Evidence-based medicine for thromboprophylaxis in hospitalized patients with cancer: why aren't we there yet?

Abstract

Venous thromboembolism is a costly disease. In those afflicted, the medical, economical, and emotional impact can be overwhelming, especially in patients living with cancer. Prevention is often an effective strategy to reduce disease burden, yet historical data have shown that rates of thromboprophylaxis in hospitalized patients with cancer are dismally low. Given the vast efforts and sweeping programs directed at preventing thrombosis in hospitalized patients in recent years, including the Surgeon General’s Call to Action in 2008 and the introduction of pay-for-performance measures in medical institutions, have oncologists and other clinicians providing care to patients with cancer embraced the practice of providing thromboprophylaxis to hospitalized oncology patients? In the article by Zwicker et al, we gain the first sneak peek at the impact of these endeavors. Using a prospective, cross-sectional design, the investigators collected data from 775 patients with cancer admitted to five academic medical centers between January to June 2013 to determine the current prescription rates of pharmacologic prophylaxis and identify factors that influenced prophylaxis prescription. Overall, 51% of patients received pharmacologic prophylaxis. But if the patients with relative contraindications to anticoagulation (nearly one-third of the total sample) were excluded, 74% of the eligible patients were prescribed pharmacologic prophylaxis. This is higher than the rates reported in previous studies that extracted data from administrative databases. Multivariable analysis identified a history of prior venous thromboembolism as the strongest predictor of prophylaxis prescription and found that patients admitted with hematological malignancies or for cancer therapy were significantly less likely to receive prophylaxis. Otherwise, patient selection for prophylaxis appeared somewhat haphazard, as only 79% of those classified as having a high risk of thrombosis by the Padua Score were prescribed prophylaxis, while 63% of those considered as low risk also received prophylaxis. The authors concluded that although pharmacologic prophylaxis is frequently prescribed in hospitalized patients with cancer, this is done without regard to the presence or absence of other risk factors for venous thromboembolism. So, what does this mean to patients, clinicians, and the healthcare burden? Most importantly, does a higher prescription rate of pharmacologic prophylaxis translate to improvement in the quality of patient care in the oncology population? To answer these questions, let us review the evidence that support prophylaxis in hospitalized patients with cancer and identify the remaining gaps that must be bridged to bring about improvement in patient outcomes. To date, the only evidence available on the efficacy and safety of anticoagulant prophylaxis in hospitalized patients with cancer comes from post hoc, subgroup analyses of trials that included a small number of selected patients with cancer. A meta-analysis of the three randomized controlled trials that compared low molecular weight heparin or fondaparinux with placebo found that among the 307 patients with cancer enrolled, no statistical reduction in the overall incidence of venous thromboembolism was demonstrated with anticoagulant prophylaxis. It is difficult to know if this finding is due to a type II error (lack of statistical power), a skewed selection of low-risk patients with cancer or a true lack of efficacy using standard doses of pharmacologic prophylaxis in this hypercoagulable population. Another major barrier for clinicians in providing appropriate prophylaxis is the identification of patients who would benefit from prophylaxis. Thromboprophylaxis based on risk stratification is strongly advocated by the most recent clinical practice guidelines from both the American College of Chest Physicians (ACCP) and the American Society of Clinical Oncology (ASCO). Although individual risk factors associated with thrombosis are well established, validated risk assessment tools for estimating the overall risk of thrombosis in hospitalized patients with cancer are not yet available. The Padua Prediction Score, used in the study by Zwicker et al and recommended by the ACCP, was empirically derived and appears promising in identifying a low-risk group of patients in whom thromboprophylaxis is likely not warranted. However, the score has not undergone rigorous external validation and a recent study found that it was not predictive of in-hospital thrombosis risk in patents with sepsis. The utility of the score is also questionable in patients with cancer as it does not include known risk factors unique to patients with cancer, such as the presence metastatic disease or the use of chemotherapy. In the current ASCO guideline, only those patients who have active malignancy and are admitted with an acute medical illness or reduced mobility are recommended pharmacologic prophylaxis. Given the heterogeneity of the cancer population admitted to hospital—from those battling with sepsis to those facing the end of JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 32 NUMBER 17 JUNE 1