CAR-T Toxicity Management and Steroid Use in High-Grade B-Cell Lymphoma: Impact on Real-World Survival Outcomes in the UK

Background

Axicabtagene ciloleucel (axi-cel) and tisagenlecleucel (tisa-cel) CD19 CAR-T therapies are licensed in the UK for relapsed/refractory large B-cell lymphoma (LBCL). Corticosteroids for CAR-T toxicity were administered to 28% and 10% of patients in ZUMA-1 and JULIET respectively, but real-world steroid use is reported to be much higher (Nastoupil et al, JCO 2020), with concerns that this may adversely impact on OS and PFS following CAR-T.

Analysis of real-world datasets may facilitate the identification of modifiable risk factors for toxicity. Here, we report the UK experience of CAR-T toxicity and its management in 341 LBCL patients with a focus on predictors for corticosteroid use and the impact of steroids on OS/PFS post CAR-T therapy.

Methods

Data were collected in 10 UK centres from January 2019 to April 2021. CRS/ICANS were graded prospectively (ASTCT). Low (LG) and high-grade (HG) CRS/ICANS were classified as grade 1-2 and 3-5 respectively. Hyper-acute CRS was defined as onset within 24 hours of cells. Steroid cumulative dosing for CAR-T toxicity was calculated as dexamethasone equivalent up to day 30. Product selection was at the discretion of the treating physician. Toxicity management in the UK follows EBMT and ASCT guidance.

Results

A total of 341 patients received axi-cel (n=261) and tisa-cel (n=80) with a median follow-up of 14.8 (3.4-30) and 13.9 (range 7.5-27) months. Median age was 59 (range 18-80), 61% were male (M=207, F=134) and 79% received bridging therapy (BT).

HG CRS, hyperacute CRS and HG ICANS with axi-cel (8.8%; 38%; 21% respectively) and with tisa-cel (7.5%; 26%; 5% respectively) were observed, with incidence of HG ICANS post axi-cel lower than published reports.

Risk factors for HG CRS were age<65yrs (p=0.014), LDH pre-lymphodepletion (LD) (p=0.04) and ECOG>0 at infusion (0.001); and hyper-acute CRS was associated with stable (SD) or progressive disease (PD) after BT (p=0.036). On multivariate analysis (MVA), ECOG >0 (OR 3.4, 95% CI 1.3-9.1, p=0.015) and SD/PD post BT (OR 1.8, 95% CI 1.03-3.3, p=0.039) were predictive of HG CRS and hyper-acute CRS respectively.

Risk factors for HG ICANS were age<65yrs (0.03), ECOG>0 (p=0.009), SD/PD post BT (p=0.011), extranodal disease (p=0.02) and LDH>normal on Day 0 (p=0.02). By MVA, LDH>normal on Day 0 (OR 3.8, 95% CI 1.4-10.5, p=0.009) was predictive of HG ICANS.

Steroids were used in 44% (n=115) of axi-cel treated patients, administered at day 6 (median; range 1-22), at a median dose of 165mg (range 10-2182) over a median of 8 days (range 1-86). Anakinra was given as an adjunctive agent for high-grade ICANS (n=26, 10%), CRS (n=10, 4%) or HLH (n=3, 1%) at a median of day 8 (range 5-43) for a median of 6 days (range 1-16). 18% of tisa-cel patients (n=14) received steroids, administered at day 4 (median; 2-26) at a median dose of 95mg (range 10-220) over a median of 4 days (range 1-8). Tocilizumab was used in 74% and 48% of axi-cel and tisa-cel patients respectively.

Clinical factors associated with steroid use included ECOG>0 (OR 1.8, 95% CI 1.2-5.6, p=0.002), and the emergence of SD/PD post BT (OR 1.7, 95% CI 1.02-2.8, p=0.04). SD/PD post BT also predicted for higher total steroid dose (>median; OR 1.7, 95% CI 1.1-2.9, p=0.03), and prolonged (>1 week) administration (OR 1.8, 95% CI 1.1-2.9, p=0.03) (Table 1).

The impact of clinical parameters including toxicity and steroids on PFS and OS post-CAR-T were assessed. Adjusted for age, sex, ECOG and LDH>normal pre-LD, HG ICANS (but not HG CRS) was associated with worse OS (HR 2.4, 95% CI 1.3-4.4, p=0.004). 100 day NRM was 3.5% for the whole cohort. Median OS was not reached.

In contrast to other published series, early steroid use (HR 1.0, 95% CI 0.6-1.6, p=0.99), higher total doses (HR 1.2, 95% CI 0.7-1.9 p=0.6) and prolonged (>1 week) steroid exposure (HR 1.3, 95% CI 0.8-2.2, p=0.25) were not associated with worse OS. Early steroid use (HR 0.7, 95% CI 0.3-1.6, p=0.4), higher doses (HR 0.5, 95% CI 0.2-1.2, p=0.13) and prolonged exposure (HR 0.9, 95% CI 0.4-2.0, p=0.8) also had no significant impact on PFS.

Conclusion

Patients who are refractory to BT or with other markers of high disease burden at infusion are more likely to require steroids for CAR-T toxicity, with axi-cel associated with more HG ICANS and cumulative steroid exposure than tisa-cel. Despite this, steroid use for CAR-T toxicity did not negatively impact on PFS or OS. Rates of HG ICANS with axi-cel were comparatively lower in this UK dataset.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal