Long-term efficacy and learning curves of Cohen and modified Glenn-Anderson pneumovesicoscopic ureteral reimplantation for primary obstructive megaureter: a retrospective cohort study with meta-analysis.

OBJECTIVE

To compare the long-term efficacy and safety of Cohen versus modified Glenn-Anderson pneumovesicoscopic ureteral reimplantation and to define the learning curves for both procedures.

METHODS

A single-center retrospective cohort study (2007-2021) included 75 children with primary obstructive megaureter (POM) (Cohen: n = 27; modified Glenn-Anderson: n = 48). Primary outcomes were reoperation rates and ureteral recovery (diameter < 7 mm) during long-term follow-up. Cumulative sum (CUSUM) analysis was used to model learning curves based on operative time. A meta-analysis of 10 studies (529 patients) established a benchmark success rate.

RESULTS

After a mean follow-up of 99.0 (Cohen) and 75.7 (modified Glenn-Anderson) months, reoperation-free rates were 92.6% and 95.8%, respectively (P > 0.05), with no significant differences in complication rates or functional improvement. CUSUM analysis indicated a learning phase of 22 cases for Cohen and 19 cases for modified Glenn-Anderson, with operative time decreasing by over 30% upon mastery. The meta-analysis yielded a pooled reoperation-free rate of 99.4%.

CONCLUSIONS

Both techniques demonstrate comparable long-term safety and efficacy for POM. The modified Glenn-Anderson technique offers a valid alternative that preserves the ureter's anatomical course. Achieving procedural proficiency requires approximately 20 cases. While global success rates are high, long-term follow-up is crucial for detecting late complications.