Title: Clinical and Biological Evaluation of the Novel CD30/CD16A Tetravalent Bispecific Antibody (AFM13) in Relapsed or Refractory CD30-Positive Lymphoma with Cutaneous Presentation: A Biomarker Phase Ib/IIa Study (NCT03192202)

Introduction: Natural Killer cells (NK) play an important role in tumor immune-surveillance. The innate cell engager AFM13 is a CD30/CD16A targeting high affinity bispecific tetravalent antibody that engages and activates NK. This study evaluates AFM13’s clinical and immunological activity. It examines the immunologic changes in the tumor and in peripheral blood (PB) as a function of the dose and method of administration of AFM13.

Methods: Subjects with relapsed or refractory CD30 expressing lymphoma with cutaneous involvement were recruited into 4 cohorts: 1.5 mg/kg IV weekly, 7 mg/kg IV weekly, 7 mg/kg continuous intravenous infusion (CIVI) over 5 days weekly and 200mg IV weekly. Each cohort consisted of 3 subjects with 3 additional subjects in cohort 4. Subjects received 8 weeks of therapy for up to 2 cycles. Response assessment was performed on week 11 of each cycle by mSWAT, photography, PET imaging and PB flow cytometry. A second cycle was administered if there was no progression. Subjects underwent skin biopsies and PB immunologic studies as follows: pretreatment, day 5, week 4 and week 8 of therapy. Biopsies were analyzed and evaluated by a pathologist and IHC image analyzer. PB samples were analyzed by flow cytometry.

Results: Fourteen subjects were accrued. The median number of prior therapies was 4 (1-11) and 4 subjects had progressed on brentuximab vedotin. The disease etiologies, treatment emergent toxicities and response by cohort are presented in Table 1. An objective response rate of 40% was observed in the study.

In the PB, flow cytometry revealed a decrease in circulating NK (CD16/56+ CD3- CD19- lymphocytes) during therapy with post therapy recovery. The activation marker CD69 on NK increased in responders (R) compared to non-responders (NR). Similarly, tumor biopsies in R showed increased infiltration and activation of CD56+ NK as opposed to NR (Figure 1). NK cytotoxicity through the expression of Granzyme B was seen in R vs. NR. Flow quantitation of circulating CD4+ CD25+ T cells (Tregs) showed a decrease in R vs. NR.

Conclusions: AFM13 demonstrated a high ORR of 40% among a population of heavily pretreated subjects with CD30 positive lymphoproliferative T-cell malignancies. AFM13 exhibited activity post brentuximab vedotin failure. In addition, biological changes in NK infiltration and activation in the PB and tissue biopsy correlated with response. These data are the first to demonstrate that innate cell engagers modulate NK and T cell populations in the peripheral blood and tumor from patients and to determine that such changes are associated with patient benefit.