Pre-CAR next generation sequencing informs patient outcomes in children and young adults with B-acute lymphoblastic leukemia

Abstract Introduction Lower baseline disease burden associates with improved survival in children and young adults with B-cell acute lymphoblastic leukemia (B-ALL) treated with chimeric antigen receptor T cells (CAR-T). Conversely, low total CD19+antigen load (an aggregate measure of malignant and physiologic B cells) at pre-CAR baseline resulted in decreased CAR-T response durability in early clinical trials. While these data used flow cytometry measures, limited data exists characterizing the role of pre-CAR next-generation sequencing (NGS) on CAR-T outcomes. NGS is an alternative minimal residual disease (MRD) measure that can detect B-ALL at 1 clonal cell/million nucleated cells. Inherent to the assay (clonoSEQ, Adaptive Biotechnologies) is a complementary measure of immunoglobulin heavy chain (IgH)/total nucleated cell (TNC) which represents the total B cell number represented in the repertoire. We conducted a retrospective single-institutional study to evaluate the role of baseline NGS on post-CAR-T outcomes and hypothesized that while lower pre-CAR NGS MRD would result in improved survival, lower IgH/TNC may represent a low aggregate pre-CAR B cell load and result in inferior outcomes. Methods We conducted a retrospective analysis of patients ≤26 years who received CD19, CD22 or CD19/22 CAR at Stanford between August 2017 and August 2024. Univariate and multivariate Cox regression analyses were used to assess the impact of demographic and disease variables on overall survival (OS) and event-free survival (EFS). These variables included age, gender, race, cytogenetic risk, CAR-T indication, type of CAR-T received, and prior therapies, as well as our primary variables of interest: pre-CAR disease burden (NGS; log peak clonal cells), and pre-CAR B cell load (NGS; total immunoglobulin heavy chain (IgH)/total nucleated cell (TNC) ratio), where total IgH/TNC of <0.5% was defined as a low aggregate B cell load. We aimed to assess the role of baseline NGS MRD (Aim 1) and baseline total IgH/TNC (Aim 2) on survival (OS/EFS) in our overall patient cohort. To account for the fact that all high disease burden patients would be expected to have ≥0.5% IgH/TNC, as an exploratory aim, we studied the impact of these measures in a limited cohort of those with low pre-CAR disease burden (<5% bone marrow blasts). Results The dataset included 76 CAR-T infusions across 60 patients, with a median age at infusion of 13 years (range 1-25). The cohort comprised 53% male and 47% female patients, with 57% receiving commercial CD19 CAR-T (tisagenlecleucel), 7% investigational CD22 CAR-T, and 26% investigational CD19/22 bispecific CAR-T. Only the first CAR-T infusion per patient was included in the analyses. Univariate and multivariate analyses of the entire cohort revealed that only pre-CAR log peak leukemic clonal cells significantly associated with OS (Hazard ratio (HR)=1.15, p=0.0038) and EFS (HR=1.15, p<0.0001). When restricting analysis to patients with <5% baseline blasts by flow cytometry (N=36), log peak leukemic cells did not significantly associate with OS or EFS (OS: HR=1.17, p=0.086; EFS: HR=1.09, p=0.10). However, in this low tumor burden cohort, univariate analysis demonstrated that baseline pre-CAR bone marrow total IgH/TNC of <0.5% associated with lower EFS (HR=0.27 for patients with >0.5% total IgH/TNC compared to <0.5%, (p=0.046)). While the sub-stratified sample size of the low-burden cohort did not support multivariate analysis, expanded multisite validation efforts are underway. Conclusion In our overall cohort, increased log peak leukemic MRD clonal cells by NGS at pre-CAR baseline associated with inferior OS and EFS. While the NGS MRD measure of disease burden did not maintain significance in the already low-burden B-ALL cohort, measures of low pre-CAR aggregate B cell load (<0.5% total IgH/TNC) associated with inferior EFS in this cohort. While multi-site validation efforts are underway, this first study of NGS as a measure of pre-CAR B cell load supports that CAR-T cells targeting B cells may require threshold levels of aggregate B-cell load for effective expansion and durable B-ALL surveillance.