Pancreatic cancer is the seventh leading cause of cancer-related deaths worldwide, and 80-85% of cases are unresectable at diagnosis (1,2). Surgical resection is the most effective treatment for pancreatic cancer. However, despite advancements in perioperative therapy and surgical techniques, the postoperative prognosis for pancreatic cancer remains poor. Several indicators, such as nutritional index and skeletal muscle mass index (SMI), have been reported as prognostic markers of pancreatic cancer (3-5). The pathophysiological mechanisms of SMI reduction are complex, with various causes, such as aging and chronic inflammation (6). Several reports have focused on SMI reduction before surgery and stated that it is a risk factor for poor prognosis (4,5). However, reports on the association between postoperative SMI reduction and prognosis of pancreatic cancer are few. We hypothesized that the postoperative SMI reduction rate may be a new predictor of long-term prognosis, especially after highly invasive surgeries, such as pancreatectomy. This study aimed to assess the postoperative SMI reduction rate for pancreatic cancer using computed tomography (CT) to determine whether this change affects overall (OS) and disease-free survival.

Patients and data collection. The data of 137 patients who underwent pancreatectomy for pancreatic cancer at the Department of Surgery, Hiroshima Red Cross & Atomic Bomb Survivors Hospital, Japan, between July 2005 and March 2022 were reviewed. Patients who did not undergo preoperative CT or died within 90 days after surgery were excluded. Preoperative and postoperative clinical data were obtained from electronic records and available for all patients. This study was approved by the ethics committee of our hospital according to the ethical guidelines of the Japanese government (approval number: 2023-030), and all patients provided consent for the research use of their clinical data.

Evaluation of skeletal muscle mass using CT images. Abdominal CT was used to examine skeletal muscle mass. The CT images used were either 1.25- or 5-mm thick contrast-enhanced or unenhanced multiphase acquisitions. Skeletal muscle borders were manually outlined (manual trace method) using a SYNAPSE VINCENT volume analyzer (Fujifilm, Tokyo, Japan). The masses of the rectus abdominis, abdominal lateral and oblique muscles, psoas, and paraspinal muscles were calculated. We measured the areas of skeletal muscle at the level of the third lumbar vertebral body (L3) on the CT images and normalized them by height in meters squared (m²) to calculate the lumbar skeletal muscle index (SMI, cm²/m²). All included patients underwent abdominal CT within two months before surgery, and the first follow-up CT scan was performed 3-9 months after surgery. The SMI reduction rate was calculated afterwards.

Statistical analyses. Comparisons of the categorical variables were performed using the χ² or Fisher’s exact tests. The independent-samples t-test or Mann–Whitney U-test was used to analyze continuous variables. Receiver operating characteristic (ROC) curve analysis of postoperative skeletal muscle loss was performed. The area under the ROC curve (AUC) was established to determine the optimal cut-off values for analyzing OS. Univariate and multivariate survival analyses were performed using Cox proportional hazard models. Cumulative OS and recurrence-free survival (RFS) rates were calculated using the Kaplan–Meier method, and differences between curves were evaluated using the log-rank test. OS was calculated as the time from the date of surgery to the date of last follow-up or death. RFS was defined as the time interval from the date of surgery to the date of disease recurrence. To identify postoperative prognostic factors, variables found to be significant in the univariate analyses were included in the overall multivariate Cox proportional model. All statistical tests were two-sided, and significance was set at p<0.05. Statistical analyses were performed using EZR software package 4.1.2 (Saitama Medical Center, Jichi Medical University, Japan).

Patient characteristics. Among the 137 consecutive patients who underwent curative pancreatic surgery between July 2005 and March 2022, 102 patients (54.9% men, mean age 69.1 years) met the inclusion criteria. The ROC curve showed that an SMI reduction rate cutoff of 14% predicts two-year mortality with a specificity of 48% and sensitivity of 92%; moreover, the AUC was 0.71 (Figure 1). A total of 76 patients exhibited an SMI reduction rate <14%, while 26 had SMI reduction rate ≥14%. Meanwhile, 61 patients (59.8%) were defined as having preoperative sarcopenia. The patient characteristics are shown in Table I. Among the 102 patients who underwent pancreatectomy at our institution, pancreatoduodenectomy (PD), distal pancreatectomy (DP), and total pancreatectomy (TP) were performed in 61 (59.8%), 39 (38.2%), and two patients (2.0%), respectively. The median amount of blood loss was 693 g (range=10-4,662 g). The median operative time was 396 min (range=160-695 min). Among the 102 patients, 71 (69.6%) were treated with complete resection (R0). Pathologically, poorly differentiated adenocarcinomas were observed in 33 patients (32.3%). Lymph node metastases were detected in 58 patients (56.9%). Neoadjuvant and adjuvant therapies were administered to 11 patients (10.8%) and 86 patients (84.3%), respectively. The medians of the preoperative inflammatory markers, including albumin, platelet count, and C-reactive protein were as follows: 4.00 g/dl (range=2.6-4.8 g/dl) g/dl, 213,900 μl (range=42,000-442,000 μl), and 0.43 mg/dl (range=0.05-9.27 mg/dl), respectively. SMI reduction rate ≥14% significantly correlated with advanced age (72 vs. 68 years; p=0.043) and higher incidence of postoperative complications (Clavien–Dindo Classification Grade III-V 34.6% vs. 10.5%; p=0.015).

