When meeting with patients and families of those affected by aggressive B-cell lymphoma, an important part of the conversation is letting them know what to expect. We have tools such as the International Prognostic Index that draw on clinical and laboratory measurements (eg, lactate dehydrogenase, age, and performance status). Additionally, studies are aiming to advance our molecular profiling capabilities so that molecular profiling might improve the personalization of treatment. 

The authors of abstract 2118 set up a national network on behalf of the Lymphoma Study Association called RT3 (real-time tailored therapy) to better characterize aggressive B-cell lymphomas. Patient samples went through conventional pathological analysis and some mutational testing (eg, MYD88, SOCS1, and MYC). Ultimately, this type of approach may allow us to combine real-time pathological and molecular information prior to R-CHOP (rituximab, cyclophosphamide, doxorubicin hydrochloride, vincristine, and prednisone) to inform subsequent treatment, potentially to include therapies that are more specifically targeted to the particular lymphoma biology identified in the mutational analysis.  

So, pathologic assessment of lymphoma is advancing from antibody-based immunohistochemistry to molecular characterization. The identification of molecular disease subsets prior to immunotherapy and the use of subsequent circulating tumor DNA (ctDNA) liquid tumor assessment may improve our ability to individualize treatment. At Stanford, we have been prospectively collecting both cellular and plasma samples from patients who receive chimeric antigen receptor (CAR) T-cell therapy. And, at ASH 2019, we reported that baseline and interim ctDNA-based minimal residual disease analyses, using the clonoSEQ platform, had prognostic significance in patients with large-cell lymphoma being treated with axicabtagene ciloleucel. 

This year, we continued to investigate our ability to monitor at a distance via plasma-based analyses. At ASH 2020, Shukla and colleagues presented the findings of the SABER (Sequence Affinity Capture & Analysis by Enumeration of Cell-Free Receptors) technique, a novel and sensitive technology that captures disease-specific sequences and could improve the sensitivity of immune receptor analysis compared with analyzing bulk DNA (abstract 199). In known T-cell lymphoma, the authors were able to detect clonal TCR-β rearrangement in 8 out of 9 cases, which was very promising. They then studied the T-cell repertoire dynamics using cell-free DNA in the plasma, and they observed what appeared to be favorable T-cell repertoire expansion as early as 1 week after CAR T-cell therapy. Although this study is too small to conclude that there is reliable predictive utility, the authors did show that TCR repertoire size 1 month after axicabtagene ciloleucel infusion was significantly associated with longer progression-free survival. I believe that monitoring immune response enhancement through TCR characterization is a promising technique deserving of prospective testing, and it may be an important way to evaluate immune-modulating therapies in cancer, in general. 

Advanced imaging also has potential utility in the study of responses to immunotherapy. In abstract 3255, Simonetta et al utilized ⁸⁹Zr-DFO-ICOS  monoclonal antibody as a cell surface marker, which can be detected by immuno-positron emission tomography (immunoPET) in an animal model system. Not only will PET scans identify where the tumor is but we may also be able to use the same platform to determine where the immune response is, which would provide us with insights into the pathways for success with immunotherapy as well as potential mechanisms of failure. After administering ⁸⁹Zr-DFO-ICOS monoclonal antibody 5 days post CAR T-cell administration, PET-computed tomography imaging performed 48 hours later found specific and sensitive detection of CAR T-cell migration and expansion in the animal model system. 

Based on these abstracts, it is clear that we will be pursuing the molecular characterization of a patient’s pathologic samples before treatment to improve prognosis. By linking this to the use of cell-free DNA to monitor disease burden, we may identify patients who require additional therapies, identify those who remain in variable responses, and improve our understanding of the biology of treatment failure following targeted therapies. While these are stepwise and incremental improvements over the early advancements with ctDNA, they demonstrate continued and steady progress in molecular profiling.