Phase I study of single-agent anti-programmed death-1 (MDX-1106) in refractory solid tumors: safety, clinical activity, pharmacodynamics, and immunologic correlates

Abstract

Purpose: Programmed death-1 (PD-1), an inhibitory receptor expressed on activated T cells, may suppress antitumor immunity. This phase I study sought to determine the safety and tolerability of anti-PD-1 blockade in patients with treatment-refractory solid tumors and to preliminarily assess antitumor activity, pharmacodynamics, and immunologic correlates.

Patients and methods: Thirty-nine patients with advanced metastatic melanoma, colorectal cancer (CRC), castrate-resistant prostate cancer, non-small-cell lung cancer (NSCLC), or renal cell carcinoma (RCC) received a single intravenous infusion of anti-PD-1 (MDX-1106) in dose-escalating six-patient cohorts at 0.3, 1, 3, or 10 mg/kg, followed by a 15-patient expansion cohort at 10 mg/kg. Patients with evidence of clinical benefit at 3 months were eligible for repeated therapy.

Results: Anti-PD-1 was well tolerated: one serious adverse event, inflammatory colitis, was observed in a patient with melanoma who received five doses at 1 mg/kg. One durable complete response (CRC) and two partial responses (PRs; melanoma, RCC) were seen. Two additional patients (melanoma, NSCLC) had significant lesional tumor regressions not meeting PR criteria. The serum half-life of anti-PD-1 was 12 to 20 days. However, pharmacodynamics indicated a sustained mean occupancy of > 70% of PD-1 molecules on circulating T cells > or = 2 months following infusion, regardless of dose. In nine patients examined, tumor cell surface B7-H1 expression appeared to correlate with the likelihood of response to treatment.

Conclusion: Blocking the PD-1 immune checkpoint with intermittent antibody dosing is well tolerated and associated with evidence of antitumor activity. Exploration of alternative dosing regimens and combinatorial therapies with vaccines, targeted therapies, and/or other checkpoint inhibitors is warranted.

Fig 1.

Objective tumor responses in patients with metastatic renal cell carcinoma (RCC) and melanoma after repeated dosing with anti–programmed death-1 monoclonal antibody (MDX-1106) at 10 mg/kg. (A) Patient 4033 with RCC experienced a partial response (PR) after receiving three doses of MDX-1106. Regression of metastases in mediastinal lymph nodes and bone (scapula) demonstrated on contrast-enhanced computed tomography scans are representative of lesions at other sites including lung, muscle, pancreas, and pericolic lymph node. Date of first treatment was January 29, 2008. (B) Patient 3019 experienced a PR after receiving 11 doses of MDX-1106. Serial core-needle biopsies of a regressing axillary lymph node metastasis were stained with anti-CD8, revealing a moderate post-treatment infiltrate. Infiltration of CD4⁺ cells was not observed (not shown). 20× objective. Rx, treatment; wk, week.

Fig 2.

Membranous pattern of B7-H1 expression demonstrated on (A) renal cell carcinoma cells in a tumor thrombus from patient 4033, and (B) melanoma cells in an axillary lymph node metastasis from patient 3019. Both patients experienced partial responses after anti–programmed death-1 monoclonal antibody (MDX-1106) therapy. 40× objective.

Fig 3.

Effects of a single dose of anti–programmed death-1 monoclonal antibody (MDX-1106; 10 mg/kg) on circulating lymphocyte numbers. (A) Twenty-four hours postdose, a decline in total lymphocyte as well as CD3, CD4, and CD8 numbers was observed (two-sided P = .004, .002, < .001, and .01 respectively; Wilcoxon signed rank test). These parameters followed similar trends, rebounding from days 2 through 29 (two-sided P < .001; two-sided P = .01 for CD8), and declining again from days 29 through 85 (two-sided P < .001; mixed model test for trend with knots [changes in slopes] and repeated measures). Means ± standard error of mean are shown; numbers of patients studied at each time point are indicated in parentheses. (B) Paired analysis of total lymphocyte numbers in 16 patients, comparing immediate pretreatment samples (day 1) with 24-hour post-treatment samples (day 2). A significant decline at day 2 was observed (two-sided P = .004; Wilcoxon signed rank test). Dotted line indicates the lower limit of normal lymphocyte counts.

Fig 4.

Pharmacodynamics of anti–programmed death-1 (PD-1) monoclonal antibody (MDX-1106). (A) PD-1 occupancy on circulating CD3⁺ T cells after one infusion of MDX-1106 is shown for single patients (Pts.) each receiving 0.3, 1, or 3 mg/kg, and for 10 patients receiving 10 mg/kg (mean ± standard error of mean; solid squares). Serum concentrations of MDX-1106 at the same time points are indicated (open diamonds). (B) Long-term PD-1 occupancy analysis in patients receiving one (top panel) or multiple doses (middle and bottom panels) of MDX-1106 at 10 mg/kg. All patients received infusions at day 1; additional infusions are indicated by arrows. Results in (B) middle and bottom panels are representative of five patients receiving multiple doses.

Fig A1.

Binding specificity of anti–programmed death-1 (PD-1) monoclonal antibody (MDX-1106). (A) MDX-1106 (20 μg/mL) was tested for binding to various T cell immunogloblulin (Ig) family members by standard enzyme-linked immunosorbent assay using plate-coated Fc-fusion proteins (R&D Systems, Minneapolis, MN). Goat antihuman immunoglobulin G (IgG; kappa chain specific) polyclonal antibody conjugated with horseradish peroxidase (Jackson ImmunoResearch, West Grove, PA) was used as secondary antibody (Ab). (B) MDX-1106 inhibits binding of both B7-H1 and B7-DC to PD-1. Transfected Chinese hamster ovary cells expressing PD-1 molecules were incubated with serial dilutions of MDX-1106 or an isotype-matched control antibody, followed by addition of fluorescein isothyocyanate–conjugated recombinant human B7-H1-Fc or B7-DC-Fc fusion proteins and flow cytometric analysis.

Fig A2.

Inflammatory colitis in patient 1008 with metastatic ocular melanoma, after receiving five doses of anti–programmed death-1 monoclonal antibody (MDX-1106) at 1 mg/kg. Routine staining and immunohistochemistry of colonoscopic biopsy specimens demonstrated mucosal ulceration and dense interstitial CD4⁺ and CD8⁺ lymphoid infiltrates, with intraepithelial lymphocyte infiltration and cryptitis. A normal colonic biopsy from another patient is shown for comparison (see inset). H/E, hematoxylin and eosin. 20× objective.

Fig A3.

Peripheral blood lymphocyte phenotypes following a single dose of anti–programmed death-1 monoclonal antibody (MDX-1106; 10 mg/kg) on day 1. (A) Percentages of circulating CD3, CD4, and CD8 cells declined steadily (two-sided P < .001), while CD19 and CD56 percentages rose reciprocally (two-sided P = .01; mixed model test for trends) over the 85-day observation period. (B) Expression of the activation markers CD25, CD45RO, and HLA-DR on circulating CD4⁺ and CD8⁺ T cells did not change significantly after treatment (Wilcoxon signed rank test and test for trends).