Patterns of left and right ventricular dysfunction and their clinical outcomes in septic patients

Abstract

Section Background

To investigate the various types of left ventricular (LV) and right ventricular (RV) dysfunction and their association with clinical outcomes in septic patients.

AbstractSection Methods

We retrospectively reviewed a cohort of sepsis and septic shock patients admitted to a tertiary hospital intensive care unit (ICU). Left ventricular dysfunction was determined by left ventricular ejection fraction (LVEF). Right ventricular systolic dysfunction (RVSD) was determined by tricuspid annular plane systolic excursion (TAPSE), fractional area change (FAC), or RV S’ velocity. Cardiac function was categorized into five types: (1) normal cardiac function (LVEF > 50% but ≤ 70%, TAPSE ≥ 17 mm, RV FAC ≥ 35%, and RV S’ ≥9.5 cm/sec); (2) biventricular dysfunction (LVEF ≤ 50%, TAPSE < 17 mm or RV FAC < 35% or RV S’ < 9.5 cm/sec); (3) isolated LV dysfunction (LVEF ≤ 50%, TAPSE ≥ 17 mm, RV FAC ≥ 35%, and RV S’ ≥9.5 cm/sec); (4) isolated RV dysfunction (LVEF > 50%, TAPSE < 17 mm or RV FAC < 35% or RV S’ < 9.5 cm/sec); and (5) hyperdynamic LV function (LVEF > 70%, TAPSE ≥ 17 mm, RV FAC ≥ 35%, and RV S’ ≥9.5 cm/sec). The primary outcome was 30-day mortality.

AbstractSection Results

A cohort of 702 septic patients was analyzed. Patients with biventricular dysfunction (n = 113) and isolated RV dysfunction (n = 117) exhibited mortality rates of 34.5% and 36.7%, respectively. In comparison, lower mortality rates were observed in patients with isolated LV dysfunction (n = 72), hyperdynamic LV function (n = 86), and normal cardiac function (n = 314), with rates of 15.3%, 15.1% and 9.2%, respectively. Cox regression analysis confirmed that biventricular dysfunction (hazard ratio [HR] 2.312, 95% confidence interval [CI] 1.291–4.139, p = 0.005) and isolated RV dysfunction (HR 2.655, 95% CI 1.455–4.843, p = 0.001) were independently associated with 30-day mortality. Neither isolated LV dysfunction (HR 1.171, 95% CI 0.463–2.960, p = 0.739) nor hyperdynamic LV function (HR 2.153, 95% CI 0.971–4.773, p = 0.059) were independently associated with 30-day mortality.

AbstractSection Conclusions

Septic patients with biventricular dysfunction or isolated RV dysfunction had significantly higher 30-day mortality rates compared to those with normal cardiac function. Notably, hyperdynamic LV function also showed a trend toward higher 30-day mortality than normal cardiac function, although this association did not reach statistical significance. In contrast, isolated LV dysfunction was not associated with 30-day mortality.

Background

Sepsis is a major global health concern and is still associated with unacceptably high mortality in critically ill patients [1, 2]. Cardiac dysfunction is a frequent complication of sepsis and plays a critical role in exacerbating hemodynamic instability, further complicating patient outcomes [3].

With the emergence of advanced critical care echocardiography, the precise definition of septic cardiomyopathy has become increasingly complex, introducing significant uncertainty into its characterization. Currently, there is no formalized or consensus-based definition of septic cardiomyopathy [4]. Previous studies have highlighted that both left ventricular (LV) and right ventricular (RV) dysfunction can occur in sepsis, either in isolation or concurrently [5,6,7]. However, comprehensive assessment of LV and RV function of sepsis patients remain limited, and the prognostic significance of septic cardiomyopathy has yet to be fully elucidated [8]. Therefore, our study aims to investigate the patterns of LV and RV dysfunction in septic patients and their association with hemodynamic outcomes and 30-day mortality. By addressing these gaps, we hope to contribute to a deeper understanding of septic cardiomyopathy and its clinical implications.

Study design and methods

This is a secondary analysis of a prospectively identified septic patients admitted from November 2019 to December 2024 at a tertiary hospital intensive care unit (ICU) [9].

