Endovascular Embolization of Left-Sided Varicocele: Outcomes by Venographic Anatomy and Coil-Adjunct Strategy.

PURPOSE: To evaluate outcomes after endovascular embolization for left-sided varicocele, with attention to venographic anatomy and embolic agent selection.

MATERIALS AND

METHODS

This retrospective single-center cohort included 46 men treated by endovascular embolization for symptomatic isolated left-sided varicocele between May 2018 and September 2025. The median age was 27 years (interquartile range [IQR], 22-32 years), and the median follow-up was 35 months (IQR, 22.5-70.0 months). Color Doppler ultrasound (US) confirmed pathologic reflux (>2 seconds during Valsalva). Venographic anatomy was classified using a 5-category scheme (Types 1-5) adapted from Jargiełło et al (Acta Radiol 2014; 56:63). Two embolic strategies were applied at operator discretion within a uniform distal-to-proximal protocol: distal coils plus proximal n-butyl cyanoacrylate glue vs distal coils plus proximal polidocanol foam. The primary outcome was recurrence (clinical and/or color Doppler ultrasound [US] defined); secondary outcomes were technical success, clinical success, and adverse events (Society of Interventional Radiology [SIR]/Baerlocher classification).

RESULTS

Technical success was 100% and clinical success was 97.8% (45/46). One recurrence (2.2%) required reintervention. Minor adverse events occurred in 2 of 46 patients (4.3%); no major adverse events occurred. Complex anatomy (Types 3-5) was present in 56.5%. In exploratory comparisons, clinical success was 94.7% with coil + glue vs 100% with coil + polidocanol (P = .41); recurrence and adverse event rates did not differ significantly by embolic strategy or anatomic complexity. The very low event rate precluded meaningful assessment of equivalence.

CONCLUSIONS

Endovascular embolization for left-sided varicocele achieved high technical and clinical success with low recurrence, including in patients with complex anatomy and prior surgical varicocelectomy. The absence of detectable between-group differences should not be interpreted as equivalence.