Targeted interferon therapy with modakafusp alfa for relapsed or refractory multiple myeloma

Multiple myeloma (MM) is an incurable cancer of plasma cells. Despite recent advances,¹ most patients eventually develop disease that is refractory to all available therapies.² Treatments that directly target myeloma cells and induce antitumor immune responses have the potential to greatly improve outcomes.

Interferon α signaling has direct antiproliferative and proapoptotic effects in myeloma cells³ but also induces broad immune activation by reducing the threshold required for innate immune cell activation and providing a stimulatory signal to adaptive immune cells.⁴ Interferon maintenance treatment of MM after autologous stem cell transplantation improved progression-free and overall survival but was limited by toxicity.^5-7 

Modakafusp alfa is an immunocytokine comprising 2 attenuated interferon α-2b moieties fused to a humanized anti-CD38 immunoglobulin G4 (IgG4) monoclonal antibody. Because IgG4 has limited effector functions,⁸ the mode of action of modakafusp alfa is driven by induction of interferon signaling within CD38-expressing (CD38⁺) cells. The interferon α-2b moieties are mutated to attenuate binding affinity for the interferon alfa receptor, allowing highly specific delivery of interferon to CD38⁺ cells.⁹ Furthermore, modakafusp alfa binds to CD38 at a unique epitope and does not compete for binding with current anti-CD38 antibodies.^9-11 In preclinical studies, modakafusp alfa stimulated type 1 interferon signaling, leading to immune cell activation and direct antimyeloma cytotoxicity.^9,12-14 

This first-in-human study evaluated single-agent modakafusp alfa in patients with relapsed or refractory MM (R/R MM).

This ongoing, open-label, phase 1/2 trial consists of a phase 1 dose-escalation portion and a phase 2 dose expansion. Eligible patients had R/R MM defined by International Myeloma Working Group criteria^15,16 with disease progression, had received ≥3 prior lines of therapy; and had disease refractory to, or were intolerant of, ≥1 proteasome inhibitor (PI) and ≥1 immunomodulatory drug (IMiD). Because of concerns that competition for binding to CD38 would obscure important safety signals (despite preclinical studies showing no such competition), in the dose escalation cohorts only, we required patients who had received an anti-CD38 antibody for ≥5 months to have a washout of 90 days. No washout from prior anti-CD38 therapy was required in the expansion cohorts. A full listing of inclusion and exclusion criteria is provided in the supplemental Appendix (available on the Blood website).

Dose escalation followed a 3+3 design to determine either a maximum tolerated dose (MTD) or optimal biological dose (defined as having evidence of antimyeloma activity and pharmacodynamic effect). Modakafusp alfa was administered intravenously over 1 hour (or >2 hours for the highest dose) in 21- or 28-day cycles across doses ranging from 0.001 to 6 mg/kg, using 1 of 4 schedules. The initial schedule of modakafusp alfa (referred to as “modified QW schedule” below) was weekly in cycles 1 and 2, then every 2 weeks (Q2W) in cycles 3 to 6, and then every 4 weeks (Q4W). Because of dose-limiting cytopenias, we explored dosing Q2W, every 3 weeks (Q3W), or Q4W. During dose escalation, modakafusp alfa was administered as a single agent, with dexamethasone given only as a premedication (20 mg with each dose of modakafusp alfa, along with diphenhydramine and acetaminophen). Based on preliminary evidence of safety and efficacy seen during dose escalation, we opened 2 expansion cohorts of single-agent modakafusp alfa at 0.4 mg/kg Q3W and 1.5 mg/kg Q4W, respectively. To understand whether dexamethasone would improve the efficacy of modakafusp alfa or antagonize its immunostimulatory effects leading to decreased responses, we also opened expansion cohorts at the same doses combined with dexamethasone 40 mg weekly (or 20 mg for patients aged >75 years; Figure 1). Preliminary safety results of the dexamethasone cohorts are included in this manuscript, but the Q4W dexamethasone cohort completed accrual after the data cutoff for this report, and we plan to report their efficacy results separately.

