Objectives: Pulmonary metastasectomy is not a standardised procedure, with no consensus regarding the optimal extent of lung resection. This international multicentre study aimed at comparing short and long-term outcomes of anatomical versus non-anatomical pulmonary metastasectomy. Methods: Retrospective database including 1647 patients aged ≥18 years, who underwent curative intent pulmonary metastasectomy between January 2010 and December 2018 at 15 European centres. Patients who underwent pneumonectomy, previous metastasectomies, and/or suffered from extrapulmonary recurrence at the time of lung surgery were excluded. Primary endpoint was overall survival. Secondary endpoints were recurrence-free survival and 30-day morbidity. Differences between the two groups were analysed using 3:1 matching. Results: In the matched cohort, 324 patients underwent anatomical resection, and 830 patients underwent non-anatomical resection. Five-year overall survival was 62.0%. Averaged over the entire follow-up, there was no significant difference in overall survival between the two groups (HR = 1.122, 95% CI = 0.909–1.385, p = 0.283). In the early period following pulmonary metastasectomy, anatomical resections were associated with worse overall survival (HR = 1.549, 95% CI = 1.135–2.114, p = 0.006). The difference in any-site recurrence-free survival between the two groups was not significant (HR = 0.832, 95% CI = 0.690–1.002, p = 0.053). Locoregional recurrence-free survival was significantly longer after anatomical resection (HR = 0.651, 95% CI = 0.520–0.817, p < 0.001). Thirty-day morbidity was significantly higher after anatomical resection (22.2% versus 13.7% for non-anatomical resections, p = 0.001). Conclusions: In a highly selected cohort, non-anatomical resection showed comparable survival and lower morbidity compared to anatomical resections, supporting the surgical strategy of favouring limited resections whenever technically and oncologically feasible. Anatomical resections remain a valid option in selected cases with acceptable outcomes.
Prisciandaro et al. (Mon,) studied this question.