Chimeric antigen receptor (CAR) T-cell therapy has revolutionized the treatment of B-cell lymphoma. Two phase III studies, ZUMA-71 and TRANSFORM,2 have demonstrated the superiority of CD19-directed CAR T-cell therapy in second-line large B-cell lymphoma (LBCL) compared to the prior standard of care (SOC), leading to the approval of axicabtagene ciloleucel (axi-cel) and lisocabtagene maraleucel (liso-cel), respectively. In fact, axi-cel not only showed an increase in event-free survival (EFS) compared to SOC salvage chemoimmunotherapy and autologous stem cell transplantation, but primary overall survival (OS) analysis also demonstrated significant benefit in favor of CAR T-cell therapy.3 This benefit spanned many poor prognosis subgroups, including patients with elevated lactate dehydrogenase (LDH) and those with high tumor burden as defined by the sum of the product of perpendicular diameters.4
High tumor burden is a well-known risk factor for poor outcomes in lymphoma.5 With the advent of 2-deoxy-2-[18F]fluoro-D-glucose (FDG) positron emission tomography-computed tomography (PET-CT), the ability to estimate metabolic tumor volume (MTV) has emerged. This may represent one of the most precise tools available to estimate the burden of viable lymphoma.6 To date, several studies have demonstrated the promising prognostic value of MTV in LBCL, including in patients receiving third-line CAR T-cell therapy for LBCL.7-11 In the present study, Frederick L. Locke, MD, and colleagues demonstrated that baseline MTV above the median portended a poor prognosis among ZUMA-7 patients with LBCL treated with either SOC or CAR T-cell therapy. High-baseline MTV was also predictive of higher rates of severe cytokine release syndrome (CRS) and neurologic events. Despite these findings, however, the use of axi-cel still achieved better outcomes compared to SOC treatment, regardless of MTV.
Baseline PET-CT scans of ZUMA-7 patients were uploaded to a central site and analyzed for MTV. Of the 359 randomized patients, 340 were evaluable for MTV. Median MTV was 231.07 mL (range: 0.04-16,669.3) with high MTV defined as that greater than the median. EFS and progression-free survival (PFS) were inferior among axi-cel–treated patients with high MTV compared to those with low MTV (hazard ratio [HR] 1.448; 95% CI, 0.980-2.139 and HR 1.660; 95% CI, 1.097-2.513). However, among high-MTV patients, axi-cel remained superior to SOC (HR 0.421; 95% CI, 0.286-0.619 for EFS and HR 0.501; 95% CI, 0.324-0.773 for PFS). For each doubling of MTV, there was a 10% increase in risk of EFS events associated with axi-cel (p=0.02) and a 14% increase associated with SOC (p=0.01). In multivariate analysis, with high and low levels of MTV based on optimal threshold, MTV was predictive of both EFS and PFS after adjustment for second-line, age-adjusted International Prognostic Index (IPI); age; and LDH.
MTV was also predictive of toxicity among ZUMA-7 patients treated with axi-cel. The median MTV was comparatively greater among patients who developed grade 3 or higher CRS and neurologic events. Interestingly, MTV did not correlate with CAR T-cell peak or area under the curve (AUC), suggesting that CAR T-cell therapy expansion may not be directly linked to metabolic tumor burden and that a lack of CAR T-cell therapy expansion does not explain the differences in outcomes among high-MTV patients. Dr. Locke and his colleagues demonstrated the utility of MTV as a prognostic marker for efficacy among patients with relapsed or refractory (R/R) LBCL undergoing second-line therapy. High MTV was also associated with severe toxicity among patients treated with axi-cel. This opens the door to investigating whether the poor prognosis of high-volume LBCL could be overcome with debulking therapy prior to CAR T-cell therapy or additional consolidation after therapy. The study findings also point to the potential to prevent high-grade toxicity by targeting high-MTV patients with prophylactic maneuvers.
In Brief
With the advent of CAR T-cell therapy and its demonstrated efficacy for R/R LBCL in the second-line setting, questions arise regarding factors that predict CAR T-cell therapy outcomes. While several studies have identified factors that predict CAR T-cell therapy outcomes, most data comes from the third-line and beyond setting.12,13 Additionally, factors that typically represent surrogates for tumor volume such as high LDH often correlate with poor outcomes. Calculating MTV harnesses the important metabolic data provided by PET-CT scans about LBCL. As demonstrated by the study authors, MTV calculations also correlate well with outcomes in second-line LBCL. Although ZUMA-7 patients with high MTV had inferior outcomes, those who received axi-cel still benefited compared to those receiving SOC.
In addition to predicting survival outcomes, MTV also appears to correlate well with severe toxicities, including grade 3 or higher CRS and neurologic events. This appeared to occur independent of CAR T-cell therapy pharmacokinetics and pharmacodynamics; it may instead be representative of the compromised immune function of patients with extensive LBCL volumes. Although ongoing studies are investigating prophylaxis for the prevention of CRS and neurologic toxicities (including the ZUMA-24 study, which uses prophylactic dexamethasone),14 risk-adapted interventions have been lacking. Using high MTV and other predictors of toxicity may identify a subgroup of patients with a greater need for or potentially greater benefit from novel prophylactic strategies. Additionally, given the poor outcomes associated with high MTV, identifying treatments that could overcome this detriment — including novel bridging, debulking, or even consolidation strategies — could potentially improve outcomes in this high-risk subgroup. In the absence of such treatments, however, the study data suggests that CAR T-cell therapy represents the best available option in this setting.
Competing Interests
Dr. Herrera has received research funding from Kite, a Gilead company. Dr. Shouse has received honoraria from Kite for consulting and as a member of a speaker’s bureau.