Study population

Sepsis was defined as a life-threatening organ dysfunction secondary to a dysregulated host response to infection, manifested by an increase in Sequential Organ Failure Assessment (SOFA) score of 2 points or more [10].

Patients were excluded if they had the following conditions including a history of chronic heart failure or cardiac surgery, severe valvular disease, acute coronary syndrome within 1 week, moderate to severe chronic pulmonary hypertension [11], or inadequate echocardiographic image quality. We also excluded patients who withheld life support.

The study was conducted in compliance with the Declaration of Helsinki and was approved by the ethics committee of our institution (Approval No. ZS-2166). Written informed consent on the review and research of the patients’ medical data was signed by the patients or their surrogates.

Echocardiography

A comprehensive transthoracic echocardiography (TTE) was performed on septic patients within 24 h of ICU admission. TTE was performed by two experienced physicians (H. Z. and Q. Z.), each with over ten years of expertise in critical care echocardiography, using a 2.5 MHZ phased array probe (Mindray, M9, Shenzhen, China). Images were stored for offline analysis and measurements were obtained from at least three cardiac cycles with interpretations adhering to the PRICES statement [12]. Intra- and interobserver variabilities for key measurements by these investigators have been previously reported [9].

A standardized TTE protocol was established prior to study initiation, incorporating comprehensive quantitative assessments of right ventricular systolic, LV diastolic and systolic function. Tricuspid annular plane systolic excursion (TAPSE), RV fractional area change (FAC), right and left ventricular end-diastolic area ratio (R/LVEDA), left ventricular outflow tract velocity-time integral (LVOT-VTI), left ventricular ejection fraction (LVEF), mitral annular plane systolic excursion (MAPSE), mitral peak E velocity (E), tissue Doppler velocity of lateral and medial mitral annuli at early diastole (e’), tricuspid regurgitation, internal diameter of inferior vena cava (IVCD) and stroke volume index (SVI) were obtained using previously described methods [9]. RV S’ was measured by positioning the sample volume on the lateral tricuspid valve ring with pulsed-wave tissue Doppler in apical 4-chamber view. Cardiac index (CI) was calculated as heart rate (HR) × SVI, with an abnormal cutoff value set at below 2.5 L/min/m2 [13]. RV dilation was defined as R/LVEDA ≥ 0.6 and acute cor pulmonale (ACP) was defined as RV dilation with septal paradoxical motion at end-systole [14]. RVSD was determined by TAPSE < 17 mm or RV FAC < 35% or RV S’ < 9.5 cm/sec [15]. Systemic venous congestion was defined as IVCD ≥ 20 mm plus hepatic vein S < D [16, 17]. LV Apical ballooning was diagnosed based on the following criteria: (1) symmetrical and apical regional wall motion abnormality with LVEF ≤ 50%. (2) Non-specific ST-T alterations with moderately elevated serum troponin; (3) Repeated echocardiography in a few days confirming complete recovery of LV function [18, 19].

Cardiac function was categorized into five types based on LV and RV systolic function: (1) normal cardiac function (LVEF > 50% but ≤ 70%, TAPSE ≥ 17 mm, RV FAC ≥ 35%, and RV S’ ≥9.5 cm/sec); (2) biventricular dysfunction (LVEF ≤ 50%, TAPSE < 17 mm or RV FAC < 35% or RV S’ < 9.5 cm/sec); (3) isolated LV dysfunction (LVEF ≤ 50%, TAPSE ≥ 17 mm, RV FAC ≥ 35%, and RV S’ ≥9.5 cm/sec); (4) isolated RV dysfunction (LVEF > 50%, TAPSE < 17 mm or RV FAC < 35% or RV S’ < 9.5 cm/sec); and (5) hyperdynamic LV function (LVEF > 70%, TAPSE ≥ 17 mm, RV FAC ≥ 35%, and RV S’ ≥9.5 cm/sec).

Clinical data collected

Upon admission to the ICU, the demographic details of the patients, along with their Acute Physiology and Chronic Health Evaluation (APACHE) II scores and SOFA scores, were meticulously recorded. Additionally, at the time of the echocardiographic examination, a set of physiological parameters was gathered for each patient. These parameters included HR, mean arterial pressure (MAP), central venous pressure (CVP), norepinephrine (NE) dose, positive end-expiratory pressure (PEEP), plateau pressure (Pplat), arterial lactate, partial pressure of arterial oxygen to fraction of inspired oxygen ratio (PaO2/FiO2), partial pressure of carbon dioxide (PaCO2) and the total volume of fluid balance. We also collected the maximum values of biomarkers within the first 24 h ICU admission, including neutrophil count, lymphocyte count, platelet count, C-reactive protein (CRP) level, troponin I level and D-dimer level.