Effects of postoperative SMI reduction on OS. The Kaplan–Meier OS curves for patients with high and low SMI reduction are shown in Figure 2, which revealed that OS (log- rank p<0.001) was significantly shorter in the SMI reduction rate ≥14% group than in the SMI reduction rate <14% group. The median OS rates in the SMI reduction rate ≥14% and SMI reduction rate <14% groups were 1.60 years [95% confidence interval (CI)=0.96-2.24] and 3.94 years (95%CI=3.28-4.61), respectively. Univariate and multivariate analyses were used to identify factors influencing OS following pancreatectomy (Table II). Univariate analysis revealed the following factors as significantly associated with poor OS: blood loss ≥500 ml [hazard ratio (HR)=1.58, 95%CI=1.00-2.54; p=0.048), lymph node metastases-positive (HR=1.72, 95%CI=1.08-2.74; p=0.024), tumor size ≥3.0 cm (HR=1.90, 95%CI=1.21-2.99; p=0.006), preoperative albumin <3.5 (HR=2.16, 95%CI=1.16-4.03; p=0.016), preoperative prognostic nutritional index (PNI) <40 (HR=4.91, 95%CI=2.24-10.78; p<0.001), and SMI reduction rate ≥14% (HR=3.23, 95%CI=1.95-5.34; p<0.001). Multivariate analysis identified the following factors as independently correlated with poor OS: preoperative PNI <40 (HR=3.93, 95%CI=1.11-13.93; p=0.034) and SMI reduction rate ≥14% (HR=2.68, 95%CI=1.57-4.59; p<0.001).

Effects of postoperative loss of SMI on RFS. The Kaplan–Meier RFS curves for patients with high and low SMI reduction are shown in Figure 3, which reveals that RFS (log- rank p<0.001) were significantly shorter in the SMI reduction rate ≥14% group than in the SMI reduction rate <14% group. The median RFS rates in the SMI reduction rate ≥14% and SMI reduction rate <14% groups were 0.95 (95%CI=0.41-1.48) and 2.50 years (95%CI=1.99-3.02), respectively. Univariate and multivariate analyses were used to identify factors influencing RFS following pancreatectomy (Table III). Univariate analysis revealed the following factors as significantly associated with poor RFS: blood loss ≥500 ml (HR=1.65, 95%CI=1.05-2.58; p=0.029), lymph node metastases-positive (HR=1.69, 95%CI=1.08-2.63; p=0.021), tumor size ≥3.0 cm (HR=2.12, 95%CI=1.37-3.27; p<0.001), preoperative albumin <.5 (HR=1.95, 95%CI=1.04-3.63; p=0.036), preoperative neutrophile-lymphocyte ratio (NLR) ≥3.0 (HR=1.86, 95%CI=1.09-3.16; p<0.022), preoperative PNI <40 (HR=3.23, 95%CI=1.44-7.23; p<0.004), and SMI reduction rate ≥14% (HR=2.61, 95%CI=1.58-4.31; p<0.001). Multivariate analysis identified the following factors as independently correlated with poor RFS: tumor size ≥3.0 cm (HR=1.68, 95%CI=1.02-2.75; p=0.040), preoperative NLR ≥3.0 (HR=1.97, 95%CI=1.13-3.44; p=0.017), and SMI reduction rate ≥14% (HR=2.82, 95%CI=1.64-4.85; p<0.001).

Regardless of cancer type, SMI reduction in the postoperative period occurs easily because of physical inactivity, malnutrition, and hypercatabolism (6). The correlation between skeletal muscle mass and prognosis has been recognized in various cancers (7). This study suggests that postoperative SMI reduction negatively affects the prognosis of patients undergoing pancreatectomy for pancreatic cancer.

Despite advancements in surgical techniques, pancreatic cancer surgery remains associated with a high risk of complications and poor nutrition, which are associated with increased morbidity and mortality (8). In addition, as pancreatic surgery often alters the gastrointestinal anatomy, muscle loss after surgery for pancreatic cancer is thought to be greater than that in other types of cancers (9). Thus, prevention and management of complications, malnutrition, and hypercatabolism are crucial for postoperative pancreatic cancer treatment. Most previous studies have focused on the preoperative loss of skeletal muscle mass (10). However, whether the loss of postoperative skeletal muscle mass affects the prognosis of pancreatic cancer is unclear.

SMI is a widely used parameter to evaluate muscle mass and calculated by dividing the cross-sectional area of all skeletal muscles at the L3 vertebral level, which is obtained by axial CT scan, by the square of the height (m²) (11). It is significantly correlated with the whole-body skeletal muscle tissue (12). Thus, the assessment of skeletal muscle mass using CT scans at the L3 vertebral level is increasingly used and has become a research hotspot in the examination of its correlation with the prognosis of various diseases (10). In this study, SMI changes were calculated using CT scans before surgery and approximately 3-9 months after surgery in patients with pancreatic cancer. Our data revealed a SMI reduction rate ≥14% after surgery was associated with poor clinical outcomes.