Toxicity was evaluated according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 5.0. Dose-limiting toxicity (DLT) was defined as a treatment-emergent adverse event (AE) of the National Cancer Institute Common Terminology Criteria for Adverse Events grade ≥3 that was, in the opinion of the investigator, clearly unrelated to the underlying disease and which occurred before the administration of the cycle 2, day 1 dose of modakafusp alfa, according to the following definitions: (1) hematologic toxicities considered to be DLTs were grade ≥3 hemolysis, grade 4 neutropenia lasting >7 consecutive days, grade 4 thrombocytopenia lasting >14 days, grade 3 thrombocytopenia with clinically significant bleeding, and any other grade ≥4 hematologic toxicity with the exception of grade 4 lymphopenia; (2) an incomplete recovery from treatment-related toxicity causing a >2-week delay in the cycle 2 infusion; and (3) nonhematologic treatment-emergent AEs were considered DLTs with the exception of asymptomatic laboratory changes (other than renal and hepatic laboratory values, and grade 4 lipase/amylase) that could be successfully managed within 72 hours (with improvement of grade 4 events to grade ≤2 and improvement of grade 3 events to grade ≤1 or baseline), grade 3 nausea/vomiting that could be successfully managed with antiemetics (with improvement within 48 hours), grade 3 fatigue lasting <72 hours, grade 3 elevation of alanine aminotransferase or aspartate aminotransferase that resolved to grade ≤1 or baseline within 7 days, and grade 3 infusion-related reactions (IRRs) that responded to symptomatic treatment, without recurrence of grade 3 symptoms.

Full details of doses for each schedule, dose-escalation rules, recommended premedications, and permitted concomitant medications are provided in the supplemental Appendix.

The trial was sponsored and designed by Takeda in conjunction with the academic authors and was conducted in accordance with the principles of the Declaration of Helsinki, the International Conference on Harmonization good clinical practice guideline, and all applicable laws. An independent ethics committee or institutional review board approved the trial protocol. All patients provided written informed consent.

The primary end points for the dose-escalation phase were the frequency and type of DLTs, AEs, and laboratory abnormalities. The primary end point for the dose-expansion phase was overall response rate (ORR), per investigator assessment. Secondary end points included pharmacokinetics, duration of response (DOR), and time to response. Pharmacodynamics were evaluated as an exploratory end point.

Response was evaluated according to International Myeloma Working Group uniform response criteria.^15-17 Minimal residual disease (MRD) was evaluated for patients with a suspected complete response (CR) using the clonoSEQ assay (version 2.0; Adaptive Biotechnologies) at a threshold of 10⁻⁵. Methods for assessment of pharmacokinetics, CD38-receptor occupancy and density, RNA sequencing, immunoprofiling, and systemic cytokines are detailed in the supplemental Appendix.

Efficacy and safety analyses were performed in the safety analysis set, which comprised all enrolled patients who received ≥1 dose of the study drug, including incomplete doses. The safety analyses included data from the on-treatment period, which was defined as the time from the first dose to 30 days after administration of the last dose of study drug. For binomial efficacy end points, frequencies, percentages, and the exact 2-sided 95% confidence intervals (CIs) were summarized. For time-to-event analysis, the Kaplan-Meier survival curves, 25th, 50th (median), and 75th percentiles, along with the associated 2-sided 95% CIs based on the Brookmeyer and Crowley method, and Kaplan-Meier probability estimates with 95% CIs at 3 and 6 months (or later time points if data permitted), are presented. No formal statistical comparisons were planned.

We treated 106 patients (56 in dose escalation, and 50 in dose expansion; Figure 1) between 9 October 2017, and the data cutoff date of 27 October 2023. We identified 2 recommended phase 2 doses of single-agent modakafusp alfa that had similar response rates and safety profiles, and our analysis focuses on the 37 patients treated at these doses: 30 patients treated at 1.5 mg/kg Q4W (5 in dose escalation; 25 in dose expansion) and 7 patients treated at 3 mg/kg Q4W. Expansion cohorts at modakafusp alfa 0.4 mg/kg Q3W alone (n = 8) and with dexamethasone (n = 3) were terminated early for lack of myeloma responses. Enrollment to the 1.5 mg/kg Q4W plus dexamethasone expansion cohort was ongoing (n = 14) at the time of data cutoff.