Outcomes

The primary outcome was the 30-day all-cause mortality among septic patients. The secondary outcomes included ICU-free days, the incidence of ACP, LV apical ballooning, low CI and systemic venous congestion.

Statistical analysis

Continuous variables are expressed as mean ± SD or the median and interquartile range. Categorical variables are presented as frequencies and percentages. The normality of the distributions for the continuous variables was evaluated using the Kolmogorov-Smirnov test. Differences among groups were assessed using the Kruskal-Wallis test, the chi-squared test, or Fisher’s exact test, as appropriate. Cumulative survival curves of the 30-day follow-up were estimated using the Kaplan‒Meier method. Prognostic factors for a 30-day mortality among the four groups were determined using univariate and multivariate Cox regression models. Based on the results of the univariate Cox regression model, the baseline covariables with p values < 0.1 were selected by the stepwise method and were included in the multivariate Cox regression models as covariables. Statistical analyses were conducted using SPSS 22.0 (SPSSInc., Chicago, Ill., USA) and Graphpad Prism (6.01 for Windows, GraphPad Software, La Jolla California, USA). Two-tailed p < 0.05 was considered significant.

Results

General characteristics

A total of 848 patients with sepsis were evaluated for study enrolment. We excluded 146 patients and enrolled 702 patients were enrolled, among whom 314 patients had normal cardiac function, 113 had biventricular dysfunction, 72 had isolated LV dysfunction, 117 had isolated RV dysfunction, and 86 had hyperdynamic LV function (Fig. 1). The five groups differed significantly in baseline characteristics including age, APACHE II score, SOFA score, vasocative inotropic score, HR, CVP, PEEP, Pplat, PaCO2, PaO2/FiO2, proportion of lung infection and acute respiratory distress syndrome (ARDS), lactate level, ICU-free days and 30-day mortality (all p < 0.01) (Table 1). All five groups exhibited comparable neutrophil counts (p = 0.429), while patients with biventricular dysfunction demonstrated the lowest lymphocyte counts (p = 0.001). Patients with isolated LV dysfunction showed the highest CRP levels, although this difference reached statistical significance only when compared to the hyperdynamic LV group (p < 0.05). Furthermore, patients with biventricular dysfunction had significantly elevated troponin I levels compared to those with normal cardiac function and hyperdynamic LV function (p < 0.05). Similarly, patients with isolated LV dysfunction also displayed higher troponin I levels compared to those with normal cardiac function and hyperdynamic LV function (p < 0.05) (Table 2).

[IMAGE OMITTED: SEE PDF]

[IMAGE OMITTED: SEE PDF]

[IMAGE OMITTED: SEE PDF]

Comparison of echocardiographic parameters among the five groups

The five groups showed significant differences in LVEF, TAPSE, RV FAC and RV S’ (p < 0.001). The five groups also differed significantly in E/e’ and TAPSE/PASP (p < 0.05). ACP was exclusively identified in patients with biventricular dysfunction and isolated RV dysfunction (p < 0.001). LV Apical ballooning was detected only in patients with isolated LV dysfunction (p < 0.001) (Table 1).

Primary outcome

Patients with biventricular dysfunction and isolated RV dysfunction exhibited mortality rates of 34.5% and 36.7%, respectively. In comparison, lower mortality rates were observed in patients with isolated LV dysfunction, hyperdynamic LV function, and normal cardiac function, with rates of 15.3%, 15.1% and 9.2%, respectively. Kaplan‒Meier curve analysis revealed that patients with biventricular dysfunction (log-rank 43.482, p < 0.001) and those with isolated RV dysfunction (log-rank 53.031, p < 0.001) had significantly higher 30-day mortality rates compared to patients with normal function. In contrast, patients with isolated LV dysfunction (log-rank 2.499, p = 0.114) and hyperdynamic LV function (log-rank 2.421, p = 0.120) showed 30-day mortality rates similar to those with normal function (Table 1; Fig. 2).

