ASH 2025 | Real-world insights on redefining post-CD19 CAR T-cell surveillance with ctDNA

This was something that I worked on during my fellowship at Stanford, and in this study, on behalf of my co-authors, I will be presenting our data on ctDNA surveillance, post-CAR T cell and lysosomal therapy in a real-world setting. So MRD is emerging as a powerful biomarker in large B cell lymphoma, with many frontline studies showing its role either as a complement or even superior to PET-CT in response assessment and disease risk prediction. So we hypothesized to see ctDNA’s role in the post-CAR T-cell therapy, and to, especially in a field where there’s more relapses, whether this could be a much better risk predictor. So in this prospective clinical trial, MRD was tested using the clonoSEQ assay, and we measured at predefined time points. We started at pre-CAR T cell therapy at the pre-LD time point, and the post-CAR T cell therapy we measured at day 14, day 28, day 90, and day 180. And we enrolled about 124 axicabtagene ciloleucel patients and 32 lisocabtagene maraleucel patients, reflective of our center’s real-world practices. Some of the key findings are that the pre-LD ctDNA burden or quantity matters. Higher the pre-LD ctDNA, shorter the progression-free survival, and overall survival. We found that a threshold of more than 1,000 lymphoma genomes per milliliter was predictive of a very short PFS of only about three months in both axicabtagene ciloleucel and lisocabtagene maraleucel cohorts. In axicabtagene ciloleucel, as early as day 14, we found that MRD detectability was predictive of worse PFS and overall survival. In lisocabtagene maraleucel, though limited by the smaller numbers, we did see a pattern only emerge as late as day 90. This could reflect a difference in product biology and expansion kinetics, but emphasizes the role that product biology could have an influence on MRD results. As well as when we looked at the day 28 non-CR PET scan patients, MRD was able to reliably detect who was going to have long-term remission, which is who we’re going to be progressing, thus adding more additive role to our current PET CT surveillance options. In longitudinal assessment, in this cohort with a median follow-up of about 18 months, we had about 64 relapses in total, and only four out of these 64 had a concurrent false negative MRD result. And three out of these four had a sub-centimeter solitary extranodal relapse, and one of them had a 0.8 centimeter cervical lymph node, which was eventually biopsied to be having lymphoma. This highlights this MRD sensitivity of this assay and its role it might have in extranodal relapses. And looking back at these data and having a much larger cohort with axicabtagene ciloleucel, we ran a simulation analysis. We picked up all the day 14 MRD negative patients and longitudinally followed them. And about 95 to 97% of them continued to remain MRD negative and PET negative at longitudinal assessment and eventually achieved durable remission. Only 5 to 6% of patients in this population actually had a subsequent MRD result, and majority of them progressed. So through these results, we propose maybe there is a different surveillance than what we do right now. Maybe there’s an MRD-adaptive surveillance strategy where we could potentially spare PET-CTs in early MRD-negative patients, those who receive axicabtagene ciloleucel therapy, and reserve PET CT more for MRD adaptive strategies only in patients who turn detectable MRD through their treatment course.

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