Psoas muscle index in sarcopenia following transjugular intrahepatic portosystemic shunt: A multicenter, retrospective study

Abstract Aims Although the skeletal muscle index at the third lumbar vertebra (L3‐SMI) is commonly utilized for the diagnosis of sarcopenia, the psoas muscle index at L3 (L3‐PMI) may serve as a reliable alternative indicator. This study aims to investigate the application of the PMI in patients who have undergone transjugular intrahepatic portosystemic shunt (TIPS). Methods This study included a cohort of 406 patients with cirrhosis who underwent TIPS between February 2016 and July 2022 across three medical centers in China. Clinical and imaging data, specifically L3‐SMI and L3‐PMI, were collected for these patients. The prognosis of the patients was assessed through re‐examinations and telephone follow‐ups, which extended up to 5 years. The diagnostic thresholds for sarcopenia, as defined by L3‐PMI and L3‐SMI, were established at 6.36 or 42.00 cm²/m² for males and 3.92 or 38.00 cm²/m² for females, respectively. Cox proportional hazards and Kaplan–Meier (K‐M) analyses were employed to evaluate patient survival. Results The findings indicated that both L3‐PMI and L3‐SMI before TIPS were independent risk factors for mortality. The results of the paired t ‐test demonstrated a significant increase in L3‐PMI 1 month post‐TIPS (5.11 ± 1.81 vs. 5.71 ± 1.90 cm²/m², p < 0.001), whereas L3‐SMI did not exhibit a significant increase until 6 months post‐TIPS (45.45 ± 9.41 vs. 48.59 ± 10.38 cm²/m², p < 0.001). Among these patients, 191 (47.0%) and 159 (39.2%) were diagnosed with sarcopenia according to the L3‐PMI and L3‐SMI models, respectively. Patients identified as sarcopenic by both indicators demonstrated a significantly lower survival rate (L3‐SMI model: hazard ratio [HR], 1.913; 95% confidence interval [CI], 1.094–3.410; log‐rank p = 0.020; L3‐PMI model: HR, 1.867; 95% CI, 1.059–3.290; log‐rank p = 0.030). In sarcopenic patients, the reversal of sarcopenia associated with improved survival occurred 1 month after TIPS in the L3‐PMI model (HR, 2.675; 95% CI, 1.245–5.735; log‐rank p = 0.012), while a similar effect was not observed until 6 months post‐TIPS in the L3‐SMI model (HR, 3.342; 95% CI, 1.477–7.560; log‐rank p = 0.004). Conclusions L3‐PMI and L3‐SMI demonstrate comparable efficacy in diagnosing cirrhosis‐related sarcopenia. Furthermore, L3‐PMI has the capacity to identify improvements in sarcopenia as early as 1 month following TIPS, thereby providing earlier insights into patient survival outcomes.