[IMAGE OMITTED: SEE PDF]

Cox regression analysis, adjusted for relevant covariates (age, APACHE II score, mechanical ventilation, PEEP and Pplat levels, vasoactive inotropic score, maximum lactate level, PaO2/FiO2, and ARDS occurrence), confirmed that biventricular dysfunction(hazard ratio [HR] 2.312, 95% confidence interval [CI] 1.291–4.139, p = 0.005) and isolated RV dysfunction (HR 2.655, 95% CI 1.455–4.843, p = 0.001) were independently associated with 30-day mortality. In contrast, neither isolated LV dysfunction (HR 1.171, 95% CI 0.463–2.960, p = 0.739) nor hyperdynamic LV function (HR 2.153, 95% CI 0.971–4.773, p = 0.059) were independently associated with 30-day mortality (Table 3).

[IMAGE OMITTED: SEE PDF]

Secondary outcomes

Patients with biventricular dysfunction, isolated RV dysfunction, and isolated LV dysfunction exhibited significantly fewer ICU-free days compared to the other two groups (p < 0.05) (Table 1; Fig. 3A). Additionally, these patients had a higher prevalence of low CI than the other two groups (p < 0.001). Furthermore, patients with biventricular dysfunction and isolated RV dysfunction showed a greater proportion of systemic venous congestion compared to the other three groups (p < 0.001). LV apical ballooning was predominantly observed in patients with isolated LV dysfunction, while ACP was primarily detected in those with biventricular dysfunction and isolated RV dysfunction (p < 0.001) (Table 2; Fig. 3B).

[IMAGE OMITTED: SEE PDF]

Discussion

This study explored the associations between different types of RV and LV dysfunctions and clinical outcome in septic patients. Our findings underscore that both biventricular dysfunction and isolated RV dysfunction were significant predictors of 30-day mortality. Notably, hyperdynamic LV function also trended toward higher 30-day mortality, though this difference was not statistically significant. However, isolated LV dysfunction was not associated with 30-day mortality.

Although LV and RV function in sepsis has been explored in previous studies, a comprehensive assessment of LV and RV function remains underexplored. This study identified RV function as a key prognostic factor in this patient population, with both biventricular dysfunction and isolated RV dysfunction carrying significant prognostic implications. Patients with biventricular dysfunction exhibited the lowest CI and a higher prevalence of systemic venous congestion, which might explain their elevated mortality. Similarly, isolated RV dysfunction also was a predictor of worse clinical outcomes, as these patients had a higher incidence of ARDS and ACP, markers associated with increased mortality [20]. Previous studies have demonstrated that RV dysfunction is related to both short- and long-term outcomes in septic patients [21, 22]. In contrast, one prior study reported that only isolated RV dysfunction, not bivenricular dysfunction, was of prognostic significance [7]. However, this study defined RV dysfunction based on RV enlargement and included only a small subset of patients (41/388) with TAPSE measurements. Our earlier research found that sepsis patients with RV enlargement but preserved RV systolic function were not associated with increased mortality [23].

We observed that more than 10% of the patients exhibited hyperdynamic LV. Notably, the hyperdynamic LV group did not show elevated CRP levels. We speculate that the hyperdynamic state was primarily attributable to reduced preload or afterload. A previous study reported that hyperdynamic LV was independently linked to higher in-hospital mortality [24]. In contrast, our study observed only a non-significant trend toward higher 30-day mortality in patients with hyperdynamic LV function. This discrepancy may be attributed to differences in the timing of echocardiography. While our protocol involved echocardiographic assessment within 24 h of ICU admission, the prior study included evaluations conducted within 3 days of admission. Further investigation remains necessary to clarify the prognostic implications of hyperdynamic LV function in sepsis.

We also did not find patients with isolated LV dysfunction was associated with 30-day mortality, despite these patients exhibiting lower CI. Sepsis-related cardiomyopathy often involve both ventricles, but this is not a necessity, particularly in cases without RV afterload increase [5, 25]. Interestingly, patients with isolated LV dysfunction demonstrated elevated CRP levels, suggesting that increased cytokine activity may play a role in this condition. Although we did not directly measure cytokine levels such as IL-6, prior studies have established a strong correlation between CRP and IL-6 in sepsis [26]. CRP is primarily synthesized by heptocytes in response to cytokines, notably IL-6 [27]. Additionally, approximately one-fifth of these patients exhibited LV apical ballooning, a condition typically confined to the LV rather than the both ventricles. This observation aligns with a previous study that identified sepsis as a potential trigger for LV apical ballooning in critically ill patients [19].

