Wishing It True Doesn't Make It So: Commentary on an article by Gregory M. Mendez, MD, et al.: "Aspirin for Prophylaxis Against Venous Thromboembolism After Orthopaedic Oncologic Surgery".

Abstract

Commentary In this article, Mendez et al. offer their experience with use of aspirin for venous thromboembolism (VTE) prophylaxis specifically in orthopaedic oncology patients. They conclude that aspirin is a “suitable” chemoprophylactic agent. I wish this were so because aspirin is such a simple agent to prescribe and use. However, few topics in orthopaedics are as controversial as the selection of appropriate prophylaxis postoperatively. Prophylaxis for VTE in general has been the focus of numerous randomized prospective clinical trials with Level-I evidence, but even those have not always agreed. Hence, study panels of the American College of Chest Physicians (ACCP), American Academy of Orthopaedic Surgeons (AAOS), and American Society of Clinical Oncology (ASCO) continue to publish guidelines based on their interpretation of the latest literature. Despite these guidelines, there is considerable flexibility in a physician’s or surgeon’s choices for VTE prophylaxis. The controversy about VTE prophylaxis has been particularly contentious with respect to aspirin. As noted by Mendez et al., the ACCP guidelines recommended against its use in 2008 but then added aspirin as an acceptable agent in 2012. ASCO guidelines recommend against aspirin use and suggest only unfractionated heparin or low-molecular-weight heparin (LMWH) be used for patients with cancer who are undergoing major surgery. However, they do allow for aspirin use in patients on chemotherapy for myeloma. Citing ambiguity in the evidence, the AAOS guidelines allow use of aspirin as well as a pneumatic compressive device or any other chemotherapeutic agent for postoperative VTE prophylaxis following hip and knee arthroplasty. Mendez et al. are correct in asserting that universal surgical guidelines do little to direct the orthopaedic oncologist in the selection of VTE prophylaxis. However, does their article provide us the missing but desired direction? I believe that it is difficult to know how to use their findings. First, the number of patients in their study was relatively small (130 patients with a total of 142 procedures). In a recent systematic review and meta-analysis of computerized clinical decision support systems for VTE prophylaxis, 11 publications (including 9 prospective cohort trials) involving 156,366 subjects (an average 14,408 subjects per trial) were reviewed1. Even among 7 prior studies of VTE in orthopaedic oncology, the average number of patients was 348 (range, 94 to 986). Moreover, 5 of those studies were focused on incidence and risk factors for VTE. The only 2 treatment studies in orthopaedic oncology involved 348 subjects (treated with aspirin and a pneumatic compression device or with LMWH and a pneumatic compression device) and 986 subjects (treated with LMWH)2,3. Second, although this was a retrospective study, and hence intention-to-treat (ITT) analysis was not applicable, the authors clearly intended for all of their patients to receive aspirin; however, only 73% did so. Reasons for protocol deviation included an outpatient operative procedure not requiring chemoprophylaxis, patient refusal, aspirin allergy, VTE history, and a desire for a readily reversible option in case an early return to the operating room was necessary. Third, the orthopaedic oncology patient population studied by Mendez et al. was broad in its scope of inclusion criteria, which included metastatic carcinoma, myeloma, lymphoma, and sarcoma, further diluting the numbers for subgroup analysis. It also differs somewhat from prior reports in that more than one-half of the patients had metastatic carcinoma whereas some larger studies in orthopaedic oncology included predominantly or exclusively patients with sarcoma4,5. Thus, we must greatly temper our enthusiasm for the conclusion that aspirin is a suitable agent with the realization that the data are based on a small number of patients relative to the large body of literature on this topic. Also, the study was not a pure evaluation of patients treated only with aspirin, but rather a real-practice evaluation in which there were protocol deviations for various reasons; hence, its conclusions should not be extended to patients who had outpatient surgical procedures or a VTE history, for instance. Finally, although a broad range of orthopaedic oncology diagnoses were included, this actually makes it difficult to know how readily the results could be applied to patients with any one of those diagnoses. In conclusion, as much as I want aspirin to be a safe and efficacious VTE chemoprophylactic agent for all of my orthopaedic oncology patients, these results should not be interpreted as definitive support until additional studies are done.