Each year, it is estimated that over 300,000 cardiac valve prostheses are implanted worldwide. When mitral valve repair is not feasible, replacement with a prosthesis becomes the offered solution. Mitral prostheses provide excellent long-term outcomes, conferring a survival advantage, albeit not reaching the effectiveness and durability seen in successful mitral valve repair. For patients under 70 years, mechanical prostheses are often preferred; however, bioprostheses also yield favorable outcomes for selected cases. The key drawback of mechanical prostheses is the requirement for lifelong oral anticoagulation (OAC), which can add morbidity that bioprosthetic alternatives might mitigate. Furthermore, even in patients needing anticoagulation for other indications, the presence of a mechanical prosthesis mandates the use of coumarin-based agents with a target INR of 3 (previous range 2.5-3.5, now less commonly used).

The potential role of direct-acting anticoagulants as substitutes for coumarins in valve prosthesis patients has been evaluated in the RE-ALIGN (Dabigatran vs. Warfarin in Patients with Mechanical Heart Valves) and RIWA (Rivaroxaban vs. Warfarin in Patients with Metallic Prostheses) trials. In RE-ALIGN, a higher bleeding risk was identified, contrasting with RIWA’s findings. However, other published cohorts reported serious valve thrombosis complications with Rivaroxaban. This evidence has led to the formal contraindication of direct-acting anticoagulants in patients with mechanical valve prostheses, with coumarins remaining the first-line therapy.

This study aims to provide new evidence by examining a less aggressive anticoagulation regimen in mitral prostheses, akin to the PROACT (Prospective Randomized On-X Anticoagulation Trial) study on the On-X prosthesis in the aortic position. From 2006 to 2020, a total of 401 patients undergoing mitral valve replacement with an On-X mechanical prosthesis were prospectively randomized across 44 centers and with the participation of 100 implanting surgeons. After three months post-surgery, patients were divided into two anticoagulation groups: 200 patients adhered to a standard anticoagulation regimen (target INR of 3, range 2.5-3.5) following clinical guidelines, while 201 patients followed a lower anticoagulation target (INR 1.5-2). All patients also received concurrent aspirin therapy. Patients younger than 18 years, those with double-valve prostheses, active endocarditis, a history of thromboembolic events within the past year, terminal illness, or urgent intervention were excluded. Candidates could be randomized between 3-12 months post-valve replacement, and INR monitoring was conducted by the patients themselves. The primary outcome included thromboembolism, valve thrombosis, and major and minor bleeding events. Secondary outcomes were all-cause mortality and adverse prosthesis-related events such as endocarditis, hemolysis, hemolytic anemia, paravalvular leakage, and structural or non-structural dysfunction. A non-inferiority analysis was conducted following FDA performance criteria for cardiac valve prostheses, with primary combined event rates set at 6% patient-years in the low-range group and 7.3% patient-years in the standard group, with a 1.5% non-inferiority margin.

The study succeeded in randomizing 401 patients over 14 years, with an average follow-up of over 4 years and 1,600 patient-years analyzed. The primary combined event occurred in 11.94% of the low-dose group and 12.01% of the standard-dose group. The absolute difference was -0.07% (95% confidence interval, -3.40% to 3.26%), with the upper limit exceeding the preset 1.5% non-inferiority margin. No differences were found in secondary outcomes, including mortality, thromboembolism, valve thrombosis, major bleeding, or event-free survival at 5 and 8 years. In a post-hoc analysis of patients with preoperative atrial fibrillation, there were also no differences in these variables.

The authors concluded that a low OAC regimen for managing On-X mitral prostheses cannot be recommended, as non-inferiority in the primary combined event was not achieved.

COMMENTARY:

Establishing OAC ranges for mechanical valve prostheses has always been a trial-and-error process. In the early 1990s, an INR range of 3-4.5 was recommended for any mechanical cardiac prosthesis. However, various teams reduced the INR range due to frequent bleeding episodes. The study by Gohlke-Barwolf et al. first established distinctions in OAC ranges between aortic and mitral prostheses. The multicenter AREVA (Anticoagulation in Patients with Mechanical Prosthetic Heart Valves) study demonstrated that an INR range of 2-3 was comparable to 3-4.5, with fewer bleeding events. The LOWERING-IT (Lowering the Intensity of Oral Anticoagulation Therapy in Patients with Bileaflet Mechanical Aortic Valve Replacement) study further reduced the range, showing the non-inferiority of an INR range of 1.5-2.5 for low-risk patients. Finally, the German Experience with Low Intensity Anticoagulation trial established an INR range for mechanical mitral prostheses of 2-3.5, and for combined multiple mechanical prostheses of 2.5-4 (provided a St. Jude prosthesis was used).

The limitations of this study include significant overlap in the OAC ranges within each group, with the low OAC group frequently surpassing the established upper limit, likely due to OAC regimen adjustments. This affects the study’s internal validity, complicating the interpretation of comparisons. The majority of patients managed their OAC at home, a scenario uncommon in our setting, which also impacts external validity. Additionally, only On-X mechanical mitral valves were studied, meaning these findings cannot be applied to other mechanical mitral prostheses. Finally, all patients received antiplatelet therapy; thus, these results do not apply to patients managed exclusively with OAC.

In conclusion, the negative findings of Chu et al. add to the evidence that for mechanical mitral prostheses, we cannot justify lower INR ranges than 3 (range 2.5-3.5), not even for the new-generation On-X mechanical prostheses. These results do not align with the PROACT study findings and, therefore, do not support the alternative anticoagulation management approach for these prostheses as in the aortic position.

REFERENCE:

Chu MWA, Ruel M, Graeve A, Gerdisch MW, Damiano RJ Jr, Smith RL 2nd, et al. Low-Dose vs Standard Warfarin After Mechanical Mitral Valve Replacement: A Randomized Trial. Ann Thorac Surg. 2023 Apr;115(4):929-938. doi: 10.1016/j.athoracsur.2022.12.031.