RE: Utility of Plasma Myostatin as a Predictive Biomarker for Post‐Intensive Care Syndrome in Patients With Sepsis

ABSTRACTAimPost intensive care syndrome (PICS) is a critical issue in postsepsis care; however, no reliable biomarker exists to predict PICS. We hypothesized that plasma myostatin, a cytokine involved in muscle and brain function, could serve as a predictive biomarker for PICS.MethodsThis single‐center prospective observational study included adult patients with sepsis admitted to the intensive care unit (ICU). Plasma myostatin concentrations were measured using enzyme‐linked immunosorbent assay on days 1, 3–4, and 6–7 after sepsis diagnosis and ICU admission. PICS was assessed 6 months post‐ICU discharge via telephone interviews using the Barthel Index, Short‐Memory Questionnaire score, and Hospital Anxiety and Depression Scale. Multivariable logistic regression analysis was conducted to determine whether myostatin levels independently predicted PICS. Predictive performance was evaluated using the area under the curve (AUC).ResultsSeventy‐seven patients were enrolled (mean age: 71 ± 14 years; median SOFA score 7 [IQR: 6–11]). Plasma myostatin concentrations were 544 (311–1015) pg/mL on day 1, 495 (302–651) pg/mL on days 3–4, and 536 (385–817) pg/mL on days 6–7. Decreased plasma myostatin levels on day 1 and on days 6–7 contributed to identifying patients with cognitive and physical impairments, respectively (p = 0.04). A decreased plasma myostatin level on day 1 yielded an AUC of 0.70 for predicting cognitive impairment, whereas the day 6–7 level yielded an AUC of 0.76 for predicting physical impairment.ConclusionLower plasma myostatin concentrations in the acute phase of sepsis may serve as a biomarker for predicting PICS‐related physical and cognitive impairments.

We sincerely thank Drs. Daungsupawong and Wiwanitkit for their thoughtful and constructive comments on our article, “Utility of Plasma Myostatin as a Predictive Biomarker for Post-Intensive Care Syndrome (PICS) in Patients with Sepsis.” Their insights provide an important opportunity to clarify several aspects of our study. First, we acknowledge that this was a single-center, exploratory study that predominantly included elderly patients. As stated in the manuscript, our findings should be regarded as hypothesis-generating, and confirmation in larger, multicenter cohorts is necessary. Major limitations include the small sample size, potential selection bias, and limited adjustment for confounding variables. Second, regarding the assessment of sarcopenia, skeletal muscle mass was evaluated using the psoas muscle index at the third lumbar vertebral level on computed tomography. Previous studies have demonstrated that computed tomography-derived psoas muscle index correlates with whole-body skeletal muscle mass and serves as a reasonable surrogate marker for sarcopenia in critically ill patients [1]. Sarcopenia was included as a covariate in our multivariable model, and lower plasma myostatin levels remained independently associated with long-term functional impairment. Because of the retrospective study design, muscle assessment was limited to a single time point. Future prospective studies should incorporate longitudinal muscle evaluation using ultrasound, which has been shown to detect low muscularity at ICU admission [1]. Serial ultrasound measurements combined with myostatin kinetics may better capture dynamic muscle changes after sepsis. Third, the predictive performance of plasma myostatin, with area-under-the-curve values ranging from 0.70 to 0.76, indicates fair discrimination. According to commonly accepted criteria, an AUC of 0.70–0.79 reflects fair accuracy, 0.80–0.89 good accuracy, and ≥ 0.90 excellent accuracy [2]. Therefore, although statistically significant, our findings should be interpreted cautiously and require external validation before clinical application. Finally, low plasma myostatin levels may reflect enhanced muscle catabolism or compensatory adaptation within the inflammatory–metabolic axis. In critically ill populations, lower circulating myostatin has been associated with systemic inflammation and unfavorable outcomes [3]. Experimental studies suggest that interleukin-6 may modulate myostatin expression via the JAK/STAT3 signaling pathway, providing a mechanistic link between inflammation and muscle atrophy. Although myostatin (GDF-8) and GDF-11 share substantial sequence homology, they are distinct gene products with different tissue distributions. Myostatin is primarily expressed in skeletal muscle and the central nervous system, whereas GDF-11 is more prominent in neural and cardiovascular tissues [4]. This difference may partly explain the stronger association observed between myostatin and PICS. Recent longitudinal studies further support a multimarker approach incorporating inflammatory, neuromuscular, and metabolic biomarkers [5]. Combining plasma myostatin with markers such as IL-6 and neurofilament light chain may therefore provide a more comprehensive framework for early identification of PICS. We appreciate the opportunity to respond to these comments and to further clarify the clinical and mechanistic implications of our findings. Sincerely, The Authors The authors thank the ICU staff of Kobe University Hospital for their support. This work was supported by JSPS KAKENHI Grant Number JP24K19491. Ethics approval was obtained from the Institutional Review Board for Clinical Research at Kobe University Hospital (Approval No. B210116). The authors declare no conflicts of interest.