Venous thromboembolism in patients on check point inhibitors.

Abstract

e14582 Background: Checkpoint inhibition (CPI) therapy, including PD-1/PD-L1 and CTLA-4 inhibition strategies, has significantly improved the prognosis for patients with advanced malignancies. Anticancer treatments like chemotherapy and antiangiogenic agents are associated with higher rate of cancer associated thrombosis. However, the association of CPI with venous thromboembolism (VTE) is not well established. Methods: We conducted a retrospective cohort study of patients with advanced malignancy who were started on CPI between 2016 and 2019 at a tertiary care medical center in Southern California. We used descriptive statistics to describe the occurrence of VTE. Logistic regression was used to identify factors that had a significant impact on occurrence of VTE. Results: The cohort comprised 240 patients with median age of 60 years (range 27-96) at initiation of CPI and majority of the patients were males (57%). 45% were Hispanic, followed by Asian (19%), Caucasian (18%) and African American (8%). 80% had metastatic disease and the most common indication for use of CPI was lung cancer (20%), followed by gynecological malignancies (19%) including ovarian, endometrial and cervical cancer, melanoma (15%) and genitourinary malignancies (15%) including renal and urothelial cancer. The most used agent was nivolumab (68%) followed by pembrolizumab (25%), with most patients receiving single agent CPI (94%). Combination CPI and CPI with other agents including chemotherapy or targeted therapy was given to 3% and 4% respectively. The incidence of VTE was 10% (24/240) – pulmonary embolism (5), upper extremity deep vein thrombosis (5), lower extremity deep vein thrombosis (10) and visceral vein thrombosis (4). Patients were also stratified by Khorana score (KS) based on laboratory values and body mass indices at initiation of CPI (median score 1, range 0-4). VTEs occurred in patients on nivolumab in 8.02% of cases and in 18.2% of cases for pembrolizumab. VTEs occurred in 7.04% of patients with initial KS 0; 8.33%, KS 1; 17.5%, KS 2; and 15.4%, KS 3. Logistic regression analysis using a significance level of p < 0.05 showed no significant association between occurrence of VTE and age, sex, race, type of cancer or Khorana score. Conclusions: In our study, VTE risk in patients receiving CPI was similar to that with other systemic therapy. Further larger studies are needed to assess clinical significance of these findings, stratify for risk factors, and account for dose-dependence.