B-cell Lymphoma and Non-Hodgkin's Lymphoma Treatment

Human gene therapy has evolved rapidly, and chimeric antigen receptor T-cell (CAR-T) therapy appears to be a promising treatment for B-cell lymphoma and non-Hodgkin's lymphoma. In 2017, the US Food and Drug Administration approved 2 CAR-T interventions, Kymriah (Tisagenlecleucel) and Yescarta (Axicabtagene ciloleucel), to treat adult relapsed or refractory B-cell lymphoma. Disease severity justified not performing controlled trials, and phase II trials provided the strongest evidence at the time for efficacy. A third CAR-T intervention, JCAR017 (Lisocabtagene maraleucel), is also under consideration for approval to treat adult nonHodgkin's lymphoma and B-cell lymphoma. However, it is unclear whether treatment response to CAR-T therapy is the same outside of trial settings, as study participants may differ in important ways from the general population with the same diagnosis. We sought to address this concern by comparing CAR-T trial and participant characteristics with a population-based sample of adult non-Hodgkin's lymphoma and B-cell lymphoma patients.

We reviewed study populations in CAR-T studies for treatment of adult B-cell lymphoma and non-Hodgkin's lymphoma. Although we searched for both experimental and observational studies, all identified studies were trials, predominantly single-arm trials. To describe the general population with comparable diagnoses for CAR-T, we used the US Surveillance, Epidemiology, and End Results (SEER) 18 1975-2016 data for patients diagnosed with non-Hodgkin's lymphoma and B-cell lymphoma between 2000 and 2015. We compared participant eligibility criteria for the research studies and characteristics of study participants against those of the SEER non-Hodgkin's lymphoma and B-cell lymphoma population.

Available trial data made it impossible to definitively compare demographics, comorbidities, or disease severity between trial participants and the SEER population. However, trial participants appeared to be younger and include more males compared with the SEER population. Furthermore, comorbidity in trial participants may have been lower than that of the SEER population, as most trials restricted participation by patients with comorbidities that would interfere with treatment or assessment. Trials also often enrolled only patients with relapsed disease that was refractory to prior treatment. Finally, the length of follow-up in trials was shorter than the average survival time of patients documented in the SEER population.

Treatment effectiveness among the general patient population may be lower than in research studies because trials were designed to demonstrate efficacy in patients with severe disease. CAR-T may also be less effective for patients in the general population, who may have more comorbidities than did trial participants. However, treatment may be more successful in general practice with patients whose disease is not as advanced as that observed in study participants. Although we found no published observational studies, newly established CAR-T patient registries aim to follow patients for longer than their untreated life expectancy. In the future, these registries may provide information about how treatment effectiveness might vary between trial participants and the general population. In the meantime, understanding differences between the trial populations and the comparable SEER population provides some insight into how these therapies will function in general practice.