A meta-analysis of clinicopathologic features that predict necrosis or fibrosis at post-chemotherapy retroperitoneal lymph node dissection in individuals receiving treatment for non-seminoma germ cell tumours

PURPOSE: Post-chemotherapy retroperitoneal lymph node dissection (pcRPLND) for residual nodal masses is a critical component of care in metastatic testicular germ cell tumour (GCT). However, the procedure is not of therapeutic value in up to 50% of individuals in whom histopathology demonstrates post-treatment necrosis or fibrosis alone. Improved diagnostic tools and clinicopathologic features are needed to separate individuals who benefit from pcRPLND and avoid surgery in those who do not. METHODS: A prospectively registered meta-analysis of studies reporting clinicopathologic features associated with teratoma, GCT and/or necrosis/fibrosis at pcRPLND for metastatic non-seminoma GCT (NSGCT) was undertaken. We examined the effect of various clinicopathologic factors on the finding of necrosis/fibrosis at pcRPLND. The log odds ratios (ORs) of each association were pooled using random-effects models. RESULTS: Using the initial search strategy, 4,178 potentially eligible abstracts were identified. We included studies providing OR relating to clinicopathologic factors predicting pcRPLND histopathology, or where individual patient-level data were available to permit the calculation of OR. A total of 31 studies evaluating pcRPLND histopathology in 3,390 patients were eligible for inclusion, including two identified through hand-searching the reference lists of eligible studies. The following were associated with the presence of necrosis/fibrosis at pcRPLND: absence of teratomatous elements in orchidectomy (OR 3.45, 95% confidence interval [CI] 2.94-4.17); presence of seminomatous elements at orchidectomy (OR 2.71, 95% CI 1.37-5.37); normal pre-chemotherapy serum bHCG (OR 1.96, 95% CI 1.62-2.36); normal AFP (OR 3.22, 95% CI 2.49–4.15); elevated LDH (OR 1.72, 95% CI 1.37-2.17); >50% change in mass during chemotherapy (OR 4.84, 95% CI 3.94-5.94); and smaller residual mass size (<2 cm versus >2 cm: OR 3.93, 95% CI 3.23-4.77; <5 cm versus >5 cm: OR 4.13, 95% CI 3.26-5.23). CONCLUSIONS: In this meta-analysis, clinicopathologic features helped predict the presence of pcRPLND necrosis/fibrosis. Collaboration between centres that provide individual patient-level data is required to develop and validate clinical models and inform routine care to direct pcRPLND to individuals most likely to derive benefits. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/prospero/, identifier CRD42021279699