INDUCE-2: A Phase I/II, open-label, two-part study of feladilimab in combination with tremelimumab in patients with advanced solid tumors

INDUCE-2 was a Phase I/II, open-label study investigating feladilimab in combination with tremelimumab with a two-part design (Supplementary Fig. 1). Part 1 enrolled patients with advanced selected solid tumors in a dose-escalation design (described below). Part 2 was planned to investigate dose expansion in patients with RR HNSCC who had progressed after receiving at least one platinum-based chemotherapy and at least one anti-PD-1/PD-L1 therapy (whether in combination or given as monotherapy). The study comprised three periods: screening (assessments up to 30 days prior to first dose); treatment (until disease progression, unacceptable toxicity, or death); and follow-up.

Eligible patients for Part 1 were ≥ 18 years of age with a documented diagnosis of a locally advanced or metastatic relapsed solid tumor of cutaneous melanoma, HNSCC, NSCLC, UC, clear cell renal cancer, or castrate-resistant prostate cancer, who had progressed following SOC therapy. Selection of these tumors was based on evidence of their reported and/or predicted response to immune checkpoint therapies [15‐20]. Patients were also required to have measurable disease as per the Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v.1.1), an Eastern Cooperative Oncology Group Performance Status of 0 or 1, adequate organ function, and consent to collection of tumor tissue samples [21, 22]. Key exclusion criteria included patients who had received prior anti-CTLA-4 or ICOS-directed therapies, patients who had received four or more lines of therapy, or those who had received systemic anticancer therapies within 30 days of first dose. The full list of exclusion criteria are as follows: patients with solid tumors other than those listed above; patients who had received prior anti-CTLA-4 or ICOS-directed therapies; patients who had received at least four lines of therapy, or who had received systemic anti-cancer therapies within 30 days of first dose; unresolved prior treatment-related toxicity; central nervous system metastasis; autoimmune disease or syndrome requiring systemic treatment within the past 2 years; systemic steroids or other immunosuppressive agents within 7 days; history of idiopathic pulmonary fibrosis or pneumonitis; history of severe hypersensitivity to monoclonal antibodies; history or evidence of cardiac abnormalities within 24 weeks of enrollment; current unstable liver or biliary disease; patients with an active infection requiring systemic therapy.

For Part 1, feladilimab/tremelimumab was administered at 8/75 mg; 24/75 mg; 8/225 mg; 80/75 mg; 24/225 mg; and 80/225 mg. Feladilimab doses were selected based on the safety and exposure data observed in the Phase I first-time-in-human study (INDUCE-1; NCT02723955), along with predicted target engagement based on in vitro potency and feladilimab exposure predicted from a population pharmacokinetic model [13, 14]. Tremelimumab doses were selected based on prior monotherapy and combination studies [23‐25]. Dose escalation followed a zone-based approach guided by the bivariate Continuous Reassessment Method model until the maximum tolerated dose or maximum administered dose combination was determined (Supplementary Fig. 1). Feladilimab was administered once every 3 weeks (Q3W); tremelimumab was administered Q3W for 6 doses, and then every 12 weeks (Q12W) thereafter. Tremelimumab was administered intravenously (IV) over an initial 60-min infusion, followed by feladilimab over a 30-min IV infusion administered 1–2 h following the end of tremelimumab infusion. The RP2D determined in Part 1 of the study was planned to be carried forward to Part 2.

The primary objective of Part 1 of the study was to determine the safety, tolerability, and RP2D of feladilimab in combination with tremelimumab. Endpoints reported here include the frequency and severity of dose-limiting toxicities (DLTs), adverse events (AEs), AEs of special interest (AESIs), serious AEs (SAEs) and DLTs/AEs/SAEs leading to dose modifications, delays, or withdrawals. Changes in laboratory parameters (for example, serum analytes including albumin, alkaline phosphatase (ALP), alanine transaminase (ALT), aspartate transaminase (AST), creatinine, and sodium), vital signs, and safety parameters were also assessed.

Secondary endpoints included investigator-assessed clinical activity of feladilimab in combination with tremelimumab, evaluated as best overall confirmed response per RECIST v1.1 and immune RECIST (iRECIST), and pharmacokinetic (PK) parameters of feladilimab in combination with tremelimumab [22]. Blood and tumor-based biomarkers were assessed as exploratory endpoints.

