Use of <scp>PD</scp>‐1 (<scp>PDCD</scp>1) inhibitors for the treatment of Richter syndrome: experience at a single academic centre

Abstract

An aggressive lymphoma in a patient with chronic lymphocytic leukaemia (CLL), known as Richter syndrome (RS), has a poor prognosis and is an important unmet clinical need (Richter, 1928; Parikh et al, 2014). Standard treatment with chemoimmunotherapy is unsuitable for many patients who are unfit for these intensive regimens or have disease features predicting poor response, such as complex CLL karyotype or relapse after prior chemoimmunotherapy, leaving no good options for these patients. (Tsimberidou et al, 2006; Rossi et al, 2011; Parikh et al, 2014; Rogers et al, 2018). Recently, the programmed death receptor-1 (PD-1, also termed PDCD1) blocking antibodies, pembrolizumab and nivolumab, have shown efficacy in prospective clinical trials with a small number of RS patients. A phase 2 trial of pembrolizumab for relapsed CLL and RS included 9 RS patients with an overall response rate of 44% and median progression-free survival of 5·4 months (Ding et al, 2017). Additionally, early results from two trials with combination ibrutinib and nivolumab reported responses, generating enthusiasm for these agents (Jain et al, 2016; Younes et al, 2017). Both nivolumab and pembrolizumab are approved by the United States Food and Drug Administration for a variety of other cancers and where evidence of tumour mismatch-repair deficiency exists. Based on the reported responses, favourable toxicity profile and lack of other options, we have prescribed PD-1 inhibitors for RS off-label when other treatments were unavailable. As experience with these agents for RS in a non-trial population has not been reported and the number in prospective studies is small, we reviewed our institutional experience. Medical records of all patients with a diagnosis of RS treated outside of a clinical trial with either pembrolizumab or nivolumab were reviewed. All patients consented to an Institutional Review Board-approved protocol permitting this research. Baseline patient characteristics, disease features, treatment history, and outcomes were abstracted from medical records. Treatment failure was defined as disease progression, start of a new treatment or death while taking a PD-1 inhibitor. Allogeneic stem cell transplant was not considered treatment failure. Time to treatment failure was calculated from first dose of a PD-1 inhibitor to a treatment failure event, censoring event-free patients at last follow-up. Overall survival was calculated from both RS diagnosis and the first dose of a PD-1 inhibitor until death, censoring patients who were still alive at last follow-up. Time to treatment failure and overall survival were estimated using the method of Kaplan-Meier. Ten patients with biopsy-proven RS with diffuse large B-cell lymphoma (DLBCL) histology were included. The clonal relationship to the underling CLL was known in 2 cases and both were clonally related. Characteristics of the cohort are in Table 1. All patients were exposed to Bruton tyrosine kinase (BTK) inhibitors, and 8/10 (80%) were taking a targeted agent at RS diagnosis. The majority of patients (6/10, 60%) received a PD-1 inhibitor as initial therapy, and the median time from RS biopsy to first dose of PD-1 inhibitor was 27 days (range 13–86). Four patients had prior RS treatment: 1 R-EPOCH (rituximab, etoposide, prednisolone, vincristine, cyclophosphamide, doxorubicin), 1 rituximab and high-dose methylprednisolone, 1 radiation therapy, and 1 multiple treatments (hyper-CVAD [hyperfractionated cyclophosphamide, vincristine, doxorubicin, dexamethasone plus methotrexate and cytarabine], pulse dexamethasone, radiation therapy and single-agent rituximab). The PD-1 inhibitor given was nivolumab in 7/10 (70%) cases and pembrolizumab in 3/10 (30%). Both were dosed according to their US label. In 4/10 (40%) cases a targeted agent was given concurrently. This was ibrutinib in 3 cases and combination ibrutinib and venetoclax in 1 case. Therapy outcome was known for all 10 patients and median duration of follow up after the first dose of a PD-1 inhibitor in surviving patients was 14·3 months (range 8·5–20). The majority of patients (9/10, 90%) had treatment failure: 3 disease progression, 2 started a new therapy due to lack of response and 4 died during treatment. Median time to treatment failure was 1·2 months [95% confidence interval (CI): 0·1–1·8] (Fig 1). Median overall survival from RS diagnosis was 4·2 (95% CI: 1·1–8·4) months and from first dose of PD-1 inhibitor was 2 months (95% CI: 0·1–5·6). Only 2/10 (20%) patients were alive at last follow-up. Only 1 patient had a sustained course of a PD-1 inhibitor without treatment failure and underwent successful allogeneic stem cell transplant. This patient presented with a brain mass while taking ibrutinib for CLL and diagnostic complete surgical resection revealed DLBCL. There was no evidence of systemic disease. This patient received pembrolizumab as initial treatment for RS. He was alive and without evidence of disease at last follow-up 20 months after the first dose of pembrolizumab and 12·4 months after transplant. The clonal relationship of the RS to CLL in this patient is unknown. This patient had disease features distinct from the rest of the cohort in that the only site of RS was brain parenchyma which was completely resected and Epstein–Barr virus (EBV)-encoded small RNA staining was positive. These features may have contributed to the response to pembrolizumab and, theoretically, viral antigen expression could increase the efficacy of pembrolizumab by increasing immunogenicity of the lymphoma cells. Additionally, in a study in primary and secondary central nervous system DLBCL there was significant sensitivity to PD-1 blockade (Nayak et al, 2017). In summary, review of our institutional experience with PD-1 inhibitors for the treatment of RS demonstrates very poor efficacy with a short time to treatment failure (median 1·2 months). Our experience was in a diverse set of patients, which better approximates the reality of prescribing PD-1 inhibitors “off-label” in clinical practice. There are several features of our cohort which have the potential to influence these results, including lack of uniform treatment of included patients, relative enrichment for poor-risk CLL features, decreased patient fitness compared to a clinical trial population, prior BTK inhibitor exposure in all patients, and the majority (80%) taking a targeted agent when RS developed. This could influence both the RS biology and the immune environment. Despite these limitations, outcomes were poor and our data do not support the use of PD-1 inhibitors for RS outside of a clinical trial. The authors would like to thank the Ohio State University Division of Hematology for support of this project. They would also like to acknowledge their colleagues who provided the clinical care for the included patients. K.A.R., J.A.W., and J.C.B. designed the study. K.A.R., E.D., and J.L. collected the data. K.A.R. wrote the initial draft of the manuscript. Y.H. performed the statistical analysis and generated the figure. All authors cared for the included patients, interpreted the data, reviewed the manuscript, and agreed to its submission. K.A.R. receives research funding from Genentech. L.A.A. receives research funding from Sanofi. F.T.A. has consulted for AbbVie, Gilead Sciences and Janssen, and receives research funding from Pharmacyclics. J.A.W. received honoraria from Janssen, has consulted for Janssen, and receives research funding from MorphoSys, Karyopharm Therapeutics and AbbVie. J.C.B. receives research funding from Genentech, Acerta Pharma, Pharmacyclics, and Janssen. Y.H., E.D. and J.L. have no disclosures.