The Impact of Treatment-Free Intervals on T-Cell Exhaustion and Clonotypic Diversity in Myeloma Patients Treated with BCMA Bispecific Antibodies

Background: BCMAxCD3 bispecific antibodies (BsAb) have changed the treatment landscape in relapsed refractory multiple myeloma (RRMM). However, despite high response rates, the majority of patients continue to relapse. Infectious complications, mainly viral, represent another challenge with BsAb, seen in up to 90% of patients with half of them requiring hospitalization. Both issues, loss of response as well as infectious complications, can be linked to impaired T-cell function potentially emerging during BsAb therapy. Most BsAbs have been developed as weekly treatments and one hypothesis is that dose dense treatment could overstimulate T cells finally leading to exhaustion and dysfunctionality. Thus, we compared T cell profiles and their functionality from patients with weekly dosing to patients who switched to a monthly schedule due to toxicity or patient request.

Methods: We longitudinally biobanked peripheral blood samples of 19 patients of whom 7 switched from a weekly (q1w) to a monthly (q4w) BsAb treatment schedule. Peripheral blood mononuclear cells (PBMCs) were analysed using 16-color flow cytometry, including T-cell subset phenotyping and the exhaustion marker PD1, TIM3, SLAMF6 and KLRG1. Furthermore, CITEseq, scTCRseq, and Seahorse metabolic studies were performed on selected patients.

Results: The enrolled patients were heavily pretreated with a median number of 6 prior therapy lines. All patients responded to BCMA BsAb with VGPR or better and were on treatment for >3 month. Only one patient from the q4w cohort relapsed during the study period. High-risk cytogenetics, lines of previous therapy, and penta-refractory disease status were equally distributed between patients on weekly and monthly schedule. Infectious complications occurred in 10/19 (q1w) (52.6%) and 1/7 patients (q4w) (14.2%).

First, we addressed the question if a monthly treatment schedule confers less T-cell exhaustion by comparing PD-1 checkpoint expression and T-cell subsets between q1w and q4w administration. Indeed, flow cytometry showed a significant decrease in the CD4+PD1+ (36.9 vs. 21.1% p<0.05, unpaired t-test) and in the CD8+PD1+ (29.44 vs. 16.54% p<0.05) T cells. We also observed a decrease in the frequency of the effector memory subsets (CD45RA-, CD62L-, TEM) of CD4+ (66.19 vs. 49.63% p<0.05) and CD8+ (53.7 vs. 40.32% p=0.09), further supporting a concept in which a treatment free interval leads to recovery of the T-cell system.

We also had access to paired samples of five patients who switched from q1w (1 ^st sample) to q4w dosing (2 ^nd sample). Here, flow cytometry showed a decrease of exhausted CD4+PD1+ T cells (33.68 vs. 21.76% p=0.046, paired t-test) and CD8+PD1+ T cells (29.2% vs. 15.1% p=0.117), confirming our cross-sectional analysis. Likewise, expression of SLAMF6, a regulator of T cell exhaustion, also decreased from weekly to monthly schedule (SLAMF6/CD4+:20.7 vs. 15.2% p=0.06; and CD8+:17.8 vs. 11.3% p=0.1), but differences were not statistically significant probably due to sample size. Our analysis on T-cell metabolism did not reveal significant differences between weekly and monthly dosing schedules.

After strict quality control, three of the paired samples were further analysed using CITE-seq and scTCRseq. In line with our flow results, patients showed a significant decrease in CD8+ pre-exhausted T-cell subset (CD8+TCF-1+CD27+) (3.55 vs. 1.54% p=0.02) and CD8+ terminal differentiated effector memory T-cells (CD8+ TEMRA) (4.66 vs. 3.57% p=0.04) from weekly to monthly schedules, again suggesting recovery of the T-cells after a longer treatment-free interval. We also observed a trend towards decrease in the CD4+ effector memory subset (CD4+TEM, 6.96 vs. 4.89% p=0.07). Interestingly, the number of hyper-expanded T cell clonotypes increased from weekly to monthly dosing in paired samples. This was somewhat unexpected but expansion could be interpreted as reactivation of a few previously exhausted T cells.

Conclusions: The optimal dosing schedule of bispecific antibodies has yet to be determined. Our results support a more extended dose schedule to allow T-cell-recovery between BsAb administrations. Whether such a strategy will result in higher antitumor efficacy or reduced susceptibility to infections, has to be evaluated in future studies.