Patients received a median of 2.5 treatment cycles (range, 1-47) overall, and 4 cycles (range, 1-41) in the 1.5 mg/kg Q4W cohort. In total, 97 patients discontinued treatment, most because of progressive disease (in 74 patients [76.3%]; Figure 1). The median follow-up was 7 months (range, 0.2-43.3) in all 106 patients, and 5.3 months (range, 0.5-36.8) in the 30 patients who received single-agent modakafusp alfa 1.5 mg/kg Q4W.

Patient demographics and disease characteristics were similar for the entire trial population and in the single-agent 1.5 mg/kg Q4W and 3 mg/kg Q4W cohorts (Table 1; supplemental Tables 1 and 2). The median age was 64.5 years. Patients had received a median of 6.5 previous lines of antimyeloma therapy (range, 3-19) over a median of 5.6 years (range, 1.4-22.7) since diagnosis. Eighty-nine (84%) patients had disease refractory to an anti-CD38 antibody and 88 (83%) had disease refractory to an IMiD, a PI, and an anti-CD38 antibody (triple-class refractory).

Details of the initial dose escalation and DLTs are shown in supplemental Table 4 and Figure 1. In the 20 patients initially enrolled on the modified QW schedule, we observed dose-limiting grade 4 neutropenia (n = 2), grade 4 thrombocytopenia (n = 1), and grade 4 neutropenia and grade 3 thrombocytopenia (n = 1), and this schedule was discontinued. The Q2W schedule (n = 8) was also discontinued because of DLTs (grade 3 thrombocytopenia, n = 1; grade 4 thrombocytopenia, n = 1; grade 3 encephalopathy; and grade 4 thrombocytopenia, n = 1). In the Q3W schedule (n = 7), no DLTs were observed at 0.75 mg/kg, but 2 of 3 patients required dose delays because of thrombocytopenia. Of 21 patients treated at various doses on the Q4W schedule during dose escalation, 2 patients (at 3 and 6 mg/kg) required a dose delay because of hematologic toxicities. At 6 mg/kg Q4W, 2 patients had DLTs of grade 3 IRR (n = 1) and grade 4 thrombocytopenia (n = 1), and the MTD was determined as 3 mg/kg Q4W.

Among all 106 patients, the most common AEs were hematologic toxicities (Table 2; supplemental Tables 5 and 6). Thirteen (12.3%) patients had serious drug-related AEs, and 13 had AEs leading to study drug discontinuation (supplemental Table 7). Twenty-four (22.6%) patients had IRRs, which first occurred after a median of 1.5 cycles (range, 1-5) and were observed in some patients through cycle 35.

Among the 30 patients in the 1.5 mg/kg Q4W cohort from escalation (n = 5) and expansion (n = 25), the most frequent grade 3 or 4 AEs were neutropenia (66.7%; grade 4, 33.3%) and thrombocytopenia (46.7%; grade 4, 23.3%; Table 2). Absolute neutrophil and platelet counts tended to nadir mid-cycle and generally recovered in time for the next scheduled cycle (supplemental Figure 1). Eleven (36.7%) patients received granulocyte colony-stimulating factor, whereas 2 (6.7%) received thrombopoietin receptor agonists, and 9 (30.0%) received platelet transfusions. One patient had significant bleeding (grade 3), and infections were reported in 8 (26.7%) patients, including grade 3 events in 5 (16.7%) patients. Three patients had serious drug-related AEs (pneumonia, n = 1; sinusitis, n = 1; and IRRs, n = 2). The only grade 4 nonhematologic AE was 1 incident of hyperuricemia. IRRs occurred in 11 (36.7%) patients (grade 1 to 2 in 10 patients, grade 3 in 1 patient), and included chills in 7 (23.3%), headache in 4 (13.3%), pruritus in 3 (10%), and pyrexia in 3 (10%) patients. In escalation, all IRRs occurred during the infusion; in expansion, 36.8% occurred during the infusion, whereas 31.6% were reported within 2 hours after the infusion.

Safety outcomes were generally similar in the 7 patients treated at the MTD of 3 mg/kg Q4W to those reported among patients treated at 1.5 mg/kg Q4W (Table 2; supplemental Table 7).

Dose delays because of AEs were reported in 9 of 30 patients in the 1.5 mg/kg Q4W cohort, and in 4 of 7 patients in the 3 mg/kg Q4W cohort; dose modifications were reported in 16 of 30 and 6 of 7 patients in these cohorts, respectively.

