Clinical utility of psoas muscle volume in assessment of sarcopenia in patients with early-stage non-small cell lung cancer

Sarcopenia influences postoperative outcomes of patients with non-small cell lung cancer (NSCLC). Imaging tools for evaluating and diagnosing sarcopenia have developed, and a novel method of psoas volume index (PVI) obtained by measuring bilateral psoas major muscle volume has been reported. However, the relationship between sarcopenia based on PVI and clinical outcomes has not been fully investigated for patients with early-stage NSCLC. This study aimed to clarify the utility of PVI values in assessing the relationshipe between sarcopenia and clinical outcomes. This study included 645 patients with stage I-II NSCLC who underwent curative lung resection between 2012 and 2017. Bilateral psoas major muscle volumes were calculated semi-automatically using a three-dimensional workstation. The cutoff value of PVI for defining sarcopenia was < 60.5 cm3/m3 for men and < 43.6 cm3/m3 for women. The avrage time to obtaine PVI was only 25 s with the 3D system, and interobserver agreements for evauating sarcopenia on PVI was 1. A total of 159 patients (24.7%) were preoperatively diagnosed with sarcopenia. On multivariate analysis, sarcopenia was an independent prognostic factor for overall survival (OS, p < 0.001), recurrence-free survival (RFS, p < 0.001), and lung cancer-specific survival (LCS, p < 0.001). The 5-year OS, RFS, and LCS were significantly worse in sarcopenic patients than non-sarcopenic patients (88.8 vs. 72.4%, p < 0.001; 80.1 vs. 65.0%, p < 0.001; 92.4 vs. 78.9%, p < 0.001, respectively). Sarcopenia diagnosed using PVI is an independent prognostic predictor of OS, RFS, and LCS in early-stage NSCLC.


Introduction
Lung cancer is the leading cause of cancer-related death worldwide (Siegel et al. 2015). Surgical resection is the mainstay treatment for patients with early-stage non-small cell lung cancer (NSCLC). The formulation of strict guidelines on operative indications, advances in perioperative management, and the widespread use of minimally invasive surgery have resulted in marked improvements in surgical outcomes, even in older and high-risk patients (Ginsberg and Rubinstein 1995;Asamura et al. 2008;Sawabata et al. 2011). However, these patients often have reduced tolerance to surgical intervention. Thus, it is critical to consider factors, such as age, body mass index (BMI), performance status, smoking status, comorbidities, and the presence of sarcopenia, when determining the course of treatment (Leduc et al. 2017;Kanzaki et al. 2021;Kawaguchi et al. 2010;Kravchenko et al. 2015;Nakamura et al. 2018).
The relationship between sarcopenia and cancer outcomes has been gaining increased attention. Sarcopenia, which refers to the loss of skeletal muscle mass and physical function with physical disability, has been suggested as an independent predictor of postoperative complications and poor overall survival (OS) in patients with cancer, including those with lung cancer (Nakamura et al. 2018;Cruz-Jentoft et al. 1 3 2010; Joglekar et al. 2015;Shinohara et al. 2020). Measuring psoas area index (PAI) of the bilateral psoas muscle areas at the third lumbar vertebra (L3) level on computed tomography (CT) is widely used for sarcopenia evaluation (Hamaguchi et al. 2016). Recently, a novel method of measuring bilateral psoas major muscle volume on threedimensional (3D) workstations has also been reported Horie et al. 2021). A previous study reported that a diagnosis of sarcopenia based on the psoas volume index (PVI) indicates a risk of respiratory complications in patients with NSCLC . Horie et al. reported that sarcopenia assessed using bilateral psoas muscle volume is likely to be more accurate in predicting oncological outcomes than PAI in patients with rectal cancer who underwent neoadjuvant chemoradiotherapy (Horie et al. 2021). However, to our knowledge, there has been no study using PVI to examine the prognostic significance of sarcopenia in patients with early-stage NSCLC. Therefore, we aimed to clarify the relationship between sarcopenia diagnosed with the PVI value and clinical outcomes in patients with clinical stage I-II NSCLC.

