Author + information
- Published online May 1, 2017.
- Chan Seok Park, MD, PhD,
- Ehimare Akhabue, MD,
- Manish Bansal, MD, DNB,
- Jarrod Marks, MD,
- Alaa Mabrouk Salem Omar, MD, PhD,
- Sean P. Pinney, MD,
- Anelechi Anyanwu, MD,
- Farooq Chaudhry, MD,
- Jagat Narula, MD, PhD and
- Partho P. Sengupta, MD, DM∗ ()
- ↵∗Icahn School of Medicine at Mount Sinai, One Gustave L. Levy Place, P.O. Box 1030, New York, New York 10029
Unanticipated right ventricular failure (RVF) is a commonly encountered complication after implantation of a left ventricular assist device (LVAD). An important observation from recent human and animal experimentation is the significance of normal interventricular septal geometry in maintenance of normal RV systolic function (1,2). We hypothesized that the dynamic changes in LV septal geometry can characterize the extent of interventricular coupling, and help predict the development of RVF following LVAD implantation.
We retrospectively reviewed clinical data from 98 patients (57 ± 11 years of age, 83% males) who underwent continuous-flow LVAD implantation. Speckle tracking echocardiography (STE) was used to measure LV and RV longitudinal strain and the actual LV radius (R) and the ideal LV radius (Ri) assuming circular shape of the LV cavity in short axis at the level of papillary muscles (Figure 1A). The ratio of the 2 (i.e., R/Ri) provided a quantitative estimate of septal motion with a value <1 indicating septal shift toward LV and a value >1 indicating shift in the opposite direction. The following indices of septal geometry were derived: end-diastolic (ED) to end-systolic (ES) R and Ri ratios (RED/RES and RiED/RiES, respectively), ED R/Ri ratio (R/RiED), ES R/Ri ratio (R/RiES), and instantaneous minimum and maximum R/Ri values (R/Rimin and R/Rimax, respectively).
A total of 24 (25%) developed RVF during the post-operative period as defined by the need for inotropic support for longer than 14 days, with 1 patient requiring an unplanned RVAD placement. On comparing patients with or without RVF, those with RVF had longer hospital stay (median stay 36 [interquartile range: 18 to 68] days vs. 20 [interquartile range: 14 to 37) days; p = 0.025) and higher in-hospital mortality (46% vs. 5%; p < 0.001).
There was no significant difference between the 2 groups for underlying heart failure etiology, comorbidities, biochemical or hemodynamic parameters, overall Michigan score, and the need for vasopressors and inotropic support. However, patients with RVF had higher peak systolic (66.6 ± 16.5 mm Hg vs. 55.4 ± 15.3 mm Hg; p < 0.012) and mean (41.7 ± 10.2 mm Hg vs. 36.5 ± 10 mm Hg; p < 0.071) pulmonary artery pressures.
The median time interval from echocardiogram to LVAD implantation was 10 days and was not different in the 2 groups. There was no significant difference in LV or left atrial dimensions, LV ejection fraction or global LV longitudinal strain (LS), RV dimensions, tricuspid regurgitation severity, RV fractional area shortening and tricuspid annular plane systolic excursion. However, the RVF patients showed lower RVLS (–9.9 ± 4.2% vs. –11.7 ± 4.8%; p = 0.095).
The patients with RVF had higher ED to ES ratios of actual (RED/RES 1.12 ± 0.06 vs. 1.08 ± 0.05; p = 0.017) and ideal (RiED/RiES 1.123 ± 0.067 vs. 1.084 ± 0.045; p = 0.001) radii as compared to those who did not develop RVF. However, in patients with RVF, the actual radii were consistently lower than ideal radii at all time points resulting in lower R/Ri ratios as compared to the patients who did not develop RVF (R/RiED 0.976 ± 0.059 vs. 1.010 ± 0.046; p = 0.004; R/RiES 0.980 ± 0.051 vs. 1.008 ± 0.046; p = 0.014; and R/Rimin 0.954 ± 0.062 vs. 0.987 ± 0.045; p = 0.007). Interestingly, in patients who developed RVF, R/Ri increased from end-diastole to end-systole whereas an opposite pattern was seen in those who did not develop RVF (Figure 1B).
Receiver-operating characteristic curves showed that RVLS, RED/RES, RiED/RiES, R/RiED, R/RiES, R/Rimin, and R/Rimax/R/Rimin were able to predict the post-operative development of RVF (area under the curve: 0.65, 0.63, 0.68, 0.68, 0.67, 0.71, and 0.64; p = 0.031, 0.062, 0.009, 0.009, 0.011, 0.002, and 0.045, respectively). RiED/RiES, R/RiED, and R/Rimin were shown to have incremental value in predicting RVF when added to RVLS using both net reclassification index (NRI) and integrated discrimination improvement (IDI) (for RiED/RiES, NRI: 0.63; 95% confidence interval [CI]: 0.18 to 1.07; p = 0.006; and IDI: 0.15; 95% CI: 0.06 to 0.25; p = 0.002; for R/RiED, NRI: 0.57; 95% CI: 0.13 to 1.02; p = 0.011; and IDI: 0.07; 95% CI: 0.01 to 0.13; p = 0.014; for R/Rimin, NRI: 0.60; 95% CI: 0.17 to 1.04; p = 0.010; and IDI: 0.07; 95% CI: 0.01 to 0.12; p = 0.027). Bland-Altman plots revealed that the intraobserver and interobserver bias and levels of agreements for R/Ri measurements were narrow (for R/RiED bias: −0.007 and −0.016 and level of agreement −0.111 to 0.097 and −0.147 to 0.115, respectively).
In summary, the present study demonstrates feasibility of STE-based measurement of dynamic changes in the interventricular septal geometry during the cardiac cycle. These findings reinforce previous observations that the interventricular septum plays a critical role in maintenance of RV function under a wide range of clinical settings, including the patients undergoing LVAD implantation (1,2) and STE, can be used to track septal geometry for predicting RVF after LVAD placement. Future prospective studies would need to confirm these observations including the predictive accuracy of R/Ri under different loading conditions.
Please note: Dr. Pinney has served as a consultant and received clinical trial support from St. Jude Medical. Dr. Chaudhry has served as a consultant for GE Medical Imaging and Lantheus Medical Imaging and has received research grant support from GE. Dr. Narula has given two talks for GE Healthcare; and his institution has received equipment grants from GE Healthcare and Philips Healthcare. Dr. Sengupta has served as an advisor for TeleHealth Robotics; and as a consultant for Hitachi Aloka Inc. and Heart Test Labs Inc. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Paul Grayburn, MD, served as Guest Editor for this article.
- 2017 American College of Cardiology Foundation