Overall, across all single-agent modakafusp alfa doses and schedules (N = 89), 20 patients had a response (Figure 2A). Among 30 patients in the 1.5 mg/kg Q4W cohort, 13 patients achieved a partial response or better, resulting in an ORR of 43.3% (95% CI, 25.5-62.6; Figure 2B). This included 1 stringent CR, 2 CRs, and 6 very good partial responses (VGPRs). The median time to response was 1.1 months, and median DOR was 15.1 months (95% CI, 7.1-26.1); median progression-free survival was 5.7 months (95% CI, 1.2-14). Among the 7 patients receiving the MTD of 3 mg/kg Q4W, the ORR was 42.9%, with 1 patient achieving a CR and 2 achieving a VGPR (Figure 2A-B); median DOR was 10.2 months (95% CI, 7.9 to not reached). Of 4 patients evaluated for MRD at a sensitivity of 10⁻⁵ (3 patients from the 1.5 mg/kg Q4W cohort and 1 patient from the 3 mg/kg Q4W cohort), 1 patient (1.5 mg/kg Q4W) was reported as MRD negative in cycle 8.

Exploratory subgroup analyses for the 37 patients in the 1.5 and 3 mg/kg Q4W single-agent cohorts showed ORRs of 38% in patients with myeloma refractory to an anti-CD38 antibody, 38% in patients with triple-class–refractory disease, 41% in penta-exposed patients (ie, exposed to bortezomib, carfilzomib, lenalidomide, pomalidomide, and an anti-CD38 antibody), 28% in patients with prior anti–B-cell maturation antigen (BCMA) therapy, 58% in patients without prior anti-BCMA exposure, and 43% in patients with extramedullary disease (Figure 2C; supplemental Figure 2). Of 4 patients who had disease progression while receiving daratumumab in the most recent line of therapy, 3 had responses of VGPR or better (1 patient withdrew in cycle 1 and could not be assessed for response); all 4 patients were treated in dose expansion with no required washout for daratumumab (time between the last dose of daratumumab and first dose of modakafusp alfa ranging from 14-55 days). After a median follow-up of 6.5 months, the median progression-free survival for all 37 patients was 5.7 months (95% CI, 2.1-13.3).

Modakafusp alfa serum concentration peaked at approximately the end of infusion and then declined in a biphasic nonlinear manner (supplemental Figure 3). Based on the dose-normalized area-under-the-curve, exposure increased in a greater than dose-proportional manner at 0.1 to 1.5 mg/kg, and approximately dose-proportional manner at 1.5 to 6 mg/kg. The apparent nonlinear pharmacokinetics is consistent with saturable target-mediated elimination. The geometric mean terminal half-life in serum for modakafusp alfa 1.5 and 3 mg/kg Q4W was 10 and 17 hours, respectively.

Modakafusp alfa was shown to bind to its target, CD38, on the surface of viable leukocytes (immune and myeloma cells) of peripheral blood (Figure 3A) and bone marrow (Figure 4A). Activation of the interferon pathway by modakafusp alfa was demonstrated by upregulation of the type 1 interferon gene signature score (comprising the 25 most upregulated genes in response to interferon beta therapy)¹⁸ in the peripheral blood (Figure 3B; supplemental Figure 4A) and bone marrow (Figure 4B). Further evidence was seen as CD38 receptor density increased on natural killer (NK) cells, T cells, and myeloma cells in the peripheral blood (Figure 3C; supplemental Figure 4B-D), with similar trends noted in the bone marrow (Figure 4C-E). Modakafusp alfa treatment resulted in increased activation of peripheral NK and CD8 T cells (Figure 3D) and increased transient production of interferon-associated cytokines, such as monocyte chemoattractant protein 2 (Figure 3E) and interferon gamma–induced protein 10 (supplemental Figure 5A), among others (supplemental Figure 5B-D). In contrast, changes in serum concentrations of interleukin-6 (supplemental Figure 5E) and tumor necrosis factor α (supplemental Figure 5F) were not observed after modakafusp alfa treatment. Pharmacodynamic findings revealed significant changes from baseline in specific markers in the peripheral blood after treatment with modakafusp alfa; however, not all changes were dose responsive (supplemental Figure 4).