Patients and follow-up
Patients who underwent pulmonary resection at our institution between July 2012 and December 2017 and were histopathologically diagnosed with primary NSCLC were included in the study. Exclusion criteria were clinical stages III-IV, incomplete surgical resection, and unavailability of CT scans obtained for the measurement of psoas major muscle volume. TNM stage was determined according to the 8th edition of the TNM classification of malignant tumors (Goldstraw et al. 2016). The follow-up schedule consisted of a clinic visit every 6 months in the first 2 years after resection and every 1 year subsequently. We aimed to continue follow-up for 10 years after resection. Follow-up evaluations included physical examination, chest radiography, and blood tests. Chest and abdominal CT scans were performed every 6 months in the first 2 years and every 1 year thereafter. Further evaluations, including brain magnetic resonance imaging and bone scintigraphy, were performed to rule out recurrence in cases of disturbing signs and symptoms. Positron emission tomography-CT scan (PET/CT) was performed when indicated. This study was approved by the Review Board of Tokyo Medical University (IRB No. T2020-0272) and was performed in line with the principles of the Declaration of Helsinki. Informed consent for the use and analysis of clinical data was obtained preoperatively for each patient.

Image analysis and assessment of sarcopenia
All preoperative PET-CT scans were obtained within 3 months before surgery. The abdominopelvic CT images were examined and the bilateral psoas major muscle area and volume were calculated (Fig. 1). We obtained the value of the psoas muscle area by adding the left and right psoas muscle areas measured at the level of the third lumbar vertebrae. The measurement of end-to-end dimensions of a muscle with a straight line semi-automatically brought the value using a 3D workstation (SYNAPSE VINCENT; Fujifilm Corp, Tokyo, Japan). One expert radiologist and two thoracic surgeons manually marked the superior edge of the liver and the superior margin of the symphysis pubis and then automatically obtained psoas muscle volume using the 3D software. The PAI and PVI were calculated in Fig. 1 a Bilateral psoas muscle (yellow green) areas at the third lumbar vertebra level, b bilateral psoas muscle volume was measured using a semi-automatic image analysis software with preoperative PET-CT. PET-CT positron emission tomography computed tomography accordance with the methods employed in previous studies as follows: PAI = bilateral psoas major muscle area at the L3 level on CT/ height 2 (cm 2 /m 2 ).
We determined the PVI cutoff value for sarcopenia as the lowest quartile by sex ; which was defined as < 60.5 cm 3 /m 3 for men and < 43.6 cm 3 /m 3 for women ( Supplementary Fig. S1).

Preoperative and Postoperative clinical factors
Data were collected on the following patient characteristics: age, sex, comorbidities (diabetes mellitus, chronic obstructive pulmonary disease, interstitial pneumonia, asthma, cerebrovascular disease, cardiovascular disease, arrhythmia, and autoimmune diseases), smoking history, preoperative BMI, spirometry test results, tumor size on CT imaging, surgical procedure, postoperative complications, histologic type, sarcopenia, and preoperative blood examination values within 1 month before surgery. In addition, the prognostic nutritional index (PNI) was determined from preoperative blood examination values and calculated as 10 × serum albumin (g/dL) + 0.005 × total lymphocyte count (cells/mm 3 ) (Shoji et al. 2016). Postoperative complications occurring within 90 days after surgery were assessed using the Clavien-Dindo classification system (Clavien et al. 2009). The complications studied included prolonged air leak, pneumonia, atelectasis, pleural effusion, empyema, chylothorax, atrial fibrillation, bradycardiac arrhythmia, pericardial effusion, acute myocardial infarction, and wound infection. A major complication was defined as one with a Clavien-Dindo grade of 2 or higher.

