Risk factors for CRS and NT with bispecific antibodies: Data from the American bispecifics and CART (ABC) consortium Free

Background: Bispecific antibodies (BsAbs) have changed the treatment paradigm in aggressive B-cell lymphomas, but these agents carry risks of cytokine release syndrome (CRS) and neurotoxicity (NT). These risks are most pronounced during the first cycle, requiring careful monitoring and hospitalization which may limit community use. Our aim was to explore risk factors for CRS and NT during the first cycle of BsAb administration.

Methods: Fourteen institutions contributed to a shared database of patients (pts) with aggressive B-cell lymphoma receiving BsAbs. Data around demographics, histology, treatment history, laboratory values at administration, and incidence of CRS and NT following Cycle 1 were collected. A chi-square test of independence was used to assess differences in CRS and NT rates across glofitamab (glofit), mosunetuzumab (mosun), and epcoritamab (epcor). Continuous laboratory measures were binned into abnormally high, normal range, and abnormally low based on reference ranges from the American Board of Internal Medicine. A series of univariate logistic regressions was performed between variables of interest and the following: 1) any grade CRS, 2) Grade 2+ CRS, 3) Grade 3+ CRS, and 4) any grade NT.

Results: A total of 262 pts received a BsAb. Median age was 67; pts were 59% male, 76% white, and 90% non-Hispanic. Prevalent histologies were de novo diffuse large B-cell lymphoma (73%), transformed follicular lymphoma (12%), and transformed marginal zone lymphoma (5%). BsAb products were glofit (45%), mosun (29%), epcor (23%), and other (3%). Products were administered as monotherapy or with steroids alone (66%), with systemic chemotherapy (7%), with targeted therapy (21%), or with XRT (5%). Line of therapy was known for 242 pts: 14% received 1^st line, 45% received 2^nd or 3^rd, and 41% received 4^thor later. Of the 244 pts with available CRS data, 79 (32%) experienced any grade CRS; 19 (8%) had G2 and 12 (5%) had G3+. Of the 231 pts with available NT data, 20 (9%) had any grade NT and 4 (2%) had G3+. CRS rates by BsAb product were 33%, 25%, and 34% for glofit, mosun, and epcor respectively (p =.43), while NT rates were 10%, 3%, and 11% respectively (p=.19). All grade CRS was more commonly seen in female pts (OR=1.79, 95% CI:1.04-3.09, p=.035), pts with high LDH (>225 IU/L, OR: 2.14, CI:1.14-4.0, p=.017), and pts receiving concurrent systemic chemotherapy (OR=5.18, CI:1.86-14.5, p=.002). Odds of any grade CRS did not vary by: (1) presence of extranodal or bulky (>5cm) disease, (2) bone marrow involvement, (3) CNS disease, (4) concurrent receipt of targeted therapy or XRT, or (5) abnormalities in other baseline lab values (ALC, ANC, platelets, CD4 count, albumin). G2+ CRS was increased in pts with high ferritin (>336 ng/mL, OR=2.90, CI:1.01-8.33, p=.048). All patients who developed G3+ CRS with sCRP data available (n=6, 100%) exhibited very high sCRP (≥15mg/L). Ferritin was also available for 5 of these pts; all of whom exhibited high ferritin (>336). Conversely only 12% pts without G3+ CRS exceeded both those thresholds. Of note, sCRP was only available for 57% of pts and ferritin for 61%. G3+ CRS was more common in pts receiving concurrent systemic chemotherapy (OR=10.8, CI:2.78-41.8, p<.001). Any grade NT was increased in pts with high ferritin (OR=6.45, CI:1.32-31.6, p=.022), high ANC (>8.25 /µL, OR=4.32, CI:1.26-14.8, p=.02), and high LDH (OR = 4.8, CI:1.08-21.3, p=.039), and increased but not significantly with high sCRP (>3mg/L, OR=3.8, CI:0.81-17.9, p=.09).

Conclusion: This study represents one of the largest multi-institution analyses of risk factors for CRS or NT in patients treated with BsAbs for B-cell lymphoma. It serves as confirmation of low incidence of G3+ CRS and any grade NT, mirroring results from clinical trials. We identified sCRP ≥15 mg/L and ferritin >336 ng/mL as possible thresholds for screening for patients requiring closer observation during drug ramp up, but these values require increased acquisition and validation. Notably, we did not reproduce previously reported associations with extranodal disease or high tumor burden and found increased risk of CRS with concurrent systemic chemotherapy, and increased risk of any CRS and NT among patients with LDH >225 IU/L. These findings require further investigation. This evidence supports the growing initiative to understand which patients are suitable for an entirely outpatient ramp-up of BsAbs and their expanded use in the community setting.