COAPT Risk Score Useful in Predicting Outcomes in Functional MR

Researchers have developed a new score to help risk stratify heart failure (HF) patients with moderate-to-severe functional mitral regurgitation (MR), one they hope can provide an accurate prognosis in this group at high risk of adverse outcomes.

The risk score, which was derived from the COAPT trial and includes clinical, echocardiographic, and treatment-related variables, had a modest discriminative ability (C-statistic = 0.74) among the derivation cohort, but did perform better than some previous scores, including those used for patients undergoing surgery and those with HF who are medically managed.

In the model, which was developed by lead author Neeraj Shah, MD (East Carolina University, Greenville, NC), and published October 1, 2022, in JACC: Cardiovascular Interventions, investigators also showed that the benefit of transcatheter edge-to-edge repair (TEER) with MitraClip (Abbott) in reducing the 2-year risk of death or HF hospitalization was seen in all patients, though there was a greater absolute benefit in those at the highest risk based on the COAPT score. 

In an editorial accompanying the paper, Michele Pighi, MD (University of Verona, Italy), and Marianna Adamo, MD (University of Brescia, Italy), point out that most of the variables included in the COAPT risk score have been previously shown to be predictive of a poor prognosis in patients undergoing TEER for MR.

“However, this simple and straightforward prediction model may have an extremely high practical utility even though a comprehensive clinical evaluation is always needed,” they write.

Andrew Goldsweig, MD (University of Nebraska Medical Center, Omaha), who wasn’t involved in the study, said risk scores in the setting of TEER for functional MR aren’t used much, but that they should be given that physicians use the SYNTAX score for guidance in coronary artery disease and the STS/ACC TVT-TAVR risk score in patients with symptomatic severe aortic stenosis. The newly developed COAPT score, he agreed, is simple enough to be routinely applied in clinical practice.

There’s definitely a need for something like this. Andrew Goldsweig

“In mitral disease, scoring-system utilization is pretty uncommon,” he said. “The other scores available are older and have been applied in research studies, but this COAPT risk score is novel. It’s a derivation paper, without validation, and that’s the key next step.”

While TEER reduces MR, the controversy surrounding its use revolves around which patients will have an improvement in clinical outcomes, said Goldsweig. COAPT was a positive study, but the MITRA-FR study failed to show any difference in the risk of death or unplanned HF hospitalizations at 1 year between guideline-directed medical therapy (GDMT) and TEER with MitraClip. 

“The pundits always point to the differences between the COAPT and MITRA-FR study populations, arguing that patients that look more like the COAPT population are more likely to benefit from mitral TEER,” said Goldsweig. “The present study derived a score that essentially quantifies how closely a patient resembles the COAPT patients who benefited from mitral TEER.” 

Scored From -3 to +15

In creating the risk score, Cox proportional hazards models were developed to identify predictors of the primary endpoint, with all predictors of the primary outcome having P values < 0.20 included as part of the score.

The researchers identified eight variables independently associated with mortality in the multivariable model and assigned each different points because the hazard ratios differed. For example, chronic kidney disease (CKD) stage 3 was assigned 1 point but CKD stage 4 or greater was assigned 3 points. The other clinical variables included NYHA functional class III/IVa (1 point), chronic obstructive pulmonary disease (1 point), and history of atrial fibrillation/flutter (1 point). The echocardiographic variables included right ventricular systolic pressure > 45 mm Hg (3 points), LVEF 25% to 35% (1 point), LVEF < 25% (2 points), left ventricular end-systolic diameter > 5.5 cm (2 points), and tricuspid regurgitation grade ≥2 (2 points). If patients were treated with TEER, they were assigned -3 points. In total, the score could range from -3 to 15 points.  

The calibration plot of the risk score model indicated “excellent calibration,” say researchers. Sensitivity analyses using only variables with P values < 0.05 were tested in the model but did not improve the C-statistic. 

When researchers looked at the 2-year event rates of the primary endpoint, there was a steady increase in both the MitraClip- and GDMT-treated patients across increasing COAPT risk-score quartiles. Similarly, higher COAPT risk scores were associated with a higher risk of death at 2 years in both the MitraClip and GDMT study arms. After removing TEER from the score, and instead relying on just clinical and echocardiographic variables, patients with higher baseline risk had a greater absolute improvement in event-free survival with MitraClip up to a risk score difference of 8 points.

Risk calculators have been developed for patients with degenerative MR, such as the MIDA mortality risk score, but given the differing pathophysiology, clinical presentation, and management of functional MR, a dedicated tool is necessary, according to the researchers. The COAPT risk score needs to be externally validated to determine its reproducibility and should also be validated in patients who don’t meet the COAPT inclusion criteria, such as the MITRA-FR study population, according to the researchers.   

Goldsweig, agreed the best validation cohort would be the MITRA-FR trial population. “In particular, it would be absolutely transformational if MITRA-FR patients with COAPT risk scores £ 4 actually experienced a reduction in death or heart failure hospitalizations,” he said. “Subsequently, the score should be tested for [transcatheter mitral valve replacement patients] as well.”  

The COAPT score, he said, could be calculated prior to the intervention and be part of the heart-team discussion. “There’s definitely a need for something like this,” said Goldsweig, adding that the creation of a website is also an important next step. “If there was a really simple website for people to calculate [the risk score], everybody would do it.”

Note: Prior to the publication of this story, TCTMD learned that Dr. Michele Pighi had passed away in a motorcycle accident, just days before the publication of his editorial. We want to express our sincere condolences to his family, friends, and colleagues.