December 2019

Synopsis: The EXCEL trial was an international, multicenter, randomized trial comparing PCI (everolimus-eluting stent (XIENCE, Abbott Vascular) to CABG in patients with significant left main stenosis (70% or more by visual estimate, or 50-69% if hemodynamically significant). The primary composite endpoint was death, stroke, or myocardial infarction. The PCI group received dual antiplatelet therapy prior to treatment and for at least 1 year following stent placement. Patients undergoing CABG were given aspirin before and this was continued after surgery. A total of 1905 patients - 948 PCI and 957 CABG – were included and 22% of patients in the PCI group and 19.2% of those in the CABG group experienced the primary composite outcome at 5 years (p=0.13; 95% CI -0.9 to 6.5). The incidence of all-cause death was 13.0% vs 9.9% in the PCI vs CABG group (OR 1.38, 95% CI 1.03-1.85). The incidence of stroke was 3.7 vs 2.9% in CABG vs PCI patients (OR 0.78, 0.46-1.31); myocardial infarction incidence was 10.6% after PCI vs 9.1 after CABG (OR 1.14, 0.84-1.55). Ischemia driven revascularization was 16.9% vs 10.0% (OR 1.84, 1.39-2.44) after PCI vs CABG. A more granular look at the distribution of adverse events over time reveals changing risk profiles over time between the two treatment groups. When the incidence of the primary composite outcome was examined at three time periods -- 0 to 30 days, 30 days to 1 year, and 1 to 5 years after treatment – the PCI group had fewer early adverse outcomes. Adverse outcomes equalized in the intermediate time period and favored CABG in the late period. Ultimately, the cumulative mean time free from adverse events was similar between PCI and CABG. 

Synopsis: The DeCOG-SLT trial is a German multicenter phase III trial in which patients with melanoma of the trunk or extremities with a positive sentinel lymph node were randomly assigned to undergo complete lymph node dissection (CLND) or observation. Previous 3-year analysis at a median of 34 months demonstrated no difference in distant metastasis-free survival (DMFS), overall survival (OS), or recurrence free survival (RFS) between arms. This study reports the final results of DeCOG-SLT at a median follow up of 72 months. Final intention-to-treat analysis included 240 patients randomly assigned to CLND and 233 to observation. There were no differences in the 5-year DMFS, OS, and RFS rates for CLND vs. observation at 64.9% vs. 67.6% (HR 1.08; 90% CI, 0.83-1.39, P=0.65), 72.3% vs. 71.4% (HR 0.99; 90% CI 0.74-1.31, P=0.94), and 59.9% vs. 60.9% (HR 1.01, 90% CI, 0.80-1.28, P=0.94), respectively. The rate of regional lymph node recurrence was 10.8% (26/240) in the CLND arm vs. 16.3% (38/233) in the observation arm, which was not statistically significant. Adverse event rates in the CLND group were 24% for all adverse events, 13% for grade 3/4 toxicity, and 8.3% for lymphedema. Tumor load in the sentinel lymph node (≤1 mm vs. >1 mm) and tumor thickness (≤2 mm vs. >2 mm) were independently associated with DFS, OS, and RFS. The findings of no differences in DMFS or OS in DeCOG-SLT corroborate those of MSLT-II,1 the larger international multicenter phase III trial of CLND vs. observation in patients with melanoma with positive sentinel lymph nodes. In contrast to DeCOG-SLT, at a median follow-up of 43 months, MSLT-II demonstrated improved disease-free survival (DFS) (68±1.7% vs. 63±1.7%, P=0.05) and a 69% decrease in regional node recurrence with CLND compared to observation. The major limitation of DeCOG-SLT is smaller sample size compared to MSLT-II, which may account for the lack of a difference in regional node recurrence and DFS between arms. Together DeCOG-SLT and MSLT-II demonstrate that there is no OS or DMFS benefit from CLND for positive sentinel lymph nodes in melanoma. There may be benefits in staging and regional disease control with CLND, which must be weighed against the risks of surgery.

¹Faries MB, et al. Completion dissection or observation for sentinel-node metastasis in melanoma. NEJM 2017;376(23):2211-22.  

Synopsis: The perioperative management of patients with atrial fibrillation (AF) taking direct oral anticoagulants (DOAC) is a common clinical scenario, but for which best practices are unclear. The Perioperative Anticoagulation Use for Surgery Evaluation (PAUSE) study is a multi-center prospective cohort study of a standardized perioperative management approach of 3,007 adult patients with AF using apixaban, dabigatran or rivaroxaban undergoing elective surgery or procedure.

Preoperativley, the DOAC regimens were stopped 1 day before a low bleeding risk procedure and 2 days before a high bleeding risk procedure. Patients using dabigatran with impaired renal clearance had longer interruption intervals. Resumption of DOAC regimens occurred on post-operative day 1 (≈24 hours) for low bleeding risk procedures and day 2-3 (≈48-72 hours) for high bleeding risk procedures (see Figure). 

Overall, the 30-day post-operative rate of major bleeding was 1.35% for apixaban (n=1,257 patients), 0.90% for dabigatran (n=668 patients) and 1.85% for rivaroxaban (n=1,082 patients). Among the 1,007 patients (33.5%) with a high bleeding-risk procedure, the rates of major bleeding were 2.96% for apixaban, 0.88% for dabigatran, and 2.95% for rivaroxaban. The rates of arterial thromboembolism were low across all three cohorts (apixaban 0.16%, dabigatran 0.60%, rivaroxaban 0.37%. 

Clinical Takeaway:
In patients with AF on DOAC therapy, a simple, standardized approach to perioperative anticoagulation without heparin bridging or coagulation function testing is feasible and associated with low rates of major bleeding and arterial thromboembolism.

1. Hornor MA, Duane TM, Ehlers AP, Jensen EH, Brown PS, Jr., Pohl D, et al. American College of Surgeons' Guidelines for the Perioperative Management of Antithrombotic Medication. Journal of the American College of Surgeons. 2018;227(5):521-36.e1.

This article provides additional guidance for perioperative management of antithrombotic medications, including warfarin and anti-platelet agents.