Limitations

This study has several limitations. First, as a retrospective single-center study, its generalizability may be limited. Validation using data from other centers is recommended to confirm our findings. Second, our study focused exclusively on patients with LV apical ballooning, the most common type of takotsubo cardiomyopathy. Previous research indicates that apical type accounts for approximately 81% of all takotsubo cases [28]. Third, data on inflammatory biomarkers were limited in this study. Fourth, in this observational study, variations in volume status—due to differences in resuscitation strategies or timing of echocardiographic examination— may have influenced cardiac function classification. Future study should include a broader range of biomarkers, such as IL-6, TNF-α, to better elucidate the relationship between inflammation and different types of cardiac dysfunction. Despite these limitations, our study offers valuable insights into the assessment of cardiac dysfunction in sepsis and paves the way for further advancements in the management of these patients.

Conclusions

Patients with biventricular dysfunction or isolated RV dysfunction exhibited significantly higher 30-day mortality compared to those with normal cardiac function. Notably, hyperdynamic LV function also showed a trend toward higher 30-day mortality than normal cardiac function, although this association did not reach statistical significance. In contrast, isolated LV dysfunction was not associated with 30-day mortality.

Data availability

All datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Abbreviations

LV:

Left ventricular

RV:

Right ventricular

ICU:

Intensive care unit

TTE:

Transthoracic echocardiography

SOFA:

Sequential organ failure assessment

TAPSE:

Tricuspid annular plane systolic excursion

R/LVEDA:

Right and left ventricular end-diastolic area ratio

FAC:

Fractional area change

LVOT:

Left ventricular outflow tract

VTI:

Velocity-time integral

LVEF:

Left ventricular ejection fraction

MAPSE:

Mitral annular plane systolic excursion

E:

Mitral annular peak velocity

e’:

Tissue Doppler velocity of lateral and medial mitral annuli at early diastole

IVCD:

Inferior vena cava diameter

SVI:

Stroke volume index

CI:

Cardiac index

HR:

Heart rate

APACHE:

Acute physiology and chronic health evaluation

CVP:

Central venous pressure

ACP:

Acute cor pulmonale

NE:

Norepinephrine

PEEP:

Positive end-expiratory pressure

Pplat:

Plateau pressure

PaCO2 :

Partial pressure of carbon dioxide

PaO2/FiO2 :

Partial pressure of arterial oxygen to fraction of inspired oxygen ratio

CRP:

C reactive protein

References

Cecconi M, Evans L, Levy M, Rhodes A. Sepsis and septic shock. Lancet. 2018;392(10141):75–87.

Gaieski DF, Edwards JM, Kallan MJ, Carr BG. Benchmarking the incidence and mortality of severe sepsis in the united States. Crit Care Med. 2013;41(5):1167–74.

Vieillard-Baron A, Caille V, Charron C, Belliard G, Page B, Jardin F. Actual incidence of global left ventricular hypokinesia in adult septic shock. Crit Care Med. 2008;36(6):1701–6.

Ehrman RR, Sullivan AN, Favot MJ, et al. Pathophysiology, echocardiographic evaluation, biomarker findings, and prognostic implications of septic cardiomyopathy: a review of the literature. Crit Care. 2018;22:112.

Pulido JN, Afessa B, Masaki M, et al. Clinical spectrum, frequency, and significance of myocardial dysfunction in severe sepsis and septic shock. Mayo Clin Proc. 2012;87(7):620–8.

Geri G, Vignon P, Aubry A, et al. Cardiovascular clusters in septic shock combining clinical and echocardiographic parameters: a post hoc analysis. Intensive Care Med. 2019;45(5):657–67.

Vallabhajosyula S, Kumar M, Pandompatam G, et al. Prognostic impact of isolated right ventricular dysfunction in sepsis and septic shock: an 8-year historical cohort study. Ann Intensive Care. 2017;7(1):94.