Statistical analysis
OS was measured from the date of surgery to the date of death due to any cause or date of the last follow-up, whichever came first. Recurrence-free survival (RFS) was measured from the date of surgery to the first date of recurrence, death from any cause, or date of the last follow-up, whichever came first. Lung cancer-specific survival (LCS) was measured from the date of surgery to the date of lung cancer death or date of the last follow-up, whichever came first. Non-lung cancer-specific survival (NCS) was measured from the date of surgery to the date of non-lung cancer death or date of the last follow-up, whichever came first. Non-lung cancer deaths were defined as deaths from any medical cause other than lung cancers. Kaplan-Meier curves were used to compare the OS, RFS, LCS, and NCS of patients with and without sarcopenia. Differences between the survival curves were estimated using the log-rank test. Univariate analysis of OS, RFS, and LCS was performed using the Cox proportional hazards regression model. Multivariate analysis was conducted using a backward stepwise variable selection for the Cox model. Variable with a threshold of p < 0.15 were adopted for the stepwise model procedure. The chisquare test and Mann-Whitney U test were used to compare groups and clinical factors for categorical and continuous data, respectively. Pearson's correlation coefficient was used to detrmine the correlation between PAI and PVI, and the value of the coefficient of detrmination (R 2 ) was provided. PAI and PVI measured by three observers were recorded, and interobserver variabliity using a random sample of 50 patients was caluculated for intraclass correlation coefficient (ICC). Statistical significance was set at p < 0.05. All analyses were performed using SPSS version 28 software (IBM Corporation, Armonk, New, USA), and graphics were constructed using R ver. 4.0.2 (R Project for Statistical Computing, Vienna, Austria).

Results
Of the 1231 patients with NSCLC who underwent operations, 645 patients were found eligible for this study. Patient characteristics are shown in Table 1. The median follow-up time for survivors was 1851 days (range: 545-3067 days). The study cohort consisted of 348 men and 297 women. The histological cancer types included adenocarcinoma (537 patients, 83.3%) and others (108 patients, 16.7%). A total of 159 patients (24.7%) were diagnosed with sarcopenia based on PVI preoperatively. There were significant differences in age (p < 0.001), BMI (p < 0.001), PNI (p = 0.007), lower forced expiratory volume 1 s/ forced vital capacity (FEV1.0%) (p < 0.001), comorbidities (p = 0.007), solid tumor size (p = 0.006), histology (p = 0.011), and the frequency of vascular invasion (p = 0.020) between the two cohorts. No significant difference was observed in the postoperative complications. The correlation analysis between PAI and PVI showed a strong correlation (R 2 = 0.75; p < 0.001; Supplementary Fig. S2). Interobserver agreements for evauating sarcopenia on PVI by the dedicated software (ICC = 1) were higher than that on PAI (ICC = 0.995, data not shown). Moreover, the avrage time to obtaine PVI (25 s) was significantly shorter than that for PAI (59 s, p < 0.001, data not shown).
We also assessed LCS and NCS according to the presence or absence of sarcopenia. Univariate and multivariate

