Vascular hitch for paediatric pelvi-ureteric junction obstruction with crossing vessels: institutional analysis and systematic review with meta-analysis.

OBJECTIVE: To compare the outcomes for laparoscopic vascular hitch (VH) and dismembered pyeloplasty in patients with pelvi-ureteric junction obstruction (PUJO) secondary to crossing vessels (CV). PATIENTS AND METHODS: Patients who underwent laparoscopic management of CV at our institution were identified between 2008 and 2020. Baseline characteristics and outcome measures were compared between those who underwent VH and those who underwent dismembered pyeloplasty. Those who underwent VH were selected intraoperatively by identifying CV in the absence of intrinsic obstruction by assessing resolution of hydronephrosis after cranially displacing the CV away from the PUJ, followed by intraoperative fluid bolus and diuretic test. In addition, a systematic search and meta-analysis were performed in June 2020 (International Prospective Register of Systematic Reviews [PROSPERO] ID CRD42020195833). RESULTS: A total of 20 patients underwent VH and 74 dismembered pyeloplasty. On multivariate analysis, VH was associated with: shorter operative time (P = 0.001; odds ratio [OR] 0.455, 95% confidence interval [CI] 0.294-0.705) and length of stay (P = 0.012; OR 0.383, 95% CI 0.183-0.803), lower use of stents (P < 0.001; OR 0.024, 95% CI 0.004-0.141) and opioid administration (P = 0.005; OR 0.157, 95% CI 0.044-0.567). From our literature search, 194 records were identified and 18 records were included (three comparative and 16 non-comparative). Meta-analysis of the comparative studies and our institutional data showed similar success rates (relative risk [RR] 1.77, 95% CI 0.33-9.52) and complication rates (RR 0.75, 95% CI 0.20-2.82). VH was associated with shorter operative time (standardised mean difference [SMD] -1.65, 95% CI -2.58 to -0.72 h) and hospital stay (SMD -1.41, 95% CI -2.36 to -0.47 days). The VH success rates ranged from 87.5% to 100% in the identified studies. Failure of VH was associated with unrecognised concomitant intrinsic obstruction in addition to CV. CONCLUSIONS: A VH, for well-selected patients with CV without concomitant intrinsic obstruction, offers a high success rate with shorter operative times.