Pathological response calculation assessment remains accurate with reduced tumor bed examination following neoadjuvant immunotherapy in clinically detectable stage III melanoma

Background: Neoadjuvant immunotherapy produces event-free survival advantage over adjuvant therapy for patients with surgically resectable macroscopic stage IIIB/C/D melanoma. Pathological response, determined as percentage residual viable tumor (% RVT), provides critical prognostic information and informs management decisions. Here, we assessed accuracy of %RVT calculation when reduced tumor bed (TB) was examined and leverage these results proposing streamlined protocols for pathological examination.

Patients and methods: Comprehensive histopathological examination was carried out on 134 patient specimens after neoadjuvant immunotherapy with ipilimumab and nivolumab. Impact on %RVT when evaluating less TB than recommended by the initial International Neoadjuvant Melanoma Consortium (INMC) protocol was assessed. Firstly, % RVT of each case was recalculated using seven modified protocols and compared with % RVT obtained under INMC protocol. Next, a simulation study was carried out recalculating % RVT by random sampling 50%, 33%, and 25% of TB slides per specimen.

Results: There was excellent accuracy in %RVT (R² > 0.97) for all the modified protocols and >90% accuracy in five protocols. Accuracy of major pathological response (MPR)/non-MPR and pathological response category classification was ≥96% in six protocols. The decrease in average slides examined per specimen ranged from 9% to 58%. In total, 85%, 79%, and 74% of simulations recalculating %RVT were within 5% of the INMC calculation when 50%, 33%, and 25% of TB slides were examined, respectively. If TB slide examination is capped at 20, %RVT calculation remains 93% accurate.

Conclusions: TB embedded for histopathological examination in neoadjuvant stage IIIB/C/D melanoma specimens can be reduced without significantly compromising accuracy of %RVT calculation. We recommend an updated pathological assessment protocol: lymph nodes ≤3 cm examined in entirety; macroscopically involved lymph nodes >3 cm should have a modified examination protocol of at least a full cross-sectional transverse slice. Capping TB slides examined at 20 appears reasonable. This refined approach results in high accuracy and significant reduction in the slides examined.