Myosteatosis: a new marker of muscle quality in chronic kidney disease strongly associated with mortality

Abstract Background and Aims Skeletal muscle composition disturbances, like sarcopenia and myosteatosis, are common in non-dialysis chronic kidney disease (ND-CKD) patients and seem to be associated with adverse clinical outcomes. Sarcopenia and myosteatosis can be evaluated by computed tomography (CT) by measuring skeletal muscle area (SMA) and muscular attenuation (MA) in Hounsfield units (HU) at the third lumbar vertebra, respectively, but the optimal cutoff points for diagnosis and outcome prediction are not established in chronic kidney disease (CKD) patients. We aimed to evaluate the prevalence of sarcopenia and myosteatosis in ND-CKD patients and to define the optimal cutoff values of SMA and MA to predict mortality. Method We conducted a retrospective cohort study including non-dialysis CKD patients referred to an outpatient clinic during a two-year period, who underwent a CT as part of clinical workup and with an available serum creatinine evaluation within a 90-days timeframe. Patients with a follow-up under 26 weeks after the CT were excluded. Area under the receiver operating characteristic curve (AuROC) analysis was used to evaluate the ability of SMA and MA to predict mortality and the Youden's index was used to determine the optimal cutoff point. Cox-regression analysis was employed to identify independent predictors of mortality. Results 167 patients (94% Caucasian, 50.9% male, 32.3% diabetics) with a mean age of 68.3 ± 16.4 years were included, most with CKD stage 3 and 4 (53.9%; mean estimated GFR 57.6 ± 33,1 ml/min/1.73m2 at baseline). During a median follow-up of 4.9 (4.2) years, 39 patients (23.4%) died. Median SMA was 127.7 (45.8) cm2 and there was a trend to increased mortality across lower SMA quartiles (1st quartile 32.6%, p = 0.026; 2nd quartile 26.8%, p = 0.095; 3rd quartile 21.4%, p = 0.261; 4th quartile 21.2% - reference). SMA showed a modest ability to predict mortality (AuROC 0.623) and the best cutoff found was 140.7 cm2. Median MA was 28.4 (13.8) HU and there was a statistically significant higher mortality across lower MA quartiles (1st quartile 42.9%, p<0.001; 2nd quartile 31.0%, p = 0.001; 3rd quartile 16.7%, p = 0.028; 4th quartile 2.4% - reference). MA showed a good ability to predict mortality (AuROC 0.733) and the best cutoff was 30 HU. Using the identified cutoff points, sarcopenia (SMA < 140.7 cm2) was present in 67.1% (n = 112) and myosteatosis (MA < 30 HU) in 56.3% (n = 94) of patients. In univariate Cox-regression both sarcopenia and myosteatosis were associated with increased mortality – Hazard ratio (HR) 4.34 (95% CI 1.68-11.19, p = 0.002) and 5.32 (95% CI 2.22-12.73, p<0.001), respectively. In multivariate Cox-regression models (adjusted for age, baseline estimated GFR and presence of diabetes) only myosteatosis kept its association with mortality – HR 2.87 (95% CI 1.15-7.16, p = 0.024). This association was also present when the model was adjusted for the presence of sarcopenia. Patients with myosteatosis were older (median age 77.3 [10.8] vs 62.7 [30.1], p<0.001) and had higher frequency of diabetes (42.6% vs 19.2%, p = 0.001), arterial hypertension (87.2% vs 57.5%, p<0,001), and heart failure (24.5% vs 6.8%, p<0.003). They had also higher BMI (29.3 [7.4] vs 25.1 [6.0] kg/m2, p<0,001), visceral obesity (77.7% vs 43.8%, p<0.001) and frequency of sarcopenia (75.5% vs 56.2%, p = 0.008). Myosteatosis was more frequent in CKD stage 3 to 5 patients, compared to CKD stage 1 or 2 (66.3% vs 42%, p = 0.002). Conclusion Sarcopenia and myosteatosis are prevalent in CKD patients, especially in advanced stages. However, reference values for this population are lacking. We found cutoff values for these muscle parameters using CT analysis in CKD patients, based on optimal stratification for mortality. Additionally, our study highlights that muscle quality (i.e., myosteatosis) may be more closely associated with mortality than muscle quantity (i.e., sarcopenia). Identifying patients at risk for these muscle abnormalities and early diagnosis are paramount for the subsequent implementation of therapeutic interventions.