New in EJOG: Surgical Management of Interstitial (Cornual) Pregnancy
Every obstetrician in America knows the Gray Journal. The American Journal of Obstetrics & Gynecology has been published continuously since 1920, it is the journal our specialty reads first, and as of today it carries the biggest paper of my career.
Surgical and Obstetric Complications of Transabdominal Cervical Cerclage by Laparoscopic, Open, and Robotic Approaches: A Systematic Review and Meta-Analysis
Marchand GJ, Gonzalez Herrera D, Robinson M, O’Connell B, Anderson D, Mera S, Palma A, Ibrahim F, Azadi A. Am J Obstet Gynecol. Article in Press, online November 1, 2026. DOI 10.1016/j.ajog.2026.10.040.
Where to read it
The article on ajog.org (Articles in Press) · DOI 10.1016/j.ajog.2026.10.040. PubMed: PMID 42967520. PROSPERO registration CRD420261422920.
Why this operation matters to me
A transabdominal cerclage is a stitch placed at the very top of the cervix, from inside the abdomen, for women whose cervix cannot stay closed through a pregnancy and for whom the usual vaginal stitch has failed or is not possible. For most of its history that meant open surgery. In 2010 Dr. Richard Demir and I presented a laparoscopic cervical cerclage performed in an 18-week pregnant uterus at the Society of Laparoendoscopic Surgeons and at the British International Congress in Belfast, and we published the technique as a training video. Dr. Demir is gone now, and I think about him every time I place one of these. In 2023 Rutgers New Jersey Medical School put that technique into its OB/GYN core competencies curriculum, which remains the proudest teaching moment I have had.
So I have a stake in this operation, and that is exactly why this paper had to be done the hard way.
What was wrong with the old evidence
My team has published two previous meta-analyses on abdominal cerclage, one on complications in JMIG in 2021 and one on efficacy in EJOG in 2022. Both of them, like every other review in the field, compared the laparoscopic literature with the open literature as two separate piles and drew conclusions from the difference between the piles. The problem is that the two piles were never the same patients, the same era, the same surgeons or the same definitions. This time we asked a narrower and much more honest question: when a single study reported both routes in the same population, did one route actually do better? It did not. The separately pooled method cannot establish an advantage for either route, and the within-study evidence does not show one.
What we did
We searched PubMed, Cochrane CENTRAL, Web of Science and Embase through June 12, 2026, carried forward every eligible study from our previous review, and screened citation lists. Screening was duplicate and blinded. Every complication count was taken from the source full text against a written schema, not from a table in someone else’s review. Risk of bias used instruments matched to the design of each study, and we graded certainty separately for the comparative question and the absolute question. The search returned 822 database records, 440 once duplicates were removed. We included 67 reports, withdrew four that overlapped, and analyzed 63 studies, 70 study arms and 5,094 unique women. That makes it the largest body of patients ever assembled on this operation, and the first analysis restricted to within-study comparisons. The comparison itself rests on six studies and 317 pregnancies, every one of them non-randomized and at serious risk of bias.
What we found
- Six studies, 317 pregnancies, compared the routes internally. Across them there was no detectable difference in any complication. Total fetal loss gave a risk ratio of 0.83 (95% CI 0.19 to 3.64) for laparoscopic versus open, and second-trimester loss 0.97 (0.21 to 4.46). Those intervals are wide enough to hide a real difference in either direction, so this is not a finding of equivalence. It is a finding that the question is unresolved, and the certainty of every comparative estimate is very low.
- Major hemorrhage and uterine perforation were uncommon after laparoscopic placement. Not one major hemorrhage was reported in 1,889 laparoscopic procedures, which caps that risk at 0.2 percent. The open literature that reported it pooled at about 2 percent, or 5.3 percent with the conventional continuity correction, but that is a comparison of two literatures, not of two routes, and we do not make it. Inside the studies that compared the routes directly, hemorrhage and perforation had no events in either arm.
- Pregnancy loss is not uncommon. In the single-route series, total fetal loss ran about 7.5 percent after laparoscopic placement and about 8.9 percent after open placement, with overlapping intervals, and a conventional continuity-corrected model pushed every estimate higher, seven-fold for perforation. We report both models so nobody has to take our word for which one flatters the result.
- The advantages reported in earlier reviews do not reproduce inside cohorts. Four of the six comparative studies were already sitting inside the largest previous synthesis, 83 studies and 3,398 women, which reported that no study compared the routes directly. The comparison was never absent. It was dissolved, because a study reporting both routes enters a separately pooled analysis as two single arms.
What it means
If you are a woman facing this operation: in nearly two thousand reported laparoscopic procedures there was not one major hemorrhage, and perforation was uncommon too. The thing to worry about is not the operation. It is the pregnancy. Roughly one in twelve was still lost after an abdominal cerclage by either route, and no route has been shown to change those odds. If you are a surgeon: pick the route you do best, for the anatomy in front of you, at the point in pregnancy you are in, weighed with the efficacy evidence and what your patient wants. Do not pick it because a review compared two separate literatures and found a difference.
If you are a researcher: the field does not need another single-arm series. It needs reports from centers performing both operations, with arm-level numerators and denominators for each complication and the threshold, the perforation definition and the timing stated. Only one of the six comparative studies reports PPROM by arm and only three report perforation by arm, and the existing cohorts almost certainly contain those numbers. Until someone publishes them we will keep writing this review.
Five rounds of review
This paper went through five rounds of revision at AJOG, and the editors pushed us on exactly the things that should be pushed on: which statistical model is primary when events are rare, how to appraise single-arm cohorts honestly, and how far a sensitivity analysis can be read. It is a much better paper for it, and I am grateful to them.
Congratulations and thanks to Daniela Gonzalez Herrera, McKenna Robinson, Brooklynn O’Connell, Dakota Anderson, Sarah Mera and Adriana Palma at the Institute, to Fatima Ibrahim at Fayoum University, and to Dr. Ali Azadi, who has been beside me on the cerclage papers since the first one. We took no funding for this work and have no conflicts to declare.
Greg J. Marchand, MD, FACS, FACOG, FICS
Related posts
