There is more than just longitudinal strain: Prognostic significance of biventricular circumferential mechanics

Abstract

Introduction: Despite the significant contribution of circumferential shortening to the global ventricular function, data are scarce concerning its prognostic value on long-term mortality. Accordingly, our study aimed to assess both left (LV) and right ventricular (RV) global longitudinal (GLS) and global circumferential strain (GCS) using three-dimensional echocardiography (3DE) to determine their prognostic importance.

Methods: Three hundred fifty-seven patients with a wide variety of left-sided cardiac diseases were retrospectively identified (64 ± 15 years, 70% males) who underwent clinically indicated 3DE. LV and RV GLS, and GCS were quantified. To determine the prognostic power of the different patterns of biventricular mechanics, we divided the patient population into four groups. Group 1 consisted of patients with both LV GLS and RV GCS above the respective median values; Group 2 was defined as patients with LV GLS below the median while RV GCS above the median, whereas in Group 3, patients had LV GLS values above the median, while RV GCS was below median. Group 4 was defined as patients with both LV GLS and RV GCS below the median. Patients were followed up for a median of 41 months. The primary endpoint was all-cause mortality.

Results: Fifty-five patients (15%) met the primary endpoint. Impaired values of both LV GCS (HR, 1.056 [95% CI, 1.027-1.085], p < 0.001) and RV GCS (1.115 [1.068-1.164], p < 0.001) were associated with increased risk of death by univariable Cox regression. Patients with both LV GLS and RV GCS below the median (Group 4) had a more than 5-fold increased risk of death compared with those in Group 1 (5.089 [2.399-10.793], p < 0.001) and more than 3.5-fold compared with those in Group 2 (3.565 [1.256-10.122], p = 0.017). Interestingly, there was no significant difference in mortality between Group 3 (with LV GLS above the median) and Group 4, but being categorized into Group 3 versus Group 1 still held a more than 3-fold risk (3.099 [1.284-7.484], p = 0.012).

Discussion: The impaired values of both LV and RV GCS are associated with long-term all-cause mortality, emphasizing the importance of assessing biventricular circumferential mechanics. Reduced RV GCS is associated with significantly increased risk of mortality even if LV GLS is preserved.

Keywords: 3D echocardiography; global circumferential strain; global longitudinal strain; heart failure; speckle tracking echocardiography.

Figure 1

Identification of the best-fit models, including left (LV) and right ventricular (RV) functional parameters by multivariable Cox regression analysis based on Akaike Information Criterion (AIC). (A) Depicts different models with only clinical characteristics (Step I) and 3D LV mechanical parameters added one by one (Step II). In Step II, adding LV global longitudinal strain (GLS) to the model resulted in the lowest (best) AIC value. (B) Shows the added value of 3D RV mechanical parameters (Step III). In Step III, adding RV global circumferential strain (GCS) to the previously established model in Step II (clinical characteristics and LV GLS) resulted in the best fit to our data as confirmed by the lowest AIC value.

Figure 2

Survival analysis of the different groups. Based on the respective median values of left ventricular global longitudinal strain (LV GLS, −15.9%) and right ventricular global circumferential strain (RV GCS, −17.9%), patients were divided into four groups. The survival of the four groups is visualized on Kaplan-Meier curves, and the log-rank test was performed for comparison.

Figure 3

Three-dimensional schematic models depict representative cases of different biventricular mechanical patterns in patients from the respective groups. The four groups were divided based on the median values of left ventricular global longitudinal strain (LV GLS, −15.9%) and right ventricular global circumferential strain (RV GCS, −17.9%). Upward arrows represent strain values better than the median (more negative), downward arrows represent strain values worse than the median (less negative). Light green mesh – left ventricular end-diastolic volume; dark green surface – left ventricular end-systolic volume; green arrow – LV longitudinal shortening (GLS); light blue mesh – right ventricular end-diastolic volume; dark blue surface – right ventricular end-systolic volume; blue arrow – RV circumferential shortening (GCS).