Safety of abbreviated dose-escalation in patients (pts) receiving glofitamab (glofit) or epcoritamab (epcor)

Abstract

Abstract Introduction Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) are unique and potentially serious adverse effects associated with bispecific antibodies (BsAbs) including epco and glofit. Measures to mitigate this risk, including step-up dosing, and a pre-treatment dose of obinutuzumab (obin, glofit only) delay administration of therapeutic doses of BsAb, which is vital in pts with rapidly proliferative disease. There is limited data on the safety and efficacy of accelerated step-up dosing or the omission of obin. To bridge this knowledge gap, we reviewed our experience in pts who received glofit or epco with rapid dose-escalation schedule (RDES). Methods We performed a multicenter retrospective study of real-world pts with non-Hodgkin lymphoma treated at nine centers in the Collaborative US Bispecific Consortium (CUBIC). RDES was defined as omission of obin (for glofit) and/or a dose-escalation schedule shorter than that established per product package insert (PI). Descriptive statistics were applied to assess pt demographics, treatment characteristics, response rates and CRS/ICANS rates. Results Out of 298 pts treated with BsAbs in the CUBIC, 30 (10%) received glofit (n=18) or epco (n=12) on a RDES. Median age was 67 years (range 34-87), 27 pts (77%) had an ECOG PS of 0-2 and 7 (23%) >2. Histologic subtypes included DLBCL (n=16, 53%), HGBCL (n=11, 37%), MCL (n=2, 6%) and CLL (n=1, 3%). Most pts had stage III-IV disease (90%), non-bulky disease (<7.5 cm, 67%) and elevated LDH (93%). Among LBCL pts, most had IPI of 3-5 (74%). Pts had a median 2 prior lines of therapy (LOT), 60% had primary refractory disease, 53% had prior CART, and 4 received gemcitabine/oxaliplatin with BsAb. Among all pts treated with RDES, CRS occurred in 16 (53%) pts (grade (G)1 n=9, G2 n=6, G3 n=2) with a median duration of 2 days (d) (1-5). CRS was treated with steroids (n=8) and tocilizumab (n=11). ICANS occurred in 7 pts (27%) (G1 n=5, G2 n=1, G3 n=1, G4 n=1) with a median duration of 2d (1-5). ICANS was treated with steroids in all 7 pts. After a median follow up of 5.1 months from start of BsAb (0-11.7), pts with measurable response had a best ORR of 54% (n=14), including 15% (n=4) best complete response (CR) and 38% (n=10) partial response (PR). The median time to best response was 34d (12-302). Of epco RDES pts, the median time to first full dose (48mg) was 6 (5-15) d, and 7 pts (58%) reached full dose in ≤7d. One pt reached full dose in 15d, with a 3d interval between the 2nd step up dose (SUD, 0.8mg) and full dose. Two (17%) pts did not receive full dose epco due to pt preference and infection, respectively, and did not receive subsequent LOT. Five (42%) pts experienced CRS - 4 (80%) cases occurred at the 1st full dose and 1 (20%) at the 2nd SUD. ICANS occurred in 2 pts (25%). Among all non-RDES epco pts in CUBIC, CRS and ICANs rates were 33% (n=41; ≥ G3 4%) and 12% (n=15; > G3 1%), respectively. In pts reaching full dose in ≤7d, 1 (14%) experienced G2 CRS and 1 (14%) experienced G1 ICANS. ORR for epco pts with measurable response (n=9) was 44% (n=4), including 22% (n=2) CR. Of glofit RDES pts, the median time to first full dose (30mg) was 14d (5-19) (vs. 21d per PI). Two pts (11%) reached full dose in ≤7d. One pt (6%) did not receive full dose glofit due to infection and death. Obin was omitted in 11 (61%) pts. Eleven (61%) pts experienced CRS - 9 (82%) cases occurred at the 1st SUD (2.5 mg). These episodes were predominantly G 1 or 2 (91%). ICANS occurred in 5 (28%) pts at the 1st SUD with the majority G1 (n=3, ≥G3 n=2). Among all non-RDES glofit pts in CUBIC, CRS and ICANS rates were 46% (n=58; ≥G3 2%) and 16% (n=19; ≥G3 2%), respectively. In pts reaching full dose in ≤7d (n=2), 1 (50%) had CRS and 0 experienced ICANS. ORR for glofit pts with measurable response (n=17) was 59% (n=10), including 12% (n=2) CR and 47% (n=8) PR. Conclusion We report a real-world analysis of pts treated with glofit and epco using RDES, including the largest reported cohort treated with glofit without obin. While numbers are limited, CRS, ICANS and response rates are comparable to those seen in pivotal trials but appear to be higher than those seen in non-RDES pts in the CUBIC. This may be due to the high-risk features of the RDES study population. Future analysis of a larger pt cohort is needed to validate this finding. Our study supports the utilization of a RDES for glofit or epco when clinically warranted.