Biomarker Assessment of Venous Thromboembolism in Cancer Patients Receiving Checkpoint Blockade

Abstract

Introduction

Immuno-oncology shows significant promise in the treatment of malignancy but is associated with a variety of immune-related adverse events. We previously reported that venous thromboembolism (VTE) occurred in over 20% of patients treated with checkpoint blockade. We therefore have attempted to identify biomarkers/risk factors for VTE in this population that might provide a means to target prophylactic anticoagulant therapy to individuals at highest risk for this complication.

Methods

We used a database of 1,686 patients who received any of six approved immunotherapies (ipilimumab, nivolumab, pembrolizumab, atezolizumab, avelumab or durvalumab) between July 2015 and December 2017 to identify patients with VTE events including deep venous thrombosis (DVT), pulmonary embolism (PE) and visceral vein thrombosis (VVT). For biomarker/risk factors assessment, we defined a subcohort of cancer patients with (n=14) or without (n=10) VTE after receiving immunotherapy that had consented to blood and data collection in compliance with guidelines approved by the Cleveland Clinic institutional review board. Age matched healthy controls (n = 9) were also included in the analysis. Phenotyping of myeloid-derived suppressor cells (MDSC) was performed on fresh peripheral blood samples following isolation of peripheral blood mononuclear cells (PBMC) and flow cytometry. We measured 1) total MDSC (CD33+HLADR−), and within that population, 2) monocytic M-MDSC (CD14+CD15−), 3) polymorphonuclear/granulocytic PMN-MDSC (CD14−CD15+), 4) immature I-MDSC (CD14−CD15−), as well as surface expression of PD1 and PDL1. Analysis of soluble plasma biomarkers was performed in parallel using the validated Meso Scale Discovery V-PLEX Human Biomarker 54-Plex assay.

Results

The study group included 7 patients with bladder cancer, 7 with melanoma, 6 with renal cell carcinoma, and 4 with lung cancer. Total MDSC levels in all cancer patients receiving immunotherapy were significantly higher than those of age matched healthy controls (4.278, n=24 vs 1.543, n=9; p = 0.0198). Cancer patients receiving immunotherapy who developed VTE had significantly more PMN-MDSC than patients without VTE (12.59, n=14 vs 1.641, n=10; p = 0.0474). Cancer patients receiving immunotherapy who developed VTE also had elevated levels of pre-immunotherapy total MDSC compared to non-VTE counterparts, though these differences did not achieve statistical significance (4.948, n=14 vs 3.341; NS). Of 54 soluble plasma biomarkers tested, vascular cell adhesion molecule 1 (sVCAM-1) and interleukin 8 (IL-8) levels were also significantly higher in patients who developed VTE compared to those who did not (1.210 vs 0.8955 mg/ml, p= 0.0285; and 221.2 vs 115.0, p= 0.0174, respectively).

Conclusions

VTE is a common but underappreciated complication in cancer patients receiving immunotherapy. There are no formal risk tools or biomarkers predictive of VTE in cancer patients initiating immunotherapy. We have identified candidate cellular and plasma biomarkers that associate with thrombotic risk in this population and provide potential insight into its pathophysiology. Prospective and mechanistic studies are underway to validate and further expand these findings.