Our first-in-human phase 1/2 trial demonstrates that modakafusp alfa is well tolerated at doses that have single-agent activity in patients with heavily pretreated R/R MM. In more than 100 patients treated across 13 doses and schedules, we identified an optimal dosing schedule of Q4W, an MTD of 3mg/kg Q4W, and an additional active dose of 1.5 mg/kg Q4W. We have identified these as our recommended phase 2 doses, and the single-agent ORR at these doses was 43%. The main toxicities associated with this anti-CD38 antibody/interferon fusion protein were hematologic, primarily neutropenia and thrombocytopenia, and we did not observe the neuropsychiatric toxicities typically associated with interferon alfa therapy. We were able to demonstrate activation of interferon signaling in both myeloma cells and immune cells, suggesting that modakafusp alfa has a dual mechanism of action, with both direct antimyeloma effect and broad stimulation of the adaptive and innate immune systems. The combination of good overall tolerability, antimyeloma activity, and a novel mechanism of action suggest that targeting interferon α signaling to CD38⁺ cells is a promising approach for myeloma therapy.

The single-agent ORR of 43% that we observed with modakafusp alfa at our recommended phase 2 doses is clinically meaningful, especially with a median DOR of 10 to 15 months. Acknowledging the inherent limitations of cross-trial comparisons, this response rate compares favorably to those of other therapies evaluated in heavily pretreated myeloma, including selinexor (in combination with dexamethasone, 21%-26%),^19,20 daratumumab (29%),²¹ belantamab mafodotin (31%-34%),²² and melflufen (in combination with dexamethasone, 31%).²³ 

In patients treated at the recommended phase 2 doses whose disease was previously refractory to anti-CD38 antibody therapy, the ORR was 38%, including responses in 3 of 4 patients who had been treated with an anti-CD38 monoclonal antibody as part of the immediate prior therapy. These responses demonstrate that the mechanism of action of modakafusp alfa is distinct from that of anti-CD38 monoclonal antibodies, because no responses were seen with the anti-CD38 monoclonal antibody isatuximab in patients whose disease was previously refractory to daratumumab.²⁴ We also observed significant numbers of responses in patients with triple-class–refractory disease, penta-exposed disease, and prior exposure to anti-BCMA therapies. For patients without prior anti-BCMA therapy, the ORR was 58%, similar to that of the bispecific T cell engager teclistamab (63%).²⁵ A lower ORR of 28% was observed among patients with prior BCMA-targeted therapy; the lower response rate may be because these patients had received more extensive overall prior therapy (median of 9 vs 5 prior lines among our BCMA-naïve patients).

Modakafusp alfa has a unique structure and mode of action. It is a fusion protein without the potential for dissociation of the attenuated interferon from the targeting antibody. The IgG4 backbone has limited effector function,⁸ distinguishing it from currently approved anti-CD38 antibodies, which have mechanisms of action dependent on IgG1 Fc effector functionality.^10,11 In addition, modakafusp alfa has a unique binding epitope on CD38 and does not compete for binding with daratumumab or isatuximab.²⁶ Our pharmacodynamic analyses demonstrate that modakafusp alfa binds to CD38 and induces interferon signaling within targeted cells, resulting in NK and CD8 T cell activation, proliferation, and enhanced cytotoxic function, as well as systemic production of interferon-associated cytokines. The observed upregulation of CD38 expression on NK, T, and myeloma cells is consistent with the CD38 gene containing interferon response elements in its promoter region.^27-29 These characteristics of modakafusp alfa demonstrate a mechanism of action that is distinct from other anti-CD38 monoclonal antibodies and suggest that modakafusp alfa might complement anti-CD38 therapy by upregulating CD38 expression.

With respect to toxicities, treatment with modakafusp alfa resulted in a relatively high rate of severe neutropenia and thrombocytopenia. However, the rate of infections with modakafusp alfa was low (26.7% all-grade and 16.7% grade 3 with 1.5 mg/kg Q4W) and serious bleeding occurred in 1 patient only. IRRs were almost all grade 1/2 and observed at a lower rate (36.7%) than what is commonly reported with intravenously administered anti-CD38 antibodies (∼50%).³⁰ Despite the immune activation observed with modakafusp alfa, we did not observe any instances of cytokine release syndrome or immune effector cell–associated neurotoxicity syndrome reported with T cell–redirecting therapies.^25,31,32 As noted above, we also did not observe the neuropsychiatric and constitutional side effects typically seen with interferon alfa therapy.

