donor | 10/353 (2.8%) | 14/111 (13%) | < .001 | .004 | 19/387 (4.9%) | 5/77 (6.5%) | .573 | .09 |
| REDV UA recipient | 4/349 (1.1%) | 2/108 (1.9%) | .630 | 4/381 (1%) | 2/76 (2.6%) | .262 | ||
| Pulsatile UV donor | 14/335 (4.2%) | 4/100 (4%) | > .99 | 17/365 (4.7%) | 1/70 (1.4%) | .330 | ||
| Pulsatile UV recipient | 97/334 (29%) | 23/99 (23%) | .307 | 100/363 (28%) | 20/70 (29%) | .884 | ||
| Absent a wave DV donor | 18/309 (5.8%) | 7/83 (8.4%) | .446 | 20/339 (5.9%) | 5/35 (14%) | .360 | ||
| Absent a wave DV recipient | 26/309 (8.4%) | 3/83 (3.6%) | .162 | .08 | 23/339 (6.8%) | 6/53 (11%) | .257 | .08 |
| Reversed a wave DV donor | 16/315 (5.1%) | 8/85 (9.4%) | .194 | 21/345 (6.1%) | 3/55 (5.5%) | > .99 | ||
| Reversed a wave DV recipient | 105/313 (34%) | 29/85 (34%) | > .99 | 107/343 (31%) | 27/55 (49%) | .013 | .007 | |
| Ascites donor | 1/261 (0.4%) | 0/98 (0%) | > .99 | .004 | 1/293 (0.34%) | 0/66 (0%) | > .99 | |
| Ascites recipient | 19/261 (7.3%) | 6/98 (6.1%) | > .99 | 20/293 (6.8%) | 5/66 (7.6%) | .906 | ||
| Pericardial effusion donor | 1/261 (0.4%) | 4/96 (4.2%) | .020 | .13 | 5/292 (1.7%) | 0/65 (0%) | .589 | |
| Pericardial effusion recipient | 29/259 (11%) | 8/96 (8.3%) | .558 | 34/292 (12%) | 3/63 (4.8%) | .117 | .07 | |
| Pleural effusion donor | 0/258 (0%) | 0/95 (0%) | 0/288 (0%) | 0/65 (0%) | ||||
| Pleural effusion recipient | 8/259 (3.1%) | 1/96 (1%) | .454 | 9/291 (3.1%) | 0/64 (0%) | .373 | .25 | |
| Subcutaneous edema donor | 1/259 (0.4%) | 0/93 (0%) | > .99 | 1/287 (0.35%) | 0/65 (0%) | > .99 | ||
| Subcutaneous edema recipient | 16/258 (6.2%) | 1/94 (1.1%) | .050 | .15 | 13/288 (4.5%) | 4/64 (6.3%) | .525 | |
| Hydrops fetalis donor | 0/354 (0%) | 0/111 (0%) | 0/388 (0%) | 0/77 (0%) | ||||
| Hydrops fetalis recipient | 27/354 (7.6%) | 5/111 (4.5%) | .389 | 26/388 (6.7%) | 6/77 (7.8%) | .805 | .04 | |
| Tricuspid or mitral regurgitation donor | 6/354 (1.7%) | 2/110 (1.8%) | .541 | 7/387 (1.8%) | 1/77 (1.3%) | .534 | ||
| Tricuspid or mitral regurgitation recipient | 152/352 (43%) | 44/109 (40%) | .933 | 154/384 (40%) | 33/77 (43%) | .092 | .07 | |
| Global cardiac dysfunction donor | 6/353 (1.7%) | 6/110 (5.5%) | .087 | 12/386 (31%) | 0/77 (0%) | .571 | .98 | |
| Global cardiac dysfunction recipient | 122/350 (35%) | 43/108 (40%) | .597 | .13 | 134/381 (35%) | 31/77 (40%) | .001 | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
