Checkpoint inhibitors and thrombosis: what’s up?

In this issue of Blood, Moik et al present data from a single-center retrospective analysis that explored both the incidence and clinical impact of thrombotic events during immune checkpoint inhibitor (ICI) therapy.¹ 

A diagnosis of cancer has a major and immediate impact on the lives of patients, their families, and their friends. When the diagnosis is advanced or metastatic cancer, patients and health care providers have to face difficult treatment decisions to balance the potential (but not guaranteed) treatment benefits against the risks. The treatment regimen and intensity usually determine the side effects and the impact on quality of life, which is particularly important, because life expectancy is usually limited in this situation. With the development of modern anticancer therapies, survival of several cancer entities has dramatically improved over recent years.²  Immune checkpoint inhibitors belong to this category of modern anticancer treatments: they impair the immune-escape mechanisms of cancer cells and allow for the immune system of the patient to help in fighting the spread of the disease.

Venous and arterial thromboembolism (VTE and ATE) are common complications of cancer³  and represent, together with infections, the leading noncancer causes of death.⁴  Thromboembolism contributes significantly to the excess mortality of patients with cancer.⁵  As a consequence, extensive research has been performed to predict and prevent VTE in high-risk populations and to improve treatment in patients with cancer.⁶  Modern anticancer strategies should pay attention to the risk of thromboembolism. Surprisingly, the risk of thromboembolism from ICI therapy has not been systematically studied, which is especially concerning, given that the immune system is interlinked with many other physiological processes, including the hemostatic system. Currently available evidence on thromboembolic risks of ICI treatment is limited to small cohort studies and case reports, sometimes describing extreme hemostatic responses to ICI treatments, such as disseminated intravascular coagulation or hyperfibrinolysis.^7-9 

Moik et al report a cohort of 672 patients treated with nivolumab, pembrolizumab, ipilimumab, atezolizumab, or avelumab with a median follow-up of 8 months. All patients were treated at a highly specialized cancer clinic, and nearly 15% were included in treatment studies, thus reducing the risk of unreported outcomes, despite the retrospective design of the analysis.

Baseline risks of thromboembolism and death were considerable in this cohort: median age 64 years, 13% with a history of VTE, and 10% with a history of ATE. Because of these risk factors, 16% of the patients received prolonged anticoagulant therapy, 20% were given antiplatelet therapy at baseline, 85% were in treatment for metastatic disease, and 15% were already in second- or third-line ICI therapy. Given this baseline, it may not come as a surprise that the cumulative incidence of VTE during ICI therapy was 13%, with an incidence of ATE of nearly 2%. Interestingly, rates of VTE were comparable among subgroups of tumor types and checkpoint inhibitors and across Khorana-score categories, indicating that this could be a class effect that was directly related to ICI therapy. Of 15 variables tested as potential VTE risk factors, only a history of VTE was found to be a significant independent risk factor (subdistribution hazard ratio, 3.7). In contrast, expected items such as age, cancer stage, body mass index, and expression of programmed death ligand 1 (PD-L1) on tumor cells were not associated with an increased risk of VTE. Baseline anticoagulation and antiplatelet therapy were not protective.

The authors also reported on the prognostic impact of VTE and ATE. The occurrence of VTE, but not of ATE, was associated with significantly shorter overall and progression-free survival. Furthermore, in patients developing VTE during ICI treatment, recurrent VTE and major bleeding were observed in 8.5% and 4.3%, respectively. VTE did not cause discontinuation of ICI therapy but led to a delay of the next treatment cycle of up to 3 weeks. In contrast, ATE led to ICI discontinuation in 1 patient and caused considerable treatment delays of up to 5 months in another 3 patients. The authors concluded that patients with cancer receiving ICI therapy are at high risk of thromboembolism, which would have a significant impact on their clinical course and prognosis.

First of all, these data clearly indicate that the phase 2 and 3 ICI trials should be reevaluated for safety, specifically concentrating on the risk of ATE and VTE. Moik et al and other groups correctly criticize the landmark clinical trial that did not include any information on the incidence of thromboembolism.⁹  This deficiency clearly indicates an important knowledge gap. Observational studies can be regarded only as hypothesis generating, and dedicated risk assessments of modern treatments should be based on randomized controlled comparisons.

Second, issues of risk prediction and primary prevention should be evaluated. In the future, VTE risk prediction models should incorporate anticancer drugs with increased thromboembolic risks, thereby identifying more patients as candidates for thromboprophylaxis. This precaution could become especially important because better prevention of VTE reduces the need for anticoagulant therapy and could reduce the resultant bleeding risks.

Third, more research should explore the impact of adverse effects of ICI treatment on the patients’ quality of life, as the occurrence of thromboembolism has been shown to add significant burden to patients with cancer.¹⁰ 

Finally, patients receiving ICI should be informed about signs and symptoms of ATE/VTE and instructed to seek immediate medical attention, to allow for prompt diagnostic and therapeutic workup for suspected thromboembolism.

All evidence taken together, Moik et al spotlight an important and underestimated safety issue of ICIs. Readers and prescribers are encouraged not only to study this article carefully but to include the question of thromboembolism in their daily treatment considerations and to invest further research into this important topic.