Nerve-oriented extracapsular dissection versus superficial parotidectomy in selected benign parotid tumours: a retrospective cohort study with propensity score–matched analysis

Abstract Background Extracapsular dissection (ECD) is proposed as a less invasive alternative to superficial parotidectomy (SP) for clinically benign parotid tumours. Because the choice of technique is driven by tumour characteristics, it is uncertain how far reported differences reflect the operation rather than the indication. We reviewed 146 consecutive patients operated on for benign parotid tumours between 2015 and 2023 (SP 104, ECD 42), recording complications, recurrence and aesthetic satisfaction on a visual analogue scale (VAS), with facial nerve function graded by House–Brackmann. Groups were compared without adjustment, by multivariable logistic regression and by 1:1 propensity score matching; post hoc exploratory analyses stratified by histology and by operating surgeon were added after peer review. Results Recurrence occurred in 16/104 (15.4%) after SP and 1/42 (2.4%) after ECD ( p = 0.025); transient facial nerve dysfunction showed the same figures ( p = 0.025), with no permanent dysfunction. Fistula, Frey syndrome and sialocele were less frequent after ECD without reaching significance, and VAS did not differ ( p = 0.173). Tumour size was associated with higher odds of Frey syndrome (OR 2.95; p = 0.028). Matching produced 38 pairs with residual imbalance in five covariates; recurrence (1/38 vs. 6/38; p = 0.25) and facial nerve dysfunction (1/38 vs. 4/38; p = 0.75) did not differ. In post hoc analyses the recurrence difference was confined to Warthin tumour (0/15 vs. 10/43; p = 0.050) and absent in pleomorphic adenoma (1/22 vs. 6/56; p = 0.67). One surgeon performed all 42 ECD and 45 SP procedures; within his own cases no difference between the techniques was detected (1/42 vs. 3/45; p = 0.62), whereas recurrence varied substantially between surgeons (4/87, 4.6% vs. 13/59, 22.0%; p = 0.003). Conclusion No statistically significant difference between the techniques was detected after adjustment or matching. The crude recurrence difference was confined to the Warthin subgroup, while the ECD arm was completely confounded with operating surgeon and with the selection of favourable tumours. These data establish neither superiority nor equivalence of either technique, and indicate that the extent of parotid surgery should be tailored to the individual lesion.