AEs were coded using Medical Dictionary for Regulatory Activities (MedDRA), grouped by system organ class, and graded by the Investigator according to the National Cancer Institute – Common Terminology Criteria for Adverse Events (NCI-CTCAE) v.5.0. Radiographic evaluations of the chest, abdomen, and pelvis were conducted according to RECIST v1.1 at 9 weeks after the first study dose, and then every 6 weeks until Week 52, and Q12W thereafter [22]. PK profiles of feladilimab and tremelimumab were measured from patient blood samples collected at protocol-defined points, and concentrations calculated using validated bioanalytical methodologies.

Whole blood samples were collected pre-dose on Weeks 1, 2, 4, 7, 13, and 25, and shipped for in-stream flow cytometry analysis for markers of T-cell activation and proliferation at Covance Central Laboratory. Serum samples were analyzed for pro-inflammatory cytokine induction using the MSD 10-plex panel (Q-Squared Laboratories). Paired tumor biopsy samples (defined as a pretreatment sample collected any time after the end of previous therapy and prior to the first dose of study intervention/SOC, and an on-treatment sample collected at Week 7) were analyzed by multiplex immunofluorescence using a custom 18-marker MultiOmyx panel (NeoGenomics Laboratories, CA, USA). Gene expression analysis was performed on paired tumor biopsies using the Nanostring PanCancer IO360 panel.

A maximum of 26 patients were enrolled for Part 1. All patients who received at least one dose of tremelimumab or feladilimab were included for the safety assessment summary. Efficacy data were initially planned to be analyzed using Kaplan–Meier analysis for time-to-event data; however, the OS and progression-free survival data were later only listed for clinical review purposes. PK parameters were presented as descriptive summary statistics. No statistical analyses were performed on blood- and tumor-based biomarker data due to small sample numbers.

A total of 42 patients were screened, of which 26 were enrolled and allocated to feladilimab/tremelimumab 8/75 mg (n = 1), 24/75 mg (n = 1), 8/225 mg (n = 5), 80/75 mg (n = 3), 24/225 mg (n = 16), respectively (Supplementary Fig. 2). At study completion, 18 (69%) patients had completed the study per protocol and 8 (31%) patients had withdrawn from the study (Supplementary Fig. 2). Based on the totality of available data from Part 1, and not due to safety, the decision was made to stop the study following feladilimab/tremelimumab dosing at 24/225 mg and not to continue to dose level 80/225 mg, or to move into Part 2 of the study.

Most patients were male (n = 17, 65%) and White/Caucasian (n = 24; 92%), and the median age was 65 (range: 36–83) years (Table 1). At initial diagnosis, primary tumor types included NSCLC (n = 9; 35%), HNSCC (n = 8; 31%), melanoma (n = 5; 19%), and bladder cancer (n = 4; 15%). The most common histology was adenocarcinoma (n = 10; 38%) followed by squamous cell carcinoma (n = 9; 35%). The median time since initial diagnosis was 410.5 (range: 27–2019) days. All patients had received prior anti-cancer therapy: 88% (n = 23) had received immunotherapy; 77% (n = 20) chemotherapy; 65% (n = 17) radiotherapy; and 85% (n = 22) had undergone cancer-related surgery (Table 1). Most patients who received immunotherapy, chemotherapy, and/or radiotherapy had received one prior treatment regimen of that type, and most patients who had undergone cancer-related surgery had undergone three prior procedures.

Median time of exposure ranged from 6.21 to 18.29 weeks for feladilimab, and 6.14 to 15.29 weeks for tremelimumab. The number of infusions ranged from three to six for feladilimab, and from three to five for tremelimumab. One patient had a dose delay for both feladilimab and tremelimumab due to an AE (hypoxia) and no patients had a dose interruption or incomplete infusion. One patient in the highest dose group (feladilimab 24 mg/tremelimumab 225 mg), experienced a DLT 18 days after the first dose of study treatment (Grade 3 diarrhea considered possibility related to treatment). A flexible sigmoidoscopy on study day 23 revealed moderately congested, erythematous mucosa with mucosal prolapse. Due to concerns of immunotherapy-induced colitis, treatment was discontinued and after intervention the Grade 3 diarrhea resolved on Day 32.

