Dr. Jacobson also noted that bridging therapy may have affected these results, as it might reduce tumor burden and increase toxicity, but bridging therapy was not used uniformly across these trials.
Most patients received bridging therapy before t-cel, none received it before axi-cel, and the use of bridging therapy was not reported in the trial of liso-cel.
“It is not possible to know whether patients treated on the ZUMA-1 trial, who were more likely to receive their CAR T cells, were healthier and more fit than patients on other studies or, because they were not allowed to receive bridging therapy, were actually sicker with a higher tumor burden and were therefore at risk for greater toxicity,” Dr. Jacobson wrote.
The fact that ZUMA-1 patients were more likely to receive CAR T cells brings up another issue—the difference between the reported results and the intent-to-treat (ITT) results in these trials. Since most patients on ZUMA-1 received the study treatment, there isn't much difference between the reported results and ITT results. However, about a third of patients who underwent apheresis on the JULIET trial did not ultimately receive CAR T cells, which means a bigger difference between the reported results and ITT results.
In ZUMA-1, 111 patients underwent leukapheresis, and 101 received treatment with axi-cel and were evaluable for efficacy. So the ORR was 75% (83/111) in the ITT population, compared to 82% in the population evaluable for efficacy.
In JULIET, 165 patients underwent leukapheresis, 111 received t-cel, and 93 were evaluable. The ORR was 30% (48/161) in the ITT population, compared to 52% in the evaluable population.
In TRANSCEND-NHL-001, 134 patients underwent leukapheresis, 114 patients received liso-cel, and 102 were evaluable. The ORR was 63% (77/122) in the ITT population, compared to 75% in the evaluable population.
Dr. Jacobson said these differences can be explained, in part, by differences in manufacturing. The time to manufacture cells was longer on the JULIET trial than on ZUMA-1, which may have been due to differences in transfection and manufacturing procedures as well as manufacturing ability.
In addition, differences in patient eligibility may have played a role, as healthier patients might be able to tolerate a longer manufacturing period than sicker patients.
Unfortunately, these differences cannot be accounted for without a randomized trial, but Dr. Jacobson said a randomized trial of these therapies is unlikely to occur.
“[S]o perhaps the best answers will come from institutions that have experience with all three products,” she wrote. “And in these cases, physicians and institutions will have to decide to what extent they would sacrifice efficacy for improved safety or sacrifice safety for improved reliability and consistency of treatment delivery.”
Dr. Jacobson disclosed relationships with Kite Pharma/Gilead Sciences, Bayer AG, Pfizer, Precision BioSciences, Novartis, Celgene, and Cowen.
SOURCE: Jacobson CA. J Clin Oncol. 2019 Feb 1;37(4):328-35. doi: 10.1200/JCO.18.01457