Radiological dynamics and SITC-defined resistance types of advanced melanoma during anti-PD-1 monotherapy: an independent single-blind observational study on an international cohort

Abstract

Background: Although the Society for Immunotherapy of Cancer (SITC) Immunotherapy Resistance Taskforce recently defined primary and secondary resistance to anti-programmed cell death protein 1 (anti-PD-1) therapy, there is lack of real-world data regarding differences in these resistance subtypes with respect to radiological dynamics and clinical manifestations.

Methods: We performed single-blind re-evaluations of radiological images by independent radiologists on a retrospectively assembled cohort of patients with advanced melanoma (n=254, median follow-up 31 months) receiving anti-PD-1 monotherapy at Massachusetts General Hospital and Peking University Cancer Hospital. Radiological characteristics and timing at multiple crucial time points were analyzed and correlated with each other and with survival. Primary and secondary resistance was defined as per the SITC Immunotherapy Resistance Taskforce definitions.

Results: The most significant target lesion measurement change took place within the first 3 months after anti-PD-1 initiation. Patients with stable disease with versus without tumor shrinkage at the initial evaluation exhibited distinct disease trajectory, as the rate of further upgrade to a partial or complete remission (CR/PR) was 44% and 0%, respectively. Eleven per cent of PR patients ultimately achieved a CR. In multivariate analyses, deeper response depth was independently associated with a more limited progression pattern, fewer involved organs, lower tumor burden, slower growth rate at disease progression (PD) (all p≤0.001), and longer post-progression survival (PPS) (bivariate analysis, p=0.005). Compared with primary resistance, secondary resistance was associated with less widespread PD pattern, lower tumor burden and slower tumor growth (all p≤0.001). Patients with secondary resistance were less likely to receive further systemic therapy (28% vs 57%, p<0.001) yet had significantly better PPS (HR 0.503, 95% CI 0.288 to 0.879, p=0.02).

Conclusions: Radiological dynamics were variable, yet significantly correlated with survival outcomes. SITC-defined primary and secondary resistance are distinct clinical manifestations in patients with melanoma, suggesting the possibility of resistance-type-based therapeutic decision-making and clinical trial design, once further validated by future prospective studies.

Keywords: immunotherapy; melanoma; programmed cell death 1 receptor.

Figure 1

Patient populations included and general research schema of this study. CR, complete remission; PD, disease progression; PR, partial remission; SD, stable disease.

Figure 2

Radiological dynamics and progression patterns. (A) Spider plot for the whole cohort (n=215). Within the entire cohort, 15 (7.0%) patients achieved complete remission (CR); 64 (29.8%) and 40 (18.6%) patients had partial remission (PR) and stable disease (SD) as their best response, respectively; 96 (55.7%) patients experienced disease progression (PD) without clinical benefit. The most drastic change in tumor size occurred within the first 3 months after anti-programmed cell death protein 1 (anti-PD-1) monotherapy initiation. (B) Tumor percent change at 3 months after anti-PD-1 monotherapy initiation (n=215). The median tumor percent change at 3-month evaluation from baseline was −69.6% (range −100% to −22.7%) in patients with CR as their best response, −36.6% (range −75.7% to −2.4%) in patients with PR as the best response, and −1.4% (range −29.6% to 17.6%) in SD patients. (C) Baseline tumor measurement distribution between different best response groups (n=215). The median baseline target lesion size of patients who had CR/PR/SD/PD as their best response was 19.3 mm (range 10.1 to 49.0 mm), 65.5 mm (range 10.0 to 255.4 mm), 48.2 mm (range 10.2 to 186.8 mm) and 46.0 mm (range 10.0 to 305.0 mm), respectively. (D) Resistance type and post-progression survival (PPS) (n=166). PPS of patients with primary resistance was significantly shorter than those developed secondary resistance (p=0.009), with median PPS of 10.3 months (95% CI 7.7 to 16.1) and not reached (95% CI 11.8 to not reached), respectively.