New Score Helps Predict Mortality After TEER Mitral Repair

Patrice Wendling

February 10, 2022

A new eight-item score may predict mortality after transcatheter edge-to-edge (TEER) repair of the mitral valve, new research shows.

The investigators found the "MitraScore" was also predictive for heart failure hospitalization and correlated with the likelihood of clinical improvement.

"Maybe with this score you can provide honest information to the patient and also to have honest information for the doctors with regard to mortality," lead author Sergio Raposeiras-Roubín, MD, PhD, University Hospital Alvaro Cunqueiro, Vigo, Spain, told | Medscape Cardiology.

Current guidelines for valvular heart disease recommend TEER for prohibitive and high-risk patients with severe symptomatic primary mitral regurgitation and for select use in secondary mitral regurgitation. TEER with the MitraClip device has grown substantially, but has also been dogged by reports of residual mitral regurgitation and the need for repeat intervention.

A 2018 study showed that only about 54% of American patients were alive and well 1 year after TEER, whereas a 2021 study showed the surgical repair rate after failed TEER was just 4.8%. Yet specific tools for the risk-stratification of patients undergoing TEER are lacking.

"In clinical practice, we have a lot of patients who have not benefited with transcatheter mitral valve repair, so we need to identify which patients have more benefit, but also to organize the follow-up," Raposeiras-Roubín said.

As reported in the Journal of the American College of Cardiology, the investigators built the score using data from 1119 TEER procedures in Europe and Canada between 2012 and 2020 in the retrospective Percutaneous Mitral Valve Repair and Nutritional Status Registry (MIVNUT). During a median follow-up of 1.6 years, 31.9% of patients died and 28.4% were readmitted by the heart failure team.

After testing 33 variables, eight variables were independent predictors of mortality: age 75 years and older, anemia, estimated glomerular filtration rate below 60 mL/min per 1.73 m2, left ventricular ejection fraction below 40%, peripheral artery disease, chronic obstructive pulmonary disease, high diuretic dose (≥80 mg furosemide/d or use of ≥2 diuretic agents excluding antialdosteronic drugs), and no therapy with renin-angiotensin system inhibitors.

Each variable was assigned one point and the MitraScore calculated by adding the number of risk variables. Patients were also classified into MitraScore risk groups by tertiles: low (≤25th percentile), moderate-risk (25 - 75th percentile), and high-risk (≥75th percentile).

For each point on the MitraScore, the relative risk for all-cause mortality increased by 55% (hazard ratio [HR], 1.55; 95% CI; 1.44 - 1.67; P <.001).

Failed TEER procedure was associated with higher mortality rates in the univariate analysis (HR, 1.76; P = .009) and showed a trend in the multivariate analysis (HR, 1.53; P = .052).

The MitraScore was validated using data from 725 patients undergoing TEER between 2016 and 2020 in the prospective Italian Society of Interventional Cardiology Registry of Transcatheter Treatment of Mitral Valve Regurgitation (GIOTTO).

The MitraScore's discrimination ability for mortality was modest in the derivation and validation cohorts (c-statistics, 0.70 and 0.66, respectively), the authors note, but an improvement in comparison to those of the surgical scores: EuroSCORE II (both 0.61) and the Society of Thoracic Surgeons score (0.57 and 0.64, respectively).

As the MitraScore risk groups progressed from low to high, the risk for death and/or heart failure hospitalization increased from a hazard ratio of 1.97 for moderate risk to 3.69 for high risk in the derivation cohort and from 2.48 to 4.44, respectively, in the validation cohort.

The MitraScore also identified patients with a better response in terms of New York Heart Association functional class, both in the derivation and validation cohorts.

Other risk tools, such as the GRASP score and the Get With the Guidelines Heart Failure Risk Score, have been tested for patients undergoing TEER but, unlike the MitraScore, require a computer or app to calculate, Raposeiras-Roubín noted. "With more than three of these parameters, you can see that the patient has a moderate or high risk, so I think that's very easy to use in front of the patient."

Commenting for | Medscape Cardiology, Mohamad Alkhouli, MD, Mayo Clinic, Rochester, Minnesota, said: "This study brings our attention to the importance of risk prediction in patients undergoing these highly sophisticated, expensive procedures. But how it would help me in practice, I think it has a lot of limitations that would make it less scalable."

He noted that the study relied on the retrospective MIVNUT registry, which lacked clinical data on frailty, anatomic factors, and procedural characteristics, and on selected all-cause mortality, one of the easier hardcore endpoints to track. Based on its c-statistics of 0.70 and 0.66, it's also only mildly predictive.

"When you tell people that a flip of the coin is point five and you have a score of point six, if you tell this to a statistician, it wouldn't make a lot of sense to them to use such a score," Alkhouli said. "But I think in the cardiovascular community, we have long accepted, well, okay, we'll take a modest score if it's simple enough that we could just use it in clinic or if we could remember it off hand."

In an accompanying editorial, Alkhouli and colleagues highlight other similarly performing risk scores in the cardiologists' armamentarium, such as CHA2DS2-VASc, ATRIA, ABC, and the TAVR in-hospital mortality model, with derivative c-statistics of 0.61, 0.71, 0.68, and 0.67, respectively.

"Furthermore, with the availability of advanced statistics and the rise of artificial intelligence research and applications, should we not be aiming at more (not less) intricate risk schemes that incorporate imaging, laboratory and clinical data to allow more precise risk estimation?" they write.

Ultimately, Alkhouli said the MitraScore is a step forward and something to keep in the back of one's mind, but it will require further refinements and validation.

"I really don't think we have a perfect way of selecting patients and that's what probably made this paper important for a journal of JACC's caliber," he said. "But I think this will stir a lot of conversations. It will be good for the field."

Financial disclosures were not provided for the study. Alkhouli and colleagues report having no relevant financial disclosures.

J Am Coll Cardiol. 2022;79:562-573, 574-576. Abstract, Editorial

Follow Patrice Wendling on Twitter: @pwendl. For more from | Medscape Cardiology, follow us on Twitter and Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.