Do Blood Immune Biomarkers Predict Response to Mosunetuzumab for FL, MZL?

Inside a clinical trial evaluating the CD3xCD20 bispecific antibody mosunetuzumab for first-line treatment of follicular lymphoma (FL) and marginal zone lymphoma (MZL), investigators conducted an exploratory study to determine if patients’ circulating immune biomarkers were predictive of response to the drug. Their findings were released at the 2025 American Society of Clinical Oncology (ASCO) Annual Meeting, held in Chicago, Illinois.

“Although CD8+ T cells are the main effector cells engaged by mosunetuzumab, our observations in previously untreated FL/MZL suggest that increased naïve CD4+ helper T cells and NK cells PreTx [at baseline], along with lower CTLA-4 levels during Tx [treatment] may better predict CR [complete response],” wrote the investigators, led by Charles J Milrod, MD, of Brown University Health, Providence, Rhode Island. Dr. Milrod also presented the study findings on-site at the Meeting.

The investigators performed this study within BrUOG-401, a phase 2 trial in which patients with untreated, high-burden FL and MZL received 8 cycles of mosunetuzumab. Disease in the patients was evaluated with PET/CT after the fourth cycle of treatment and at the end of the 8-cycle treatment. Those who did not have CR by the fourth cycle of treatment were given additional lenalidomide for cycles 4 through 8, with an option to extend treatment up to 12 cycles.

Peripheral blood samples were collected from the patients at treatment baseline, on day 8 of the first cycle, on day 1 of the second cycle, at the fourth cycle, and the end of treatment. The investigators used a multiplex Luminex assay to measure 25 plasma cytokines associated with T-cell activation and regulation, along with flow cytometry to assess immune cell subsets. They then compared the markers using rank-sum tests and mixed-effects generalized linear models.

Response to mosunetuzumab by the end of the 8-cycle treatment was evaluable in 34 patients. Among them, 65% achieved CR by the fourth cycle and 85% by the end of treatment. Of 22 patients who had baseline cytokine level information available, 15 had achieved CR by the fourth cycle and 19 by the end of treatment. The investigators found no significant association between patient baseline cytokine levels and CR at the fourth cycle or at end of treatment.

Regarding T-cell activation markers, specifically GZMA/B, IFNg, IL2, and IL7, the investigators found by cycle 1 day 8 that these had increased overall (P < .05 for all). However, only lower CTLA-4 levels were significantly predictive of achieving CR by cycle 4 (P = .0031), and persistently lower CTLA-4 levels at cycle 4 were found to correlate with CR at cycle 4 (P = .008). None of the cytokines was significantly associated with achieving CR at the end of treatment.

In the 26 patients in whose samples flow cytometry was performed, the investigators observed an overall increase in circulating NK cells during treatment (P<.05 at all timepoints). They found that having greater NK cell amounts at baseline significantly correlated with achieving CR at cycle 4 (P = .043). Having elevated NK or elevated HLA-DR+ NK cells at cycle 2 was found to correlate with achieving CR at cycle 4 (P=.011 and P=.0042, respectively).

Having elevated CD4+CD45RA+ T cells at baseline or reduced CD4+CD45RO+CD25- T cells at baseline was associated with CR achievement at the end of treatment (P=.0035 and P=.0061, respectively). At cycle 4 the CD8+CD45RO+CCR7-CD27- effector memory T cell subset was found to be elevated overall in patients (P=.025), but at no timepoint was CD8+ subset predictive of CR.

The investigators noted that the P values calculated were not corrected for multiplicity.

“These findings suggest that cytokine-driven immune priming could influence response to BsAbs [bispecific antibodies], providing a basis for future investigations of combinations therapies to enhance their efficacy,” Milrod and colleagues concluded.