Hollenberg SM, Singer M. Pathophysiology of sepsis-induced cardiomyopathy. Nat Rev Cardiol. 2021;18(6):424–34.

Zhang H, Zhang D, Wang X, et al. Prognostic implication of a novel right ventricular injury score in septic patients. ESC Heart Fail. 2023;10(2):1205–13.

Singer M, Deutschman CS, Seymour CW, et al. The third international consensus definitions for Sepsis and septic shock (Sepsis-3). JAMA. 2016;315(8):801–10.

Zaidi A, Knight DS, Augustine DX, et al. Echocardiographic assessment of the right heart in adults: a practical guideline from the British society of echocardiography. Echo Res Pract. 2020;7(1):G19–1941.

Sanfilippo F, Huang S, Herpain A, et al. The PRICES statement: an ESICM expert consensus on methodology for conducting and reporting critical care echocardiography research studies. Intensive Care Med. 2021;47(1):1–13.

VanDyck TJ, Pinsky MR. Hemodynamic monitoring in cardiogenic shock. Curr Opin Crit Care. 2021;27:454–9.

Jardin F, Dubourg O, Bourdarias JP. Echocardiographic pattern of acute Cor pulmonale. Chest. 1997;111(1):209–17.

Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, Flachskampf FA, Foster E, Goldstein SA, Kuznetsova T, Lancellotti P, Muraru D, Picard MH, Rietzschel ER, Rudski L, Spencer KT, Tsang W, Voigt JU. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American society of echocardiography and the European association of cardiovascular imaging. Eur Heart J Cardiovasc Imaging. 2015;16:233–70.

Beaubien-Souligny W, Rola P, Haycock K, Bouchard J, Lamarche Y, Spiegel R, Denault AY. Quantifying systemic congestion with Point-Of-Care ultrasound: development of the venous excess ultrasound grading system. Ultrasound J. 2020;12:16.

Spiegel R, Teeter W, Sullivan S, Tupchong K, Mohammed N, Sutherland M, Leibner E, Rola P, Galvagno SM Jr, Murthi SB. The use of venous doppler to predict adverse kidney events in a general ICU cohort. Crit Care. 2020;24:615.

Chockalingam A, Mehra A, Dorairajan S, et al. Acute left ventricular dysfunction in the critically ill. Chest. 2010;138:198–207.

Park JH, Kang SJ, Song JK, et al. Left ventricular apical ballooning due to severe physical stress in patients admitted to the medical ICU. Chest. 2005;128:296–302.

Mekontso Dessap A, Boissier F, Charron C, et al. Acute Cor pulmonale during protective ventilation for acute respiratory distress syndrome: prevalence, predictors, and clinical impact. Intensive Care Med. 2016;42:862–70.

Lanspa MJ, Cirulis MM, Wiley BM, et al. Right ventricular dysfunction in early Sepsis and septic shock. Chest. 2021;159(3):1055–63.

Vallabhajosyula S, Shankar A, Vojjini R, et al. Impact of right ventricular dysfunction on Short-term and Long-term mortality in sepsis: A Meta-analysis of 1,373 patients. Chest. 2021;159(6):2254–63.

Zhang H, Huang W, Zhang Q, et al. Prevalence and prognostic value of various types of right ventricular dysfunction in mechanically ventilated septic patients. Ann Intensive Care. 2021;11:108.

Dugar S, Sato R, Chawla S, et al. Is left ventricular systolic dysfunction associated with increased mortality among patients with Sepsis and septic shock. Chest. 2023;163:1437–47.

Vieillard-Baron A, Cecconi M. Understanding cardiac failure in sepsis. Intensive Care Med. 2014;40(10):1560–3.

Oberhoffer M, Karzai W, Meier-Hellmann A, Bögel D, Fassbinder J, Reinhart K. Sensitivity and specificity of various markers of inflammation for the prediction of tumor necrosis factor-alpha and interleukin-6 in patients with sepsis. Crit Care Med. 1999;27(9):1814–8.

Vincent JL, Donadello K, Schmit X. Biomarkers in the critically ill patient: C-reactive protein. Crit Care Clin. 2011;27(2):241–51.

Templin C, Ghadri JR, Diekmann J, et al. Clinical features and outcomes of Takotsubo (Stress) cardiomyopathy. N Engl J Med. 2015;373:929–38.