Mean Muscle Attenuation within Paraspinal Muscle as a Determinant of Survival in Patients with Gastric Cancer Undergoing Gastrectomy

Background: This study aimed to evaluate the prognostic potential of skeletal muscle mass and muscle quality at the level of the third lumbar vertebra (L3) using computerized tomography (CT) images in patients with stage I–III gastric cancer (GC) who underwent curative gastric resection. Methods: Patients with stage I–III GC who underwent curative gastric resection between October 2006 and June 2014 were enrolled in this study. Demographic and clinical parameters were collected and analyzed. The muscle-related parameters (MRPs), such as the area of total abdominal wall musculature (SMA), area of paraspinal muscle (PMA), mean muscle attenuation (MA) within the total abdominal wall musculature (SMMA), and MA in the paraspinal muscle (PMMA), were measured at the L3 level using CT images. The Kaplan-Meier curve analyses and Cox proportional hazards model were applied to the MRPs, demographic, and clinical parameters to explore the overall survival (OS) and disease-free survival (DFS). Results: Overall, data from 339 patients (233 men and 116 women) were analyzed in this study. In the multivariate Cox model, the elderly (hazard ratio [HR] 2.08, 95% condence interval [CI] 1.30–3.32, p=0.002), total gastrectomy (HR 2.14, 95% CI 1.32–3.48, p=0.002), stage (HR 3.41, 95% CI 2.10–5.54, p<0.001), perineural invasion (HR 2.01, 95% CI 1.02–3.93, p=0.042), prognostic nutritional index (PNI) (HR 0.91, 95% CI 0.88–0.95, p<0.001), and PMMA (HR 2.36, 95% CI 1.47–3.78), p<0.001) were prognostic factors for OS. Similarly, the elderly (HR 1.68, 95% CI 1.09–2.59, p=0.018), total gastrectomy (HR 2.16, 95% CI 1.37–3.38, p<0.001, stage (HR 4.16, 95% CI 2.64–6.54, p<0.001), PNI (HR 0.91, 95% CI 0.87–0.94, p<0.001), and PMMA (HR 2.22, 95% CI 1.42–3.47, p<0.001) were prognostic factors for DFS. Conclusions: PMMA is suggested as a determinant of OS and DFS in stage I through III GC patients who underwent gastrectomy. Because PMMA is a newly characterized parameter in GC, external validation before clinical applications is a prerequisite. nature of the variables. Variables without well-known cutoff points were analyzed as continuous variables without an intentional dichotomy. Disease-free survival (DFS) was dened as the interval from the date of GC surgery to the date of recurrence or death from any cause, whichever came rst. Overall survival (OS) was dened as the interval from the date of GC surgery to the date of death from any cause. Patients who did not experience cancer recurrence or death from any cause were censored at the last follow-up. NLR: PLR: platelet-to-lymphocyte ratio; LMR: lymphocyte-to-monocyte ratio; PNI: prognostic nutritional index; SMI: skeletal muscle index; SMMA: mean attenuation within total abdominal wall musculature; PMI: paraspinal muscle index; PMMA: mean attenuation within paraspinal muscle.

microscopically margin-negative resection (i.e., R0 resection) [34]. Since there are no validated thresholds for the CT measurements of muscle quantity, more research is required. Therefore, the purpose of the current study was to evaluate the prognostic potential of the four MRPs (i.e., SMI, SMMA, PMI, and PMMA) using preoperative CT images at the L3 level in a cohort of stage I-III GC patients who underwent R0 resection.

Patients
Patients who underwent gastric resection for GC between June 2006 and October 2014 at a single university hospital were evaluated retrospectively. The inclusion criteria were as follows: (i) primary GC, according to Lauren's histological classi cation [35]; (ii) stage I through III, according to the 7th edition of the American Joint Committee (AJCC) on cancer staging [36]; (iii) underwent R0 resection; and (iv) underwent gastrectomy by a gastroenterology expert who participates in more than 50 GC resections annually.
The exclusion criteria were: (i) previous malignancies within the last 5 years, or concurrent second malignancies; (ii) received neoadjuvant chemotherapy, radiation therapy, or other anti-cancer treatment prior to surgery; (iii) stage IV; (iv) absence of R0 resection; (v) history of positive tests for human immunode ciency virus, severe infections within 4 weeks prior to surgery, or active autoimmune diseases that require systemic or immunosuppressive agents; (vi) chronic kidney disease (stage 4 or 5); (vii) chronic obstructive pulmonary disease; (viii) underwent surgical interventions for lumbar spine disorders [37]; (ix) without Korean resident registration number; and (x) without a preoperative abdominal CT scan that can be analyzed.

The baseline clinical characteristics
Records of demographic (e.g., age, sex, height, body weight, and body mass index [BMI]), and pathological parameters (e.g., tumor location, tumor size, type of gastrectomy, stages, Lauren's histological classi cation, and lymphatic, vascular, and perineural invasion) were collected and analyzed.
Selected blood tests included peripheral blood leukocyte count, absolute neutrophil count (ANC), absolute monocyte count (AMC), absolute lymphocyte count (ALC), hemoglobin concentration, platelet count, and serum albumin level. Blood test results were analyzed through tests performed within 7 days before surgery. If more than one test result was available, the test result closest to the date of surgery was selected for further analysis. The diagnosis of anemia was based on hemoglobin concentrations below 13 g/dL in men and 12 g/dL in women. Blood test results were also used to calculate NLR, PLR, LMR, and prognostic nutritional index (PNI). The details of the calculation are as follows: NLR = (ANC/ALC) × 100 [38]; PLR = (platelet counts/ALC) ×100 [38]; LMR = (ALC/AMC) ×100 [39]; and PNI = 10 × serum albumin level (g/L) + 0.005 × ALC [40].
The muscle-related parameters Only CT images taken as part of a routine staging work-up within 30 days prior to surgery were analyzed. After identi cation of the landmark at the L3 level, the corresponding single, enhanced axial CT image was extracted by a musculoskeletal radiologist and was saved as a DICOM image le [41]. The slice thickness in this study was 5 mm. The image was analyzed by a trained nurse using ImageJ ® v.1.37, a Java-based image analysis program, which went through an inspection process by a physician who is an expert in the musculoskeletal system.
For measuring the SMA, total abdominal wall musculature was selected as the region of interest (ROI). As for the PMA, a group of muscles, such as the erector spinae, multi dus, quadratus lumborum, and the psoas, were selected as ROIs (Fig. 1). The SMA and PMA were adjusted by the square of height, which produced the SMI and PMI, respectively [42]. In this study, as the muscle mass was evaluated using CT without considering body functions, the term sarcopenia was not applied to mean low SMI or PMI to avoid confusion. To measure the SMMA and