Diffuse large B-cell lymphoma (DLBCL), the most common form of non-Hodgkin lymphoma, is aggressive but curable in a majority of those diagnosed with the disease.1 Over the last five years, there has been an influx of new treatment approvals for both untreated and relapsed/refractory (R/R) DLBCL,2-8 with chimeric antigen receptor T-cell (CAR-T) therapy likely the most impactful to the field.6-8 Originally approved for third-line or later (3L+) R/R DLBCL, the three agents have provided a lifeline for those with difficult-to-treat disease, with long-term follow-up demonstrating cure. All three agents have been explored in pivotal phase III studies.9-11 The results of the Zuma-7 and TRANSFORM trials subsequently catapulted and entrenched both axicabtagene ciloleucel (axi-cel) and lisocabtagene maraleucel (liso-cel) into the second-line (2L) space for patients with refractory disease.9-10 While patients aged 75 and younger with relapsed disease remain candidates for intensive chemo-immunotherapy (CIT) followed by autologous stem cell transplantation (ASCT) consolidation, preference “creep” has also begun in this space, with some preferring the CAR-T products approved for those deemed too frail for intensive CIT.
The efficacy of CAR-T therapy is without question, but the cost, limited accessibility, and logistical requirements needed to administer these treatments remain major obstacles to expanding their reach. For most patients, the logistical requirements begin with insurance approval and end approximately six to 10 weeks later, with receipt of the cellular product. This time gap will generally require other treatments to “bridge” to this end goal, especially in those with aggressive disease. While most will not experience any substantial response to bridging therapy, some will achieve complete metabolic response (CMR), which brings into question the next best step. Some have argued that proceeding to CAR-T remains the best option for these patients, which is supported by data indicating excellent outcomes in those who receive CAR-T while in CMR. Others have advocated for ASCT in these patients, given that sensitivity to CIT has generally been a good barometer of long-term response to ASCT. In the absence of a truly randomized prospective trial, some researchers have evaluated retrospective data to help address this question.
For the present study,12 Mazyard Shadman, MD, MPH, and colleagues used data from the Center for International Blood and Marrow Transplant Research. They retrospectively evaluated outcomes for adult patients with DLBCL treated at member institutions between 2015 and 2021 with 2L+ ASCT or between 2018 and 2021 with any commercially approved CAR-T while in complete response (CR) determined by CT or PET scan and per the 2014 Lugano classification for initial evaluation, staging, and response assessment. The primary endpoints of the study were progression-free survival (PFS) and overall survival (OS). PFS was defined as the time from either ASCT or CAR-T to relapse, progression, or death from any cause.
The authors identified 360 patients with DLBCL, of which 281 had received ASCT while in CR and 79 had received CAR-T while in CR. The median follow-up for survivors was 49.7 months for the ASCT group and 24.7 months for the CAR-T group. Those who received ASCT had a two-year PFS of 66.2% (95% CI 60.4-71.8) versus 47.8% (95% CI 36.4-59.4, p<0.001) for those who received CAR-T. Those who received ASCT also had a lower two-year relapse/progression rate of 27.8% (95% CI 22.6-33.4), compared with a rate of 48% (95% CI 36.4-59.7, p<0.001) for those who received CAR-T, as well as a two-year OS of 78.9% (95% CI 73.9-83.6) versus 65.6% (95% CI 53.6-76.6, p=0.037), respectively. No significant difference in non-relapse mortality was noted between the two groups.
With respect to patients with primary refractory disease or early relapse (within 12 months), the data again indicate a greater benefit for those who received ASCT as compared to CAR-T, with a two-year PFS of 68.2% (95% CI 61.1-74.9) versus 48.4% (95% CI 35.7-61.3, p=0.001), respectively, as well as a two-year relapse/progression rate of 25% (95% CI 18.8-31.7) versus 46.3% (95% CI 33.5-59.3, p<0.001), respectively. No statistically significant difference in OS was noted between the two treatments in this group of patients. In a multivariable regression analysis, treatment with ASCT was associated with a lower rate of relapse/progression (hazard ratio [HR]= 2.18; 95% CI 1.48-3.2; p<0.0001) and a superior PFS (HR=1.83; 95% CI 1.27-2.63; p=0.0011) compared to CAR-T. In this analysis, no significant difference was noted between ASCT and CAR-T with respect to non-relapse mortality (HR=0.59; 95% CI 0.19-1.83; p=0.36) and OS (HR 1.44; 95 CI 0.91-2.28; p=0.12).
In Brief
The authors concluded that there remains a role for ASCT in chemo-sensitive R/R DLBCL based on the noted PFS and OS in the clinical study. This raises several questions about why the use of ASCT has declined substantially since the approval of 2L CAR-T. One is how applicable the study results are to our patient populations. Since the cutoff used for the study population, there has been approval of two CAR-T agents in patients with primary refractory or early relapse (relapse within 12 months of frontline therapy). As these patients were excluded from this analysis, the study cannot answer whether ASCT holds the same advantage for patients in CR as compared with 2L CAR-T. Overall, ASCT has the same issues as CAR-T regarding restriction to select centers, logistical requirements, and increased short-term toxicity. Additionally, while not formally powered to evaluate this question, the results of both Zuma-7 and TRANSFORM13 seem to indicate that 2L CAR-T outperforms ASCT even in those who are in CR at the time of receiving either treatment modality. The second question, which might hint at the need for an update to this analysis, is what role the availability of CAR-T products may have played in influencing the study results, as all approved products experienced longer time from approval to apheresis to product receipt than what we currently experience. Further, even though we do not have head-to-head efficacy data for the CAR-T products, almost none of the patients received liso-cel. Lastly, the study results warrant the undertaking of a more formal prospective trial of CAR-T versus ASCT in patients who are in a complete or true partial remission prior to receipt of either agent. Given the financial toxicity of CAR-T therapy to the health care system, could we “park the CAR” in some patients (those who obtain a CR to bridging therapy), potentially saving CAR-T therapy for those who actually need the treatment as opposed to the current drive to provide CAR-T therapy to all those who are eligible?
Disclosure Statement
Dr. Phillips indicated no relevant conflicts of interest.