All patients experienced at least one AE and 19 (73%) patients experienced treatment-related (TR)-AEs (Table 2), the most common of which were diarrhea (n = 9; 35%) and fatigue (n = 8; 31%). SAEs occurred in 12 (46%) patients, of which two (8%) were considered to have experienced TR-SAEs (one case each of diarrhea and colitis). AEs (colitis, diarrhea, fatigue, acute kidney injury, and dyspnea) led to permanent discontinuation in four (15%) patients (Tables 2 and 3). One fatal SAE, considered related to disease progression, occurred during the study (Table 2). All deaths (n = 17; 65%) were attributed to disease under study, and most (n = 15; 88%) occurred over 30 days after the last dose of study treatment (data not shown).

The most frequently reported AEs of any cause were fatigue, diarrhea, and nausea (Supplementary Table 1). Most patients experienced a Grade ≥ 3 AE (n = 15; 58%), the most frequently reported being fatigue, acute kidney injury, anemia, diarrhea, dyspnea, embolism, and hypoxia (data not shown). Grade ≥ 3 TR-AEs occurred in five (19%) patients, four of which received feladilimab 24 mg/tremelimumab 225 mg (Supplementary Table 2). Nine (35%) patients experienced at least one AESI, and of these events all five skin-related AESIs, both cases of colitis, one of two cases of nephritis and renal function, and the one case of hepatitis were considered treatment-related (Table 3).

One (4%) patient with NSCLC in the feladilimab 24 mg/tremelimumab 225 mg group achieved a confirmed partial response (PR), completed the study per protocol, and remained on study treatment for 337 days (Fig. 1). Stable disease (SD) was observed in five (19%) patients (including one patient with HNSCC, three patients with NSCLC, and one patient with melanoma). Of these patients, one was in the feladilimab/tremelimumab 8/225 mg group, two in the 80/75 mg group, and two in the 24/225 mg group (Fig. 1). Three patients (two in the 80/75 mg group and one in the 24/225 mg group) had SD ≥ 18 weeks.

Investigator-assessed best percent reduction from baseline in tumor measurement is shown in Supplementary Fig. 3a, and percent change in tumor measurement over time from baseline in Supplementary Fig. 3b.

Median feladilimab plasma concentration–time data during Treatment Cycle 1 are shown in Supplementary Fig. 4 and Supplementary Table 3. Feladilimab exposure in combination with tremelimumab was dose proportional and generally consistent with expected patterns of accumulation for a mAb (Supplementary Table 3).

The longitudinal assessment of peripheral immune cell populations showed a slight decrease in the percentage of CD3+ , CD4+ , and CD8+ T cells post-treatment across most dose levels (Fig. 2a). Similar trends were observed in pooled dose analyses showing the maximum pharmacodynamic changes for individual patients (Supplementary Fig. 5). Of the 10 evaluated cytokines from peripheral blood, only induction of interferon gamma (IFN-γ) was noted in most of the patients. At Week 2 (one week post-first dose), a ≥ twofold induction was observed in 11 out of 25 evaluated patients (Fig. 2b).

Multiplex immunofluorescence and gene expression analyses of paired tumor biopsies did not show any clear or consistent changes in immune cell markers. Pooled analysis of data from six paired tumor biopsies across several dose levels and tumor types did not show any statistically significant changes in T-cell populations or immune markers following combination treatment (Fig. 3a). Four paired biopsies were further evaluated for gene expression changes and there was minimal consistency in their gene expression profiles after combination treatment. Additional analyses focusing on a subset of genes associated with ICOS biology and/or T-cell activation showed no overlap in genes exhibiting a ≥ twofold increase in gene expression among all four patients (Fig. 3b, c).

Among the 26 patients enrolled, a PR was achieved in one patient with study treatment (24 mg feladilimab plus 225 mg tremelimumab). Flow cytometry analysis showed that this responder displayed the highest levels of CD3+ and CD8+ T-cells relative to other patients enrolled in the same dose-level cohort (Fig. 4, red line). Further, this patient had a marked increase in the percentage of proliferating (Ki67+ /CD4+) and activated (HLA-DR+ CD8+) T-cells at Week 2 (one week post first dose).