The Use of the Montreal Cognitive Assessment to Predict Neurotoxicity Post CD19 and BCMA CART

Background: Chimeric antigen receptor T-cell (CART) therapy is approved for non-Hodgkin lymphoma (NHL), multiple myeloma (MM), and acute lymphoblastic leukemia (ALL), and is available in many centers across the US. One of the risks of this treatment is immune effector cell-associated neurotoxicity syndrome (ICANS), which can be potentially fatal. Predicting neurotoxicity with pre-CART evaluations may help risk stratification, determine the need for preemptive strategies, and determine risks with outpatient management. There is sparse literature demonstrating the recommendation of utilizing neurocognitive testing, such as the Montreal Cognitive Assessment (MoCA), to predict the risk of ICANS in CART (Möhn et al., 2022). This retrospective review of CD19 and BCMA CART patients at the University of Kansas Medical Center aims to identify the correlation between baseline pre-infusion MoCA scores and the incidence of ICANS.

Methods: We identified patients who underwent CART infusion for NHL, MM, and ALL from 2017 to 2023. All neurocognitive testing was performed by staff who were MoCA trained. Permission was obtained to utilize the MoCA for retrospective research. The MoCA exam is an 11-question exam scored on a 30-point scale to predict normal cognition (26-30 points), mild cognitive impairment (18-25 points), moderate cognitive impairment (10-17 points), and severe cognitive impairment (<10 points). For this study the MoCA score was categorized into the following: high scores >25 (normal cognition), medium 17-25 (mild cognition impairment), and < 17 low (moderate-severe cognitive impairment). For patients with impaired eyesight or seen via telehealth, a MoCA-BLIND was administered, resulting in a possible 22 points which was converted back to the 30-point scale. Other clinical and laboratory variables that may be associated with the incidence of ICANS were evaluated using univariate logistic regression analysis. Correlation of MoCA score with ICANS incidence, grade of ICANS, and other outcomes of interest such as steroid usage, intensive care unit (ICU) stay, 30-day overall survival (OS), and 100-day OS was reported.

Results: 212 patients who had MoCA scores were included in this analysis, (148 NHL, 56 MM, and 8 ALL) leading to comparisons of 155 CD19-directed CAR T patients and 57 BCMA-directed CAR T patients. 99% of these were baseline scores pre-CART infusion. The median age was 65(range 22-84 yrs.). 99 (47%) developed ICANS, including 19% with Grade ≥ 3 ICANS. Overall, 52% of patients obtained high MoCA scores, 44% obtained medium MoCA scores, and 4.2% obtained low MoCA scores. Patients with high scores had a 91% significantly lower odds of developing ICANS (odds ratio=0.09, p-value=0.023) compared to patients with low scores. Patients with medium scores trended towards a decrease in odds of developing ICANS as compared to patients with low scores (odds ratio=0.12, and p-value=0.052). There was no significant difference in the odds of developing ICANS between patients with high scores and medium scores (p-value=0.223). Additionally, low MoCA scores correlated to the need to use steroids (p= 0.042), need for ICU stay (p=0.038), and 30-day OS (p= 0.035), respectively. However, MoCA scores were not correlated with 100-day OS (p=0.478) or prolonged admissions (p=0.612). Other independent factors that are univariately associated with developing ICANS include: age (p= 0.029), types of diagnosis (NHL, MM, ALL) (p<0.001), lower baseline KPS (p=0.007), higher ferritin at infusion (p<0.001), high CRP at infusion (p <0.001), low WBC at infusion (p=0.028) CD19 CART compared to BCMA (p<0.001), and CD28 compared to 4-1BB (p<0.001)

Conclusion: This is the largest report of MoCA scores and their association with ICANS in CD19 and BCMA CART recipients. Our results suggest that high MoCA scores (normal cognition) are associated with a significantly decreased risk of developing ICANS, while low scores (moderate-severe cognitive impairment) are associated with an increased risk of ICANS. In addition, patients with low scores are more likely to have increased steroid use, need for ICU stay, and lower 30-day OS. MoCA scores provide critical information to determine risk of ICANS and should be considered when determining eligibility for outpatient administration. Other factors may also affect ICANS outcomes, and further study, including multivariate analysis and multicenter reports, will be important.