Optimization of immune checkpoint inhibitor treatment planning for relapsed or refractory extranodal NK/T cell lymphoma

Pembrolizumab (anti-programmed cell death-ligand 1 inhibitor) is a promising salvage therapeutic option for relapsed/refractory extranodal NK/T-cell lymphoma (R/R ENKTL). However, the appropriate duration of pembrolizumab use in R/R ENKTL patients and the optimal timing for administering pembrolizumab remain undetermined. We collected and analyzed clinical information on R/R ENKTL 58 patients who received pembrolizumab to evaluate the optimal treatment durations and clinical information for considering treatment interruption. Treatment outcomes were assessed by 18F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) and Epstein Barr virus DNA (EBV DNA) every 3 months. Nineteen (32.8%) patients had been treated with more than three chemotherapies before pembrolizumab administration. The best response rate towards the first try of pembrolizumab was 38.9% (31.5% complete response rate (CR), 7.4% partial response (PR)). During the 41.8-month median follow-up duration, the median progression-free survival (PFS) was 3.1 months, and the median overall survival (OS) was 7.1 months. The failure group, which was characterized by Deaville score (DS) 3–4 and circulating EBV detection, or DS 5 with/without EBV detection, had the worst PFS (p < 0.001) and OS (p < 0.001), followed by the high (DS 1–2 and EBV detection, or DS 3–4 and EBV not detected) and low-risk groups (DS 1–2 and EBV not detected). Among the 21 patients who achieved the best response at the first pembolizumab try, the patients who received planned 24 cycles presented better PFS than those who received incomplete cycles (57.6 months vs 20.9 months, P-value = 0.012). Among 13 patients who received avelumab or pembrolizumab in advance, a few who responded to the second trial of pembrolizumab administration had over one year of chemotherapy vacation. Determining the discontinuation or continuation of pembrolizumab would be considered in selected cases assessed by PET-CT and EBV monitoring. Disruption of pembrolizumab treatment may be advisable for the low-risk group(DS 1–2 and EBV not detected), whereas continuation could be warranted for the high-risk group (DS 1–2 and EBV detection, or DS 3–4 and EBV not detected). Moreover, it might be critical to maintain over 24 cycles to improve the survival outcome of R/R ENKTL.