In our study, the SMI reduction rate significantly correlated with advanced age and a higher incidence of postoperative complications. Age-related loss of skeletal muscle mass is caused by decrease in motor nerve activity, impairment in neuromuscular junctions, and reduction in hormone production (13). Low preoperative SMI was also correlated with postoperative complications in patients with pancreatic cancer (14,15). Previous studies have suggested that SMI reduction is associated with increases in the concentration of circulating inflammatory cytokines and with high complication rates (16,17).

Multivariate analyses revealed that low PNI and SMI reduction rate ≥14% were independent factors for OS, while large tumor size, high NLR, and SMI reduction rate ≥14% were independent factors for RFS. Although the strong link between SMI reduction and prognosis remains obscure, we hypothesize that it is associated with multiple factors, including poor caloric intake and physical activity. The PNI is an independent indicator of postoperative outcomes, and cytokine levels are correlated with tumor size and NLR in pancreatic cancer (3,18,19). We have reported that the postoperative NLR level represents the levels of circulating cytokines and is an independent predictor of survival after pancreatectomy (20). The skeletal muscle contains a large number of leukocytes, which play critical roles in antitumor immunity and are the same as those in total blood (4,21). These findings indicate that high cytokine levels and loss of antitumor immunity would lead to cancer progression, resulting in a shorter RFS. Therefore, we suspect that the combination of tumor aggressiveness and patient factors, such as nutritional and physical conditions, affect the prognosis of patients with pancreatic cancer. To improve the prognosis after surgery of pancreatic cancer, prophylactic strategies, including nutritional support and physiotherapy, are important to prevent skeletal muscle loss. According to some studies, enteral nutrition is preferred over total parenteral nutrition (TPN) as postoperative nutritional support (22). TPN is associated with a longer period of first bowel movement, weight loss, and increased complications. Although oral feeding is the best way to provide nutrition postoperatively, patients often need to abstain from food because of complications or delayed gastric emptying and require alternative nutrition therapy. Jejunostomy tube (J-tube) placement is a safe method commonly used for early enteral feeding after pancreatic surgery. However, there were no differences in complications or prognosis between patients with and without J-tube feeding after pancreaticoduodenectomy (23). Thus, nutritional support therapy after pancreatic surgery remains a challenging and important issue. Some nutrients have been shown to be associated with muscle mass and complication rates. The early postoperative arginine-rich enteral nutrition could improve nitrogen balance and suppress weight loss in patients after abdominal surgery (24). The arginine-rich nutrition also contributes wound healing and increases hormone secretion with protein synthesis. In animal models, it could suppress bacterial translocation in the intestinal lumen (25). Another report indicates vitamin D is associated with muscle mass, and low vitamin D level is one of the risk factors of postoperative complications (26). The blood testing for such nutritional deficiencies would be important to reduce muscle mass loss and complication after surgery.

Postoperative complications of pancreatic cancer were significantly reduced not only by postoperative rehabilitation, but also by preoperative rehabilitation called prehabilitation (27,28). Home prehabilitation has proven to reduce muscle mass loss in the early postoperative period (29). In addition, the personalized home-based prehabilitation can reduce the rate of postoperative complications after abdominal surgery (30). Enhanced recovery of patient care pathways (ERAS) is becoming the standard of care for surgery (31). According to the ERAS 2019 recommendations, avoidance of postoperative nasogastric tubes, early diet, early mobilization, and near-zero fluid balance to avoid excessive fluid administration are encouraged to improve patient outcomes (32,33). A multi-disciplinary team approach for perioperative management, such as nutritional and physical therapy, is becoming increasingly important as the number of patients with sarcopenia increases with population aging. In our case, preoperative sarcopenia was identified in 60% of patients. Despite previously published reports, sarcopenia was not found to be associated with prognosis (34,35). This might be because more patients were sarcopenic in our study than in previous reports, and there was no statistically significant difference between the SMI reduction rate and preoperative sarcopenia.

Study limitations. This retrospective, small-cohort study was conducted at a single center using only CT image analysis. However, the strength of this study is that we used the objective index of CT to evaluate postoperative SMI changes. In the future, large-scale investigations using different analytical methods should be conducted.

SMI reduction rate after pancreatic surgery is an independent prognostic factor for survival in patients with pancreatic cancer. We believe that nutritional support and physiotherapy focused on preventing skeletal muscle loss require further development.

The Authors declare that they have no competing interests in relation to this study.

JM designed the study and wrote the initial draft of the article. ET and TM contributed to interpretation of the data and critical revision of the article for important intellectual content. All other Authors (NH, RM, YO, AS, ST, RM, TO, TK, SY, KK, KH) contributed to data collection and interpretation and critically reviewed the article.

The Authors would like to thank Editage (www.editage.com) for English language editing.

None of the Authors received funding regarding this study.