Muhammad Rizwan Husain, MD

Robinson AA, Trankle CR, Eubanks G, Schumann C, Thompson P, Wallace RL, Gottiparthi S, Ruth B, Kramer CM, Salerno M, et al. Off-label Use of Direct Oral Anticoagulants Compared With Warfarin for Left Ventricular Thrombi. JAMA Cardiol. 2020;5:685-692.

In this entry, I discuss the RED-VELVT observational study by Austin A. Robinson and colleagues about treating left ventricular (LV) thrombi with direct oral anticoagulants (DOACs) compared to warfarin and looking at rates of embolic events between the two treatment modalities.

Anecdotal data from the nineties, as well as the early 2000s, demonstrated a lower risk of systemic embolization with the use of warfarin or low molecular weight heparin (LMWH) compared to antiplatelets for LV thrombi. However, with the arrival of DOACs, anticoagulation therapy has been revolutionized due to ease of administration, no requirement for daily INR checks, no need for dietary alterations, and lower risk of bleeding events, to name a few. The benefit of DOACs for treatment of left atrial (LA) thrombus was noted in the X-TRA trial, and further benefit of DOACs was also seen in patients requiring anticoagulation for cardioversion (ENSURE-AF, EMANATE). Noting the benefits of DOACs in treating LA thrombus, many physicians have extrapolated the use of DOACs to treat LV thrombus. However, there has been no study to date to evaluate if DOACs fare better, worse, or the same compared to warfarin to reduce embolic events in patients with LV thrombi, and the authors in this study have attempted to answer this question.

This was a retrospective study conducted at three academic centers between October 2013 and March 2019, in which the authors included all patients with echocardiographically diagnosed LV thrombi. The patients were stratified as to which type of anticoagulation they were treated with (DOAC or warfarin). The outcome studied was a composite of clinically documented stroke and systemic embolization (SSE). A presenting embolism had to occur within the first 30 days of the initial echocardiogram, and any SSE that occurred during the initial bridge with parental therapy to oral anticoagulation was excluded.  Various clinical and echocardiographic factors were collected.

There were a total of 514 patients with LV thrombi, of which 421 were treated with an oral anticoagulation. A total of 300 patients were treated with warfarin, and a total of 185 patients were treated with a DOAC (Apixaban 141; Rivaroxaban 46, Dabigatran 9). 64 patients had a switch between the groups (warfarin and DOAC), which resulted in a total of 236 patient treated exclusively with warfarin and 121 patients treated exclusively with a DOAC. No patients were treated with edoxaban. The median time to switch from a warfarin to DOAC was 277 days, and the median time to switch from a DOAC to warfarin was 43 days. Cost was the major reason (31.6%) to switch from a DOAC to warfarin, while convenience (19.2%), adherence (15.4%), and labile INR values (15.4%) were the major reasons to switch from warfarin to a DOAC.

The median follow-up was 351 days, during which 54 SSE (36 ischemic strokes and 18 systemic emboli) occurred. SSE occurred in 17 patients in the DOAC group, and in 14 patients in the warfarin group. The event rate was 0.065 per patient year of follow up.

In the univariate analysis, SSE was significantly associated with DOAC treatment versus warfarin (HR, 2.71; 95% CI, 1.31-5.57; P = .01). A history of prior SSE was also associated with SSE (HR, 2.13; 95% CI, 1.22-3.72; P = .01). In the multivariate analysis, again, DOAC use when compared to warfarin (HR of 2.64 95% CI, 1.28-5.43; P = .01), as well as a history of prior SSE (HR, 2.07; 95% CI, 1.17-3.66; P = .01), were significantly associated with SSE.

A total of 115 deaths also occurred during the follow-up period (14 in DOAC group vs 32 in warfarin group). In an unadjusted Cox analysis of a combined end point of death or SSE, DOAC vs warfarin was associated with death or SSE (HR, 1.55; 95% CI, 1.05-2.30; P = .03).

Even though there was a clear significant difference demonstrating DOACs being more likely to lead to SSE compared to warfarin, the authors conducted various other analyses as well.

First, due to the potential risk of early embolism in an acute LV thrombus, all SSE in the first 10 days were excluded. Again, DOACs were associated with SSE compared to warfarin (HR, 2.67; 95% CI, 1.25-5.70; P = .01).

Second, due to a delay in the divergence of the survival curves, the authors looked at events at 3 months, 6 months and to the end of the follow-up period. At the 3-month mark, DOAC vs warfarin showed an HR of 2.33 (95% CI, 0.63-8.74; P = .2), which did not reach statistical significance, and there were no events in the 3- to 6-month period to allow for an analysis. However, the analysis from the 3-month mark to the end of the follow-up period again demonstrated DOACs causing SSE compared to warfarin (HR, 2.88; 95% CI, 1.22-6.80; P = .02).

Third, the authors conduced a sensitivity analysis including only patients who completed a 90-day course of anticoagulation and then stopped it. This again showed DOACs to result in SSE compared to warfarin (HR, 2.69; 95% CI, 1.37-5.26; P = .004).

Fourth, the authors included all events (15 SSE; 6 warfarin, 9 DOAC) that occurred only during the first year. This, too, showed DOAC causing SSE compared to warfarin (HR, 3.10; 95% CI, 1.10-8.73; P = .03).

After multiple comparisons, the analysis strongly suggests that warfarin should be the drug of choice thus far to treat LV thrombi. Considering the success of DOAC in preventing stroke and systemic embolism in the treatment of atrial fibrillation(AF), there are a few potential explanations as to why this benefit might not extend to treating LV thrombi. One, DOACs are used to prevent thrombus formation in the left atrial appendage compared to an already formed LV thrombi. Two, the thrombi formed in AF are due to stasis, whereas LV thrombi are a mix of stasis and LV endocardial changes. Third, the study of a DOAC showing benefit to treat left atrial appendage clot had no direct comparison to warfarin, so no head-to-head comparison is available. Lastly, there are many late embolic events that are not related to the left ventricle itself, such as leaflet thromboses, septic vegetation’s and vascular calcifications.

There are also a few important limitations to note besides the retrospective nature of the study: 1) the authors did not clarify what constituted a systemic embolism besides stroke; 2) there was no central radiological evaluation of the echocardiograms; 3) bleeding events were not considered a primary end point, especially since DOACs have shown to cause less bleeding risk than warfarin; and 4) patient compliance with anticoagulation and time spent in therapeutic INR range was not studied. Overall, the study results should further push physicians to the continued use of warfarin and not DOACs to treat LV thrombi until further randomized trial results become available.