CAR-T Meeting 2026 | Using real-world evidence to shape CAR-T care in R/R DLBCL

Yeah, so I think that real-world evidence is very important to kind of support the data we have from clinical trials. So especially in the CAR T-cell setting, we have really good results, you know, in terms of efficacy with CAR T-cell therapy and relapsed refractory large B-cell lymphoma, but it’s very important to see that those results are reproduced outside of the trial setting when we’re treating, you know, a wider range of patients in terms of age, comorbidities, and prior lines of therapy. So in this sense, you know, we’re grateful there’s a lot of real-world evidence, data sets with hundreds of patients coming from the U.S. and coming from Europe, treated with axi-cel, liso-cel, and tisa-cel in the relapsed refractory large B cell lymphoma setting. And in this respect, you know, I think it’s also interesting to see how the different data sets have made an effort when they are comparing different constructs, such as axi-cel and liso-cel in these recent meetings, to try to balance baseline variables and try to make both populations, let’s call it comparable. So in the absence of a head-to-head randomized clinical trial, which would be the ideal way to compare different CAR-T constructs, we try to make, you know, let’s say, for example, the axi-cel and the liso-cel patients similar to some extent so that we can compare both their toxicity and efficacy outcomes. 

So at the EU CAR-T meeting this year in Mallorca, we discussed different ways from a statistical point of view that you can match these populations and therefore try to compare more robustly their outcomes. So we discussed the different ways, you know, with multivariable analyses, propensity score matching methodologies, and one of the ones that’s most commonly used currently in these data sets is the IPTW, or inverse probability treatment weighting, where you assign different weights to the patients in the different cohorts, and therefore you try to match, especially the high-risk features in terms of the rate of patients who are primary refractory, who have a high LDH, high-risk IPI score, and therefore try to make the results better in the sense of making them comparable between one data set and the other, and trusting those comparisons further. 

You know, there are still limitations in these comparisons. We have to take that with a grain of salt. It’s not a head-to-head randomized clinical trial. But I think definitely the effort being made in these real-world cohorts in that sense and in the sense of trying to match treatment periods, because we know that axi-cel was available at least in the EU and in Europe before liso-cel, so there are more patients treated with axi-cel at an earlier time point. So trying to match treatment eras so that you’re not skewing the way you’re selecting and managing these patients is another important factor. So I think these statistical methodologies we mentioned, PSM, IPTW, and trying to match them in terms of treatment eras will help us be able to compare more correctly or adequately the outcomes in terms of both efficacy and toxicity of axi-cel and liso-cel patients in the relapsed/refractory large B-cell lymphoma setting.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.