The main limitation of our study is the small number of patients treated at each dose level. Because both 1.5 and 3 mg/kg Q4W regimens resulted in good response rates with similar safety profiles without a clear pharmacodynamic distinction as to which would be optimal, a randomized study has been completed to compare dose levels of 120 and 240 mg (fixed-dose equivalents to 1.5 and 3 mg/kg) Q4W to define the single-agent dose with the optimal benefit-to-risk profile (manuscript in preparation).

In summary, modakafusp alfa is an immunocytokine that delivers interferon α to CD38⁺ cells, resulting in immune activation and antitumor activity. The immune stimulatory effects of modakafusp alfa, along with its manageable toxicity profile, monthly administration, and distinct CD38 binding epitope, suggest a potential for combination with a wide range of antimyeloma therapies including IMiDs, PIs, anti-CD38 antibodies, and T cell–redirecting therapies. The association between modakafusp alfa–induced immune stimulation and myeloma response remains to be determined and presents a unique opportunity to elucidate the role of the immune system in oncologic disease. The single-agent activity warrants continued exploration of modakafusp alfa for the treatment of MM.

The authors thank the patients and their caregivers, as well as the physicians, nurses, trial coordinators, and research staff for participation in the trial. The authors also thank all modakafusp alfa clinical team members, particularly LiLi Yang, of Takeda Development Center Americas, Inc (TDCA) for overseeing the pharmacokinetics assay validation and sample analysis; Joanna Pye and Shuli Li of TDCA for providing programming and statistical support; and Faith Dunbar of TDCA for additional biostatistics support. Medical writing support for the development of this manuscript, under the direction of the authors, was provided by Luisa Madeira of Ashfield MedComms, an Inizio company, funded by Takeda Pharmaceuticals U.S.A., Inc, and complied with Good Publication Practice guidelines.³³ 

This trial was funded by TDCA.

Contribution: D.T.V., O.N., N.B., K.S., C.L., S.C., X.P., and J.L.K. were responsible for study conception and design; D.T.V., S.A., S.A.H., O.N., D.B., M.C., K.S., C.L., Y.L., S.C., X.P., and J.L.K. were responsible for the acquisition, analysis, or interpretation of data; and all authors contributed to writing the manuscript, reviewed and approved the final version before submission, and are accountable for all aspects of the work.

Conflict-of-interest disclosure: D.T.V. reports consulting fees from Takeda, GSK, Genentech, Janssen, Karyopharm, and Oncopeptides; and grants or funds from Takeda and Active Biotech. S.A. reports consulting fees from Janssen and GSK, and grants or funds from Amgen, Janssen, GSK, Bristol Myers Squibb, Karyopharm, and Incyte. S.A.H. reports participation on advisory council or committee for Takeda; consulting fees from AbbVie, Bristol Myers Squibb/Celgene, Janssen, Oncopeptides, Secura Bio, and Takeda; and grants or funds from Bristol Myers Squibb and Oncopeptides. O.N. reports participation on advisory council or committee for Janssen, Bristol Myers Squibb, Karyopharm, Sanofi, Takeda, GPCR Therapeutics, GSK, and Adaptive Biotechnologies; consulting fees from GPCR Therapeutics; and grants or funds from Janssen and Takeda. M.C. reports employment with Bristol Myers Squibb. N.B. reports consulting fees from Janssen, Sanofi, Bristol Myers Squibb, GSK, AbbVie, and Pfizer; and participation in speakers bureau for Janssen. K.S., S.C., and C.L. report employment with and ownership of stocks/shares in Takeda. Y.L. and X.P. report employment with Takeda. X.P. reports owning stock/shares in Takeda. J.L.K. reports consulting fees from Bristol Myers Squibb, Ascentage, Roche, Sanofi, and Sebia. D.B. declares no competing financial interests.

Correspondence: Dan T. Vogl, Abramson Cancer Center, Perelman School of Medicine, University of Pennsylvania, Perelman Center 12-176, 3400 Civic Center Blvd., Philadelphia, PA 19104; email: dan.vogl@pennmedicine.upenn.edu.