Exploration of Language As a Barrier to the Assessment and Management of CAR T-Cell Therapy Associated Toxicities

Introduction: Chimeric antigen receptor T-cell therapy (CAR T) is a novel therapy associated with a unique toxicity profile, namely cytokine release syndrome (CRS) occurring in (37–93%) of patients (pts) and immune effector cell associated neurotoxicity (ICANS) occurring in 23-67% of pts [Zahid, 2020]. Grading and management of these toxicities varies amongst institutions and patient populations. Furthermore, assessment of ICANS utilizes the Immune Effector Cell-Associated Encephalopathy (ICE) scoring system which includes 5-components that are dependent on a patient’s English- speaking proficiency [Lee 2019]. We previously reported a higher-grade of neurotoxicity assessed in non-English speaking pts in a cohort of 127 pts at our institution [Melody, 2023]. We sought to confirm our previous findings by analyzing the impact of language spoken on the incidence, severity, and management including resource utilization for CAR T associated toxicities, in a larger multi-institutional cohort.

Methods: We compiled retrospective data from 13 academic institutions for pts with aggressive B-cell lymphoma treated with CD-19 directed CAR T between 2015 through 2022. Baseline characteristics, lymphoma treatment, along with incidence, severity and management patterns of CRS and ICANS were compared between non-native English-speaking (NNES) and native-English speaking (ES) pts with Pearson Chi-squared test. OS was estimated via Kaplan-Meier method.

Results: In a cohort of 582 pts treated with CAR T, 555 pts were ES and 27 were NNES pts; 17= Spanish, 3= Cantonese, 7= other (including Arabic, Burmese, Polish, Igbo, Japanese, Korean, & Mandarin). Demographic and clinical data for NNES and ES cohorts are included in Table 1. There was no difference in baseline disease characteristics between the two pt populations with the exception of younger pts and a higher proportion of women in the NNES group (p =0.02). Translator services were utilized for 20 (74%) of NNES pts at any point during CAR T-cell treatment.

Two-thirds (66.7%) of NNES pts developed CRS (4=grade 0, 6= grade 1, 9= grade 2, 1= grade 3, 2 =grade 4, 5=Missing data) vs 55.5% of ES pts developed CRS (63=grade 0, 131= grade 1, 134= grade 2, 56= grade 3, 1 =grade 4, 1=grade 5, missing= 169), p < 0.001. For CRS management, there was no difference in the use of tocilizumab (48.1% vs 46.0%, p= 0.85) or escalation to ICU care (13.3% vs 10.1%, p= 0.66) for NNES vs ES groups, respectively.

Thirty-seven percent of NNES pts developed ICANS (6=grade 0, 4= grade 1, 2= grade 2, 3= grade 3, 1=grade 4, 11=missing) compared with 39.8% of ES pts who developed ICANS (120=grade 0, 75= grade 1, 60= grade 2, 75= grade 3, 8= grade 4, 3= grade 5, missing= 241), p= 0.92. ICE scoring for ICANS assessment was recorded in 48% and 51% of NNES and ES pts, respectively. For management of ICANS, there was no difference in the use of steroids (51.9% vs 43.4%, p= 0.43), lamotrigine (61.5% vs 46.0%, p=0.39), or anakinra (10% vs 11.7%, p= 1.0) between NNES vs ES pts. There was no difference in escalation to ICU care for management of ICANS between groups (15.4% vs 19.1%, p=1.0). Neurology consults were obtained more frequently for NNES vs ES but this difference was not statistically significant (61.5% vs 40.5% respectively, p=0.16). Similarly, neuroimaging was obtained more frequently in NNES vs ES pts though not statistically significant (69.2% vs 53.4% respectively, p= 0.39).

There was no difference between re-admission rates within 90 days post CAR T for NNES and ES pts (34.6% vs 26%, p = 0.36). With median follow-up of 35.3 mo (1.4-82.9), there was also no statistical difference in median overall survival of 17.8 mo in NNES pts compared to 28.9 mo in ES pts (p= 0.50; Figure 1).

Conclusions:

Despite our previous observations of increased severity of CRS and ICANS in NNES pts at our own institution, expansion to a larger multi-institutional cohort did not identify differences in occurrence, assessment, or management of toxicities between NNES and ES pts. Analyses were limited by small sample size of NNES patients and missing data for our patient cohort as a whole. This may be explained by decreased access to tertiary care for CAR T in NNES pts and highlights the lack of standardization in assessment and management of CRS/ICANS more globally. None-the-less, the assessment of language as a barrier to care warrants further exploration as does validation of current ICANS assessment tools in NNES pts.