Serum interferon-gamma-induced protein 10 levels can help predict sarcopenia development in patients with primary hepatocellular carcinoma: A retrospective cohort study.

Sarcopenia is characterized by a low skeletal muscle mass, weakness, and decreased overall physical performance [ 1 ]. Based on the evaluation criteria developed by the Japan Society of Hepatology (JSH) for sarcopenia in liver disease cases, patients with chronic liver disease exhibiting low grip strength and muscle mass, as determined by computed tomography (CT) or bioelectrical impedance analysis, are considered to have sarcopenia [ 2 ]. Recently, there has been a growing emphasis on evaluating sarcopenia in terms of muscle quality, especially focusing on decreased grip strength or increased intramuscular fat mass [ 3 – 6 ]. Sarcopenia reportedly is a poor prognostic factor in patients with hepatocellular carcinoma (HCC) [ 7 – 11 ]. Thus, sarcopenia has recently gained more attention in those with chronic liver disease. However, the development of sarcopenia in patients with HCC is often associated with primary sarcopenia due to age and secondary sarcopenia, making it difficult to understand the pathogenesis in some cases. Several theories about sarcopenia development, besides primary sarcopenia, have been proposed; however, the precise mechanism in patients with HCC still requires elucidation. The decreased levels of amino acids and testosterone, hyperammonemia, hypermetabolic state, and suppression of the mammalian target of rapamycin pathway due to lack of mobility and systemic treatment are all thought to contribute to development of sarcopenia [ 12 ]. Chemokine interferon-gamma (IFN-γ)-induced protein 10/C-X-C motif chemokine ligand 10 (IP-10), a downstream molecule of IFN-γ, may potentially be involved in the mechanism [ 13 , 14 ]. IP-10 levels were reported to promote the proliferation and differentiation of satellite cells by signaling through the receptor CXCR3. Intramuscular recombinant IP-10 treatment in aged mice induces the proliferation of satellite cells and provokes the increase in the number of regenerated myofibers [ 15 ]. Contrarily, the reduction in muscle atrophy via down-regulation of IP-10 levels has been reported in tumor-bearing mice [ 16 ]. The serum IP-10 levels correlated with the clinical degree of muscle involvement in patients with systemic sclerosis, suggesting that high IP-10 levels do not usually indicate increased muscle regeneration [ 17 ]. The IP-10 levels may produce an opposite effect on muscle regeneration depending on age, presence of chronic disease, presence of cancer, and so on. Therefore, in the present study, we aimed to examine the association between serum IP-10 levels and sarcopenia development in patients with HCC.

The median age of the patients (n = 120) was 71 years (range: 51–87 years), and 73 patients were male. The Child–Pugh scores were 5, 6, and 7 in 86, 22, and 12 cases, respectively. Regarding the Tumor-Node-Metastasis (TNM) staging, 52 and 68 patients were categorized as having Stage 1 and 2, respectively. Forty patients did not experience sarcopenia (Non-Sarco group), 40 had sarcopenia after 3 years (Sarco-develop group), and 40 patients were assigned to the Sarco-base group. A summary of the features of these patients is shown in Table 1 . The IP-10 levels only showed a very weak correlation with Albumin-bilirubin (ALBI) score, Child-Pugh score, Branched-chain amino acids (BCAA), Branched-chain amino acids/tyrosine ratio (BTR), platelet counts, alpha-fetoprotein (AFP), and tumor size, with no alternative indicators ( Table 2 ). The baseline IP-10 levels were significantly lower in the Sarco-base group compared to the rest (88 vs. 110 pg/ml, p = 0.016) ( Fig 1a ). No association was found between the presence of sarcopenia at baseline and IP-10 ratio (0.98 vs. 1.0, p = 0.81) ( Fig 1b ). Contrarily, IP-10 levels at 1 year after the confirmed diagnosis of HCC were lower in the Non-Sarco group compared to the rest (25 vs. 62 pg/ml, p < 0.001) ( Fig 1c ). The IP-10 ratio after 1 year was also lower in the Non-Sarco group compared to the rest (0.91 vs. 1.1, p = 0.044) ( Fig 1d ). (a) The association between serum IP-10 levels and the presence of sarcopenia at baseline, (b) The association between IP-10 ratios and the presence of sarcopenia at baseline. (c) The association between serum IP-10 levels at 1 year after the first HCC occurrence and sarcopenia development. (d) The association between IP-10 ratios at 1 year after the first HCC occurrence and sarcopenia development. Further comparisons were performed between the Sarco-base, Sarco-develop, and Non-Sarco groups. The baseline IP-10 levels were higher in the Sarco-develop group than in the Sarco-base (p < 0.001) and Non-Sarco (p = 0.0017) groups ( Fig 2a ). The IP-10 ratios were higher in the Sarco-develop group than in the Non-Sarco group (p = 0.025) ( Fig 2b ). The outcomes of the comparative analysis of the patient characteristics among the three groups are summarized in Table 1 . (a) Baseline IP-10 levels among three groups. (b) IP-10 ratios among three groups. In all the patients, independently related factors for sarcopenia development were age > 68 years old, male, body mass index <24, with diabetes mellitus, high IP-10 levels at baseline, and high IP-10 ratios at 1 year ( Table 3 ). After propensity score matching using the factors; age, gender, body mass index, and diabetes mellitus, high IP-10 levels at baseline and high IP-10 ratios at 1 year were still independently associated factors for the development of sarcopenia ( Table 3 ). Patients without sarcopenia at baseline but who had high IP-10 levels at baseline or high IP-10 ratios after 1 year exhibited several distinguishing characteristics, which were as follows: high ALBI grade, low BCAA, low BTR, low platelet counts, more patients opting for transarterial chemoembolization (TACE) as treatment for primary HCC, and recurrence beyond up to seven criteria during the 3-year follow-up period ( Table 4 ).

