Introduction: CD19 directed CAR T cell therapies, such as axicabtagene ciloleucel (axi-cel) and lisocabtagene maraleucel (liso-cel), are standard treatments for relapsed/refractory large B-cell lymphomas (LBCL); despite high initial response rates, only 30–40% of patients achieve durable remissions. Multiple mechanisms of CAR T resistance have been identified including high tumor burden, hostile tumor microenvironment and downregulation of target antigen. We previously demonstrated CD19 downregulation in ~30% of pts treated with axi-cel and highlighted the utility of quantitative flow cytometry for assessing CD19 antigen density (Spiegel, Nat Med, 2021). Here, we present findings from a large cohort of pts treated with CD19 CAR T, incorporating quantitative flow cytometry-based profiling of CD19, CD20, and CD22 expression, including analysis of paired pre- and post-treatment biopsies.

Methods: Pts with r/r LBCL treated with commercial CD19-directed CAR T at Stanford University between January 2018 and December 2024 who enrolled on an IRB approved biorepository protocol and had fine needle aspirate (FNA) biopsies with available quantitative flow cytometry data were included. A 10-color flow cytometry panel and calibrated fluorochrome-conjugated quantitation beads (BD Quantibrite) were run concurrently to quantitate CD19, CD20, and CD22 expression. Median antigen density values, reported as molecules/cell (m/c), were used for analysis. Comparison between groups (pre-CAR T cohort vs progression cohort) was performed by Mann-Whitney. Univariable and multi-variable logistic regression was used to associate pre-CD19 levels with clinical outcomes.

Results: We identified 116 unique pts (107 axi-cel, 9 liso-cel) with 134 total unpaired biopsy samples (n = 17 paired), pre-CAR T (n=75) and at progression (n=59). Stage III/IV disease was seen in 72.4%, bulky disease in 15.5% and 62.1% were male. Prior lines of therapy were 1 in 23%, 2 in 39.2% and 37.8% of pts received ≥3.

To evaluate changes in CD19 antigen levels after CAR T, we compared unpaired biopsies at progression (n=59) to pre-CAR T (n=75) levels. CD19 antigen density was significantly reduced at progression (2541 m/c, IQR 486–5340), relative to pre-treatment (5493 m/c, IQR 2011–8825) (p=2.8x10-4). CD22 was reduced at progression (3293 m/c, IQR 1529-7942) vs pre-treatment (6019 m/c, IQR 2230–9333)(p = 0.062) while CD20 was comparable (11239 m/c, IQR 1303-30398 vs 16286 m/c, IQR 2517–56156 m/c, p = 0.16). In paired biopsies, CD19, CD22 and CD20 were not lower at the time of progression, potentially due to lower pre-treatment levels (CD19 3650 m/c, IQR 1811–8039; CD22 2537 m/c, IQR 1853–3613, CD20 3699, IQR 1951–20167, p=0.28). At progression, 8 (14%) pts had biopsies with pronounced decreases in CD19, CD20 and CD22 below 3000 m/c and 2 (3.3%) pts had densities below 1000 m/c across antigens, suggestive of antigen loss.

We next tested association of pre-CAR T CD19 antigen density with clinical outcomes in a cohort of 74 pts. Best response to CAR T was 74.3% complete response, 9.5% partial response, 4.1% stable disease and 12.2% progressive disease. After 34.5 months median follow-up, median overall survival was not reached, median progression-free survival (PFS) was 23.1 months (6.19 – NA), and 7 pts died of non-relapse mortality. By univariate logistic regression, pre-CAR T CD19 antigen density associated with PFS (Odd Ratio (OR) 0.36, 95% CI 0.11–1.09, p = 0.08) and progression (OR, 0.34, 95% CI 0.09–1.03, p = 0.07). Selecting 3000 m/c as a cut point based on prior data (Spiegel, Nat Med, 2021), PFS was significantly better in those with CD19 ≥3000 m/c (median 38.6 months, IQR 15.1 – NE) vs those with CD19 <3000 (2.9 months, IQR 0.97 – NE), p=0.009. In a multi-variable regression model with LDH, sex and prior lines of therapy, CD19 antigen density showed a potential association with progression (OR 0.35, 95% CI 0.1–1.1, p = 0.084), corresponding to a 65% decrease in progression for every log increase of CD19 antigen density.

Conclusion: In this large cohort of pts, quantitative flow cytometry demonstrated overall CD19 antigen density was significantly lower at time of progression. Lower pre-treatment CD19 antigen density was also potentially associated with risk of progression in a multivariable regression model. These findings suggest tumor antigen expression levels, when profiled with high resolution methods, merits further exploration as a biomarker for CAR T outcomes.

This content is only available as a PDF.
2025
Sign in via your Institution