Blood test predicts CAR-T response in non-Hodgkin lymphoma
Click Here to Manage Email Alerts
A risk model based on low-pass whole-genome sequencing of cell-free DNA predicted which patients are more likely to respond to chimeric antigen receptor T-cell therapy for large B-cell lymphoma, study results showed.
Researchers from The University of Texas MD Anderson Cancer Center developed a noninvasive blood-based assay to determine a patient’s level of genomic instability. Investigators combined this with other traditional markers of tumor bulk to develop a model that can reliably risk stratify patients before they receive CD19-directed CAR T cells.
“Incredible advances like CAR T-cell therapy for patients with lymphomas are improving our ability to potentially cure more patients, but we desperately need better tools to predict which therapy may benefit which patient,” Jason Westin, MD, MS, director of lymphoma clinical research and section chief for aggressive lymphoma at MD Anderson, told Healio. “Our model suggests that current CAR T-cell therapies work very well in certain patients, and that others should be considered for clinical trials to further improve outcomes.”
Background and methodology
A high proportion of heavily pretreated patients achieve durable responses to treatment with one of the three commercially available CAR-T cell therapies. However, about half will experience disease progression within 3 months, Westin said.
“We struggle to predict outcomes for our patients — specifically which patients are most likely to benefit from a treatment like CAR T-cell therapy,” he said. “Having better predictors may allow us to personalize therapy to improve outcomes or select other options for patients.”
Westin and colleagues aimed to develop a novel approach to risk stratification prior to CAR T-cell therapy by using low-pass whole-genome sequencing of cell-free DNA to define DNA copy number alterations. The researchers developed the high focal copy number alterations score (FCS) to denote genomic instability.
The investigators conducted an analysis of pretreatment samples from a cohort of 122 patients (median age, 60.5 years; range, 18-88; 71% men) with relapsed or refractory large B-cell lymphoma undergoing leukapheresis prior to manufacturing of standard-of-care CD19-directed CAR T cells.
The study included patients treated at MD Anderson Cancer Center with CD19-directed CAR-T between May 2018 and Feb. 2021.
Median follow-up was 20.7 months (range, 0.5-36).
Key findings
Multivariate analysis showed a significant association between high FCS and inferior 3-month complete response rate (28% vs. 56%; P = .0029).
Likewise, patients with high FCS achieved significantly lower 3-month OS (33% vs. 59%; HR = 2.1) and shorter median PFS (3.5 months vs. 12.7 months; HR = 2.11) than those who had low FCS.
Researchers also identified 34 unique focal copy number alterations in 108 (89%) patients. Among these, they found deletion 10q23.3 leading to loss of FAS death receptor to be most highly associated with poor outcomes, including inferior PFS (HR = 3.49) and OS (HR = 2.68).
Clinical implications
The results have allowed researchers to combine data from the assay along with other clinical factors to develop a classification system based on low, medium and high risk for lack of response to CAR T-cell therapy, Westin said.
“Studies like ours could help us transform our approach from a one-size-fits-all model into one where we personalize treatments,” he told Healio. “Patients with a high risk for lymphoma progression may need additional therapies either added to or in replacement of a CAR T-cell therapy.”
Efforts are ongoing to validate results from this study and in the context of other disease states, including for patients with newly diagnosed lymphoma and for the assessment of other targeted therapies, Westin added.
For more information :
Jason Westin, MD, MS, can be reached at The University of Texas MD Anderson Cancer Center, Department of Lymphoma and Myeloma, 1515 Holcombe Blvd., Unit 429, Houston, TX 77030; email: jwestin@mdanderson.org.
Perspective
Back to Top
Craig Sauter, MD
Investigators at The University of Texas MD Anderson Cancer Center reported prognostication of patients with large B-cell lymphoma proceeding to commercially available chimeric antigen receptor modified T cells directed against CD19 based upon a relatively limited resource intensive whole-genome sequencing platform of cell-free DNA from leukapheresis plasma samples at time of mononuclear cell collection for CART19 generation.
This low-pass whole-genome sequencing technology (lpWGS) essentially identifies and quantitates copy number alterations and the investigators were able to generate a focal computed score per patient. The focal computed score was then inserted as a variable in a multivariable analysis for outcome endpoints, including complete response, PFS and OS to CD19-directed CAR-T based upon a binary high or low value (evenly splitting the cohort based upon the median focal computed score.
Of the 122 patient cohort analyzed, the majority (92%) received axicabtagene ciloleucel and nearly three-quarters had received more than two prior lines of therapy. In multivariable analysis, researchers observed superior complete response, PFS and OS among patients with low focal computed score vs. high focal computed score.
A three-variable risk-model based upon: focal computed score, lactate dehydrogenase pre-CAR-T and more than one extranodal site reliably risk-stratified toward outcomes with clear PFS and OS separation of low (score 0), intermediate (score 1-2) and high (score 3) values.
Clearly, this report identifies copy number alterations by lpWGS as a valuable prognostic tool for patients with large B-cell lymphoma proceeding to CD19-directed CAR T-cell therapy. Whether this tool simply represents a surrogate for tumor bulk, which clearly defines negative prognosis per multiple prior publications, is difficult to discern in this manuscript. In the iteration of the current report, the limitation of this tool is restriction to prognosis alone.
Beyond prognosis, the lpWGS platform could potentially serve as a valuable determiner of risk-stratification in future prospective interventional studies for LBCL patients intended to proceed to CD19-directed CAR-T vs. other therapies, or in addition to adjunctive interventions.
Whether lpWGS provides prognostic value beyond conventional measures of tumor bulk by standard PET would need to be confirmed in rigorously conducted prospective investigation. Regardless, this report is a welcomed contribution to the knowledge base of the burgeoning clinical arena of CD19-directed CAR-T for large B-cell lymphoma.
Collapse
Cherng HJ, et al. Blood. 2022;doi:10.1182/blood.2022015601.
Click Here to Manage Email Alerts