Patients with sarcopenia at baseline were more likely to present with low IP-10 levels than those without. Conversely, those not presenting with sarcopenia at the first occurrence of HCC and subsequently developed sarcopenia after 3 years had higher baseline IP-10 levels and higher IP-10 ratios at 1 year. This indicates that the IP-10 levels may have different effects on the muscles depending on the timing after the confirmed diagnosis of HCC. Furthermore, monitoring the IP-10 levels may be an effective tool for defining high-risk groups in patients with primary HCC, with a predisposition to develop sarcopenia, which is a significant prognostic factor for patients with primary HCC.

From a total of 738 patients at our hospital with a confirmed diagnosis of primary HCC from January 2008 to January 2021, 239 patients with Barcelona Clinic Liver Cancer (BCLC) stage A who had satisfactory imaging at baseline and were aged ≥20 years were selected, whereas those without sufficient blood samples for IP-10 assay, those with missing data, those diagnosed with HCC other than BCLC stage A, or those with a shorter follow-up (<3 years) were excluded from the analysis. The presence of sarcopenia was further assessed at baseline and after 3 years, and patients were classified into the following three groups; patients with sarcopenia at baseline were classified as the Sarco-base group, patients who met the criteria for sarcopenia after 3 years were classified as the Sarco-develop group and patients who never met the criteria at follow-up were classified as the Non-Sarco group. In each group, 40 patients whose first visit date were closer to the present, i.e., with a short serum sample cryopreservation period were selected, and total 120 patients whose serum IP-10 levels were measured both at baseline and after 1 year, were enrolled in our research. We accessed their data for research purposes on 01/04/2024. HCC was diagnosed based on the outcomes of the pathological evaluation or the non-rim hyperenhancement in the arterial phase of dynamic CT or gadolinium ethoxybenzyl diethylenetriamine penta-acetic acid-contrast-enhanced magnetic resonance imaging (MRI) and non-peripheral washout or threshold increase, where only nodules that exhibited LR-4 and LR-5 using Liver Imaging Reporting and Data System were categorized as HCC [ 18 ]. All patients provided written informed consent prior to study participation, and the study was approved by the Human Ethics Review Committee of Yamanashi University Hospital (approval number: 1326), in accordance with the Declaration of Helsinki. According to the evaluation criteria for sarcopenia in liver disease established by JSH, the manifestation of decreased grip strength and low muscle mass are indicative of sarcopenia. Nonetheless, due to the retrospective nature of our research, adequate grip strength analysis was not possible. Hence, CT values were used for assessing skeletal muscle quality. The psoas muscle mass index (PMI) at the level of the third lumbar vertebra, as determined by CT imaging, was applied as an indicator of muscle mass volume. CT images obtained during the primary HCC diagnosis served as the baseline data. The cross-sectional areas of the bilateral psoas muscles were evaluated by manual tracing, and PMI was determined by normalizing these areas to a square of a patient’s height in meters. The cut-off value for PMI was set as 6.36 and 3.92 cm 2 /m 2 for men and women, respectively, as per the JSH criteria [ 19 ]. The CT values for the multifidus muscle at the third lumbar vertebral level were indicative of skeletal muscle quality [ 20 – 22 ]. The cut-off value for low CT values was established at 44.4 and 39.3 Hounsfield Unit (HU) for men and women, respectively [ 3 , 4 ]. The measurement of muscle mass volume and CT values was carried out by two hepatology specialists who were experts in this field. In our research, sarcopenia was outlined as the presence of both low PMI and CT values. Serum samples were gathered from 9 mL of blood obtained at baseline and 1 year after the confirmed diagnosis of HCC. These samples were distributed into aliquots and stored at −80 °C until further analysis. The serum IP-10 levels were measured using 50-μL of the stored serum and an enzyme-linked immunosorbent assay kit, as per the manufacturer’s instructions. The levels were determined using standard calibration curves and expressed in pg/mL. The IP-10 ratio was defined as the ratio of IP-10 levels obtained at baseline and 1 year after the confirmed diagnosis of HCC. All experimental data were expressed as medians (ranges). Between-group comparisons were conducted using the Mann–Whitney U-, Kruskal–Wallis, and Friedman tests along with nonparametric analysis of variance. If the one-way analysis of variance yielded significant results, differences between individual groups were analyzed with Fisher’s exact test. A p  < 0.05 was regarded as statistically significant. All the analyses were done with EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). Specifically, EZR is a modified version of the R commander developed to employ statistical functions commonly applied in biostatistics [ 23 ].