Prognostic value of preoperative peak atrial longitudinal strain for patients undergoing mitral valve surgery

Aims

In asymptomatic severe degenerative mitral regurgitation (MR), indications for surgery mainly rely on 2D echocardiographic criteria. Preoperative peak atrial longitudinal strain (PALS) has prognostic value for predicting clinical outcome after surgery for MR while Bernard demonstrated the prognostic value of an echocardiographic staging assessment of extra-valvular cardiac damage in patients with at least moderate MR.

The primary aim was to assess the additive prognostic value of PALS to Bernard's staging classification in patients undergoing surgery for degenerative MR.

Methods

Ambispective multicenter cohort study of patients with severe degenerative MR undergoing surgery. Patients were assigned to Bernard's stages based on pre-operative echocardiographic data, from which PALS values were obtained. Follow-up assessed the composite primary endpoint: all-cause mortality, hospitalization for heart failure, acute myocardial infarction, stroke, life-threatening bleeding or wound infections requiring re-intervention, failure requiring re-intervention.

Results

Three hundred patients (mean age 63 ± 13 years; 65% men) were enrolled with a median follow-up of 24 months. Bernard's staging predicted outcomes in our population, as demonstrated by survival analysis with Kaplan–Meier curves (log-rank 13.3; P  = 0.01) and Cox model (hazard ratio of 1.51; P < 0.001). Adding PALS improved prognostic performance ( χ ²  = 20.96 vs. 16.15), with an area under the curve of 0.70 vs. 0.64 at 24 months, a net reclassification improvement of 0.18 ( P  = 0.031), an integrated discrimination improvement of 0.02 ( P  = 0.020). We identified an optimal PALS cutoff of 18% associated with the primary endpoint.

Conclusions

PALS has a potential additive role in estimating the prognosis of patients undergoing surgery for severe degenerative MR.