A First-in-Human Phase I Study of Subcutaneous Outpatient Recombinant Human IL15 (rhIL15) in Adults with Advanced Solid Tumors

Abstract

Purpose: Preclinical data established IL15 as a homeostatic factor and powerful stimulator of NK and CD8⁺ T-cell function, the basis for clinical testing.Experimental Design: A first-in-human outpatient phase I dose escalation trial of subcutaneous (SC) rhIL15 was conducted in refractory solid tumor cancer patients. Therapy consisted of daily (Monday-Friday) subcutaneous injections of rhIL15 for two consecutive weeks (10 total doses/cycle). Clinical response was assessed by RECIST. Pharmacokinetics of rhIL15 and immune biomarkers were evaluated.Results: Nineteen patients were treated with rhIL15 at dose levels of 0.25, 0.5, 1, 2, and 3 mcg/kg/day. Fourteen patients completed ≥ 2 cycles of therapy that was well tolerated. One serious adverse event (SAE), grade 2 pancreatitis, required overnight hospitalization. Enrollment was halted after a patient receiving 3 mcg/kg/day developed a dose-limiting SAE of grade 3 cardiac chest pain associated with hypotension and increased troponin. No objective responses were observed; however, several patients had disease stabilization including a renal cell carcinoma patient who continued protocol treatment for 2 years. The treatment induced profound expansion of circulating NK cells, especially among the CD56^bright subset. A proportional but less dramatic increase was found among circulating CD8⁺ T cells with maximal 3-fold expansion for the 2 and 3 mcg/kg patients.Conclusions: SC rhIL15 treatment was well tolerated, producing substantial increases in circulating NK and CD8⁺ T cells. This protocol establishes a safe outpatient SC rhIL15 regimen of 2 mcg/kg/day dosing amenable to self-injection and with potential as a combination immunotherapeutic agent. Clin Cancer Res; 24(7); 1525-35. ©2017 AACR.

Trial registration: ClinicalTrials.gov NCT01727076.

Figure 1.. rhIL-15 pharmacokinetics and cytokine responses.

Blood samples were collected from enrolled subjects before, and 10 minutes, 1, 4 and 24 hours after the first dose of rhIL-15 given subcutaneously during cycle 1. Serum was frozen until batch ELISA testing was performed for IL-15 (A), IL-6 (B), and IFN-γ (C) levels, as described in Materials and Methods. Mean results from each dose cohort are shown with error bars (±1SD); the y-axis is shown in log (A) or linear scales (B, C). Error bars for panels (B) and (C) were omitted to facilitate viewing of the data because error bars for all dose cohorts overlapped, even at the 4-hour peak time point.

Figure 2.. Circulating lymphocyte, NK and CD8+ T cell numbers before and during rhIL-15 treatment.

Absolute lymphocyte counts were calculated from CBC data obtained daily from individual subjects’ local labs for the 2 mcg/kg/day (yellow, N=6) and 3 mcg/kg/day (blue, N=3) dose cohorts, and includes a pre-cycle 1 day1 (C1D-P) time point (A). Each line/symbol represents mean results from each dose cohort with error bars (±1SD). Green bars represent periods of daily rhIL-15 treatment. Absolute cell frequencies (cells/mcL) of CD45+CD3-CD56+ NK cells in fresh whole blood were measured using Trucount tubes and are shown as means (±1SD) grouped by dose cohort (B, left). Absolute cell frequencies (cells/mcL) of CD45+CD3+CD8+ T cells among subjects treated with 2 or 3 mcg/kg/day of rhIL-15 is shown (B, right). Each line/symbol represents results from a single individual from the 2 and 3 mcg/kg/day dose cohorts as shown in (A). Mean fold-increases for whole blood NK and CD8+ T cell frequencies during cycle 1 at days 11 or 15 (whichever was available and/or maximal) compared to baseline (day 1) for all treated subjects is shown, grouped by dose cohort (C). Mean CD56+ NK cell fold increases (± 1SD) for the 0.25 (N=3), 0.5 (N=3), 1 (N=3), 2 (N=6), and 3 (N=3) mcg/kg dose cohorts were 2.3 (± 1.2), 3.3 (± 2.3), 4.4 (± 3.2), 10.8 (± 8.2), and 13.5 (± 6.6) respectively. Mean CD8+ T cell fold increases (± 1SD) for the 0.25, 0.5, 1, 2, and 3 mcg/kg dose cohorts were 1.1 (±0.2), 1.0 (±0.2), 1.1 (±0.2), 3.3 (± 3.8), and 2.8 (± 0.6) respectively.

Figure 3.. Circulating NK cell subset expansion and NK cell function during rhIL-15 treatment.

Whole blood samples were analyzed for CD3-CD56+ NK cell subset frequencies of CD56^bright (left) and CD56^dim (right) NK cells using multiparametric flow cytometry as described (A). Individual subject data are represented by a single line/symbol, in yellow for the 2 mcg/kg/day (N=6) and blue for the 3 mcg/kg/day (N=4) dose cohorts (A). Green bars represent periods of daily rhIL-15 treatment. Mean fold-increases during cycle 1 at days 11 or 15 (whichever was available and/or maximal) compared to baseline (day 1) for treated subjects, grouped by dose cohort, are indicated by each column for whole blood CD56^bright and CD56^dim NK cells (B). Mean CD56^bright NK cell fold increases (± 1SD) for the 0.25 (N=3), 0.5 (N=3), 1 (N=3), 2 (N=6), and 3 (N=4) mcg/kg dose cohorts were 4.6 (± 1.6), 6.4 (± 4.1), 6.9 (± 2.2), 39.7 (± 54.4), and 74.6 (± 74.4) respectively. Mean CD56^dim NK cell fold increases (± 1SD) for the 0.25, 0.5, 1, 2, and 3 mcg/kg dose cohorts were 2.2 (± 1.2), 3.4 (± 2.5), 4.3 (± 3.6), 7.9 (± 4.7), and 9.8 (± 7.0) respectively. Mean values are shown above each column. Cryopreserved samples obtained before rhIL-15 initiation (N=11) and at Day 11 (N=11) and Day 15 (N=8) during/after treatment were assessed for evidence of active proliferation by intracellular Ki67 labeling of NK cells, T cells and T cell subsets as described (C). NK cell degranulation (C, left panel) or intracellular TNF-α (C, right panel) measured after no stimulation or stimulation with K562 for 5 hours is shown. Cryopreserved samples obtained before rhIL-15 initiation (N=11) and at Day 11 (N=11) and Day 15 (N=8) during/after treatment were assessed for evidence of active proliferation by intracellular Ki67 labeling of NK cells, T cells and T cell subsets as described (D).

Figure 4.. Development of anti-rhIL-15 antibodies and assessment of neutralizing capacity.

All enrolled subjects were tested for the presence of anti-rhIL-15 serum antibodies at baseline, at the start of each cycle, and at their final blood draw for correlative testing. Of 19 subjects, only 3 developed antibodies against rhIL-15, and serum antibody levels for these 3 subjects re shown (A). For the graph x-axis in panel A, “F” refers to the 6-month follow-up visit after study withdrawal. Serum from two subjects, Subject 3 (0.25mcg/kg/day, white and black bars) and Subject 5 (0.5mcg/kg/day, red and black bars) were tested for their ability to inhibit the proliferation of NK92, an IL-15 dependent cell line, in the presence of rhIL-15 (B) or rhIL-15 + IL-15 receptor-α (C) in a ³H-thymidine incorporation assay. Means of triplicate values ±1 SD are shown.