RV Strain a Prognostic Marker for Mortality in COVID-19

Evidence of right heart dysfunction—especially right ventricular longitudinal strain (RVLS) on two-dimensional speckle-tracking echocardiography (2-D STE)—is tightly linked to higher mortality among patients with COVID-19, new data out of Wuhan, China, suggest.

This measure outperformed conventional echocardiographic measures, including RV fractional area change (RVFAC), tricuspid annular plane systolic excursion (TAPSE), and tricuspid tissue Doppler annular velocities (S’).

“It has been reported that SARS-CoV-2 infection could cause both pulmonary and systemic inflammations, which may contribute to RV failure through RV overload and direct damage to [the] cardiomyocyte,” write Yuman Li, MD, PhD, and colleagues, all based at Huazhong University of Science and Technology and the Hubei Province Key Laboratory of Molecular Imaging in Wuhan. Their results were published online last month in JACC: Cardiovascular Imaging.

Li et al say that RV dysfunction is worthy of close attention, since it is “not only a sign of increased pulmonary pressures, but also directly contribute[s] to cardiac insufficiency.” COVID-19 patients with this high-risk characteristic “might benefit from vigilant monitoring,” they add.

Commenting on the study for TCTMD, Rebecca Hahn, MD (NewYork-Presbyterian/Columbia University Irving Medical Center, New York, NY), listed some of the common indications for echo in a patient who is presumed or confirmed to be COVID-19-positive. These include hemodynamic instability/shock; signs of a major acute cardiovascular event, including chest pain or rising troponin; new arrhythmia; known history of cardiac disease; and evaluation for sepsis, endocarditis, or cardioembolic source of stroke.

Strain imaging, Hahn continued, is a particularly useful tool for gauging myocardial function in this setting. “It’s this beautiful way of really looking at much finer and smaller segments of the muscle and seeing if they’re contracting,” Hahn explained, adding that many studies have shown that LV strain is an “earlier and perhaps better marker of myocardial dysfunction than ejection fraction, which is what we’ve traditionally used.”

Right ventricular strain is of interest in Hahn’s area of expertise, tricuspid and mitral valve disease, where it also holds prognostic value, she noted. “Once you start to dilate the ventricle or have dysfunction of the ventricle either resulting from regurgitation or as the primary cause of regurgitation, then you’re in trouble clinically. Those patients have a worse outcome,” Hahn observed.

Although in her view Li et al’s study has some limitations, “no matter what, it’s interesting to look at the COVID population,” she said, “because we have seen in our experience at Columbia a lot of right heart dilatation and dysfunction. And the etiologies of these findings are numerous.”

Here, the investigators have been able to show that multiple echo parameters of RV function, including free-wall strain, TAPSE, and RVFAC “all predicted mortality,” Hahn pointed out. “But the best predictor of mortality was the RV longitudinal strain.”

RVLS for Risk Stratification

Li and colleagues enrolled 120 consecutive patients with COVID-19 (mean age 61 years; 48% men) who underwent echocardiography at their center. The patients were divided into tertiles based on RVLS values: 10.3% to 20.5%, 20.6% to 25.4%, and 25.5% to 35.7%.

Patients with the lowest RVLS values tended to have higher heart rates, higher levels of D-dimer and C-reactive protein, greater use of high-flow oxygen and invasive mechanical ventilation, and higher incidence of acute heart injury, acute respiratory distress syndrome, and deep vein thrombosis than those with the highest RVLS measurements. They also had higher mortality.

Over a median follow-up period, 18 patients died in total (15%). Nonsurvivors were more likely to have enlarged right-heart chambers, worsened RV function, and elevated pulmonary artery systolic pressure.

On multivariate Cox regression, RVLS more accurately predicted mortality, with a C-index of 0.89, than did models based on RVFAC and TAPSE (C-index = 0.84 and 0.83, respectively). Values for area under the curve were 0.87, 0.72, and 0.67, respectively, for the three parameters.

For mortality prediction, the best RVLS threshold was 23%, with a sensitivity of 94.4% and specificity of 64.7%.

“Our study demonstrated comprehensive assessment of RV function by 2-D STE may be essential for risk stratification in COVID-19 patients,” the researchers concluding, adding that additional studies done at other centers and using other software to perform 2-D STE are required.

Hahn agreed on the need for replication.

“There are a number of interesting things about the Li outcomes, and one of them is that for all of the [RV parameters], the cutoffs that they came up with as predicting outcomes are very close to what we could consider normal values,” she observed. Additionally, as the investigators note in the paper, here “RVLS was calculated as the mean of the strain values in the three segments of RV free wall.” For Hahn, it’s preferable to look at “global” strain that incorporates not only the free wall but also the interventricular septum.

One reason the authors may have found near-normal values predicting worse outcomes is the difference in patient populations compared to studies reporting normative data. For instance, obesity, which can adversely affect right heart function, is much more common in the United States than in China. “Therefore [this] probably deserves additional studies, particularly with a group of patients in the United States,” she advised.

Despite these caveats, this is “yet another study that shows that looking at a more granular parameter of mechanical function of the myocardium is going to give us a better idea of outcomes,” Hahn concluded.

One obstacle is that RVLS isn’t typically measured by clinicians in the COVID-19 setting, she said. “Most of us in the echo world now are trying to limit the time in the room, so this type of analysis typically will have to be done off-line or retrospectively. But that’s fine, [RVLS assessment] can be done in a very automated fashion right now . . . within a couple of minutes.”

Because of concerns about exposure to healthcare workers, recently released American Society of Echocardiography guidelines advise, among other safety measures, restricting echocardiography to patients whose management might be changed by its findings. As such, it would be hard to glean enough echo data at a single center; Hahn suggested that a multicenter registry based in New York might enable investigators to obtain cutoffs for RVLS relevant to the US population.

Photo Credit: Rebecca Hahn, MD