Discussion
The main findings of this study were as follows: 1) sarcopenia on CT-based PVI was a better prognostic indicator for long-term outcomes than that on PAI; 2) PVI provides robust results in consistency and simplicity for evaluating sarcopenia comapred to PAI, and 3) age, BMI, PNI, FEV1.0%, presence of comorbidities, solid tumor size, non-adenocarcinoma, and vascular invasion were found to be significantly associated with preoperative sarcopenia. Various imaging techniques, such as dual-energy X-ray absorptiometry, magnetic resonance imaging, skeletal muscle ultrasound, and bioimpedance analysis, are used to measure the quantity of skeletal muscle mass and adipose tissue. Among these, PAI at the L3 level on cross-sectional CT  is widely used to diagnose sarcopenia or muscle wasting and predicts mortality in patients with various solid tumors (Nakamura et al. 2018;Joglekar et al. 2015;Shinohara et al. 2020;Williams et al. 2021;Sakamoto et al. 2020). However, only one slice is examined for PAI. The cross-sectional area of the psoas muscle widely varies along its length and differences in measurement position can cause inconcistency in its overall measured size. Therefore, specific technical skills and multiple attempts are required to accurately measure these indices. Although we can semi-automatically obtaine PAI by simply measuring end-to-end dimensions of a psoas muscle with a straight by the dedicated software, oue study suggests that volumetric measurements on a 3D scale were more accurate than single axial or cross-sectional CT measurements. As 3D workstations enable rapid and automatic acquisition of volumetric data of the psoas, we hypothesized that PVI is a simple, reliable, and reproducible measure and can be applied as a surrogate for total muscle mass. Preoperative sarcopenia was associated with OS, RFS, and LCS on multivariate analysis, and patients with sarcopenia had significantly higher grade malignant features, such as larger solid tumor size, non-adenocarcinoma, and the presence of vascular invasion. Kawaguchi et al. showed that patients with pathlogical stage I-III NSCLC and sarcopenia had larger tumors, more lymph nodes, and more extensive vascular invasion than patients without sarcopenia, which is consistent with our findings (Kawaguchi et al. 2021). Sarcopenia can be divided into ''primary'' or age-related sarcopenia when no other cause is evident other than aging itself, or ''secondary'' when one or more causes can be identified (Cruz-Jentoft et al. 2010;Williams et al. 2021). Diseaserelated, especially cancer-related, sarcopenia is a secondary sarcopenia, and many tumor-associated factors can affect the progression of sarcopenia in patients with advanced cancer (Williams et al. 2021;Sakamoto et al. 2020). These include advanced organ failure, chronic inflammation, cachexia, and immunosenescence. Given that cachexia is not usually seen in most patients with early-stage cancer, other factors may affect sarcopenic status and poor outcomes.
Skeletal muscle is known to be a secretory organ, and accumulating data have shown that muscle cells produce and secrete several hundreds of cytokines and other peptides, termed myokines, which influence various systemic responses (Pedersen and Febbraio 2012). Among numerous myokines, myostatin, a member of the transforming growth factor-β (TGF-β) superfamily, is a key regulator of skeletal muscle mass and is associated with the promotion of tumor progression. Hence, skeletal muscle loss impairs the immune reaction by increasing TGF-β levels, leading to cancer progression and recurrence (Neto et al. 2018;Lenk et al. 2010). Furthermore, skeletal muscle cells modulate immune function by signaling through myokines such as interleukin (IL)-15 in cell-to-cell interactions (Afzali et al. 2018). IL-15 modulates the proliferation, activation, and distribution of NK cells and CD8 T cells (Conlon et al. 2015). NK cells and CD8 T cells are necessary for the clearance of viral pathogens and destruction of tumor cells. The results of this study suggested that the impairment of anti-tumor immunity due to muscle degradation may occur from an earlier stage of a cancer's progression than expected, possibly affecting cancer recurrence and death. Skeletal muscle mass is under investigation as a risk factor for recurrence in renal cell carcinoma, malignant melanoma, colorectal cancer, hepatocellular carcinoma, pancreatic cancer, and NSCLC (Horie et al. 2021;Sakamoto et al. 2020;Kawaguchi et al. 2021;Noguchi et al. 2020;Liao et al. 2021;Youn et al. 2021). Elucidation of the etiology and molecular mechanisms of sarcopenia may hold the key to overcoming cancer progression and reducing recurrence. Further studies are required to confirm this association.
BMI that is more predictive of body fatness than weight alone does not truly reflect changes in body composition although many reports showed that sarcopenia was significantly related to lower BMI. Prado et al. demonstrated that some individuals showed high BMI and low muscle mass, indicating sarcopenic obesity, and those with sarcopenic obesity had reduced functional status, poor survival, and increased chemotherapeutic toxicity (Prado et al. 2008). Sarcopenic obesity in the elderly was reported to be associated with abnormalities in performance-based tests of balance and gait, aerobic capacity, strength, balance, and walking speed (Goisser et al. 2015). Therfore, it is essential to assess total body composition even in obese patients with lung cancer because a heavier weight may mask muscle degradation and sarcopenia.
This study had some limitations. First, this was a retrospective study with a limited number of patients from a single institution. Second, because our study included only patients who were Japanese, the results cannot be generalized to patients of all ethnicities. Third, the definition of sarcopenia used in our study may have influenced the results. Several parameters can be used to define sarcopenia, including PAI, PVI, skeletal muscle index, and thoracic skeletal muscle area (Hamaguchi et al. 2016;Lee et al. 2020;Kim et al. 2020;Madariaga et al. 2020). Fourth, a standard cutoff value for PVI has not been established because the PVI value is a novel assessment of sarcopenia.
In conclusion, the current study demonstrated that the PVI, as measured using a 3D workstation, could serve as a measure of frailty and therefore be useful for clinical outcomes, survival and recurrence risk stratification in patients with early-stage NSCLC.