CAR19 expansion is dependent on lymphoma histology. (A) MCL histology has increased D14, D21, and D28 CAR19 expansion relative to LBCL based on a multivariate linear mixed effects model of longitudinal data. (B) CAR19 AUC in MCL is greater than LBCL with almost fourfold higher CAR19 exposure (Wilcoxon test). (C-D) Breakdown of the LBCL histology indicates there are differences between HGBCL, TFL, and other LBCL pathologies in CAR19 expansion. These differences are not significant in mixed effects modeling (C) but there is a significant AUC difference between HGBCL and other lymphoma histologic subtypes when directly compared (D; Wilcoxon test). (E) Greater LDH is associated with significantly more persistence of CAR19s on D21 and D28 using a multivariate mixed effects model. Plot is representative of statistical output that uses continuous values. (F) Discrete analysis of CAR19 expansion (Wilcoxon test) by AUC based off prelymphodepletion LDH greater than the upper limit of normal does not demonstrate statistically significant total exposure differences in CAR19 expansion. (G) Greater age is associated with reduced CAR19 persistence on D21 and D28 using the same modeling. Plot is representative of statistical results that uses continuous values. (H) Discrete modeling of total CAR19 exposure over the first 28 days does not demonstrate significant differences in AUC using a standard age threshold (Wilcoxon test). However, highly significant effect sizes in mixed effects modeling for both LDH and age were modest. ∗P < .05; ∗∗P < .01, ∗∗∗P < 0.001; ns; error bars are standard error of the mean; plots are representative of modeling results. Significance values in line charts represent the interaction between CAR19 expansion and day; graphs are only representative of the continuous statistics. ns, not significant.