A Richter transformation PD-1 block party

In this issue of Blood, Ding et al demonstrated anti–programmed death 1 (anti–PD-1) responses that may benefit chronic lymphocytic leukemia (CLL) patients with Richter transformation (RT) but did not show activity in non-RT CLL patients.¹ 

Since the identification of RT by Maurice Richter in 1928, and the subsequent description of Richter syndrome as the progression of CLL to a “malignant reticulopathy” by Lorthoraly et al in 1964, management of these patients has remained challenging. Approximately 10% of CLL patients will develop a RT, primarily representing a clonal CLL transformation that typically leads to a highly aggressive diffuse large B-cell lymphoma.²  The aggressive nature of RT appears to be associated, at least in part, with p53 altered dysfunction. The vast majority of these transformed lymphoma cells have undergone somatic hypermutation, display post–germinal center cell markers (ABC subtype), and carry Notch 1 mutations, all well-recognized high-risk features.³ 

The challenge has been that most patients display limited responses to standard chemoimmunotherapeutic regimens with a limited median survival of ∼8 to 12 months.⁴  In addition, use of novel targeted agents such as Bruton tyrosine kinase inhibitors (BTKi) or phosphatidylinositol 3-kinase inhibitors have not shown even incremental improvements in outcomes for RT patients. While recent work using a second-generation BTKi (acalabruitinib) in a cohort of RT patients resulted in a nearly 40% response rate, progression-free survival and durations of remission were still only a handful of months, and thus, we remain in dire need of meaningful therapeutic agents for RT patients.⁵ 

In this issue of Blood, Ding et al explored PD-1 checkpoint blockade in 25 patients with relapsed CLL, including 9 with RT. Patients were treated with a single-agent PD-1 inhibitor, pembrolizumab, in a phase 2 clinical trial in which significant responses were seen in 4 of the 9 with RT. Perhaps equally important, however, was that none of the 16 relapsed CLL patients without RT responded to anti–PD-1 therapy per International Workshop on Chronic Lymphocytic Leukemia 2008 criteria. Specifically, no tumor responses lasting more than 2 or 3 months were seen among those who enrolled after immediate prior chemotherapy, yet all 4 RT patients who responded had ibrutinib just prior to receiving pembrolizumab. Although it was a small number of RT patients studied, these preliminary data suggest that immediate prior ibrutinib may be associated with the observed clinical responses to PD-1 blockade.

Despite these promising data, it is unclear which RT patients will respond to PD-1 blockade. The clinical response to PD-1 blockade did not correlate with PD-1 expression on circulating T cells or CLL cells, as the expression of PD-1 in peripheral blood circulating lymphocytes did not differ significantly among RT patients with confirmed responses compared with those without a response, although variable PD-1 expression was detected in individual patients. Based on the work by Ding et al, the clinical responses to PD-1 blockade also could not be correlated with the baseline peripheral blood T-cell frequency or absolute number.

A small number of the 16 relapsed CLL patients without RT had a brief nodal tumor reduction that was maintained for <2 months, but all 16 patients had continuous marrow CLL progression. This striking lack of efficacy in human CLL seems perplexing given the promising anti–PD-1 activity in a well-established CLL Tcl-1 transgenic mouse model.⁶  As the authors suggest, one hypothesis has been that CLL and RT patients may generate different tumor-specific antigens and tumor-reactive T cells. Perhaps RT patients only recognize RT tumor antigens and not CLL antigens due to a manipulated T-cell immunity because of the long malignant cell presence in CLL patients. Such tumor-reactive T cells may only recognize CLL-specific antigens as self-antigens. Conversely, the RT phase of the tumor has been newly generated and may still be recognized as foreign entities by host T-cell immunity, as has been shown with neoantigens detected in lung cancer and melanoma tumors.⁷  This may be different from true human CLL and RT biology, where the human CLL and RT tumors are situated in much more complex tumor microenvironments consistently exposed to different cellular components and infectious antigens. Human CLL has been known to be significantly more complex, as multiple tissue sites, including lymph nodes, bone marrow, peripheral blood, and spleen are commonly involved with disease. However, studies using the CLL Tcl-1-tg mouse model have shown predominant disease in the murine spleen with less located in lymph nodes. Moreover, current knowledge with regard to major immune subsets in different human and mouse tissue sites has not been well elucidated. It remains unknown if humans have similar immune subsets in different tissues compared with mouse tissue sites. It may be possible that tumor-reactive T cells to CLL antigens are present in Tcl-1-tg mouse spleen but may not be present in human lymph nodes or other tissues. Taken together, it may not be surprising that the immunotherapy results observed in this relevant CLL mouse model do not recapitulate the human disease.

Despite the promise of PD-1 blockade in RT patients, this approach did not prevent the progression of underlying clonally related CLL, suggesting that these interesting anti–PD-1 findings might be best translated to combination therapies for RT/CLL patients. To this end, early work in a pilot study from the MD Anderson Cancer Center using BTKi therapy with PD-1 blockade has been ongoing and may add to these results.⁸  Nonetheless, to establish further advances for RT therapy, there remains a great need for a better understanding of the mechanisms involved in the development of RT. The important work described by Ding et al may be the initial step by demonstrating single-agent PD-1 blockade can benefit RT patients with modest improvements of survival. While other early-phase trials are ongoing, these results clearly must be confirmed in a larger phase 3 study testing the efficacy of checkpoint inhibition in RT patients. Overall PD-1 blockade may finally be the next move toward significant progress in the treatment of RT since conventional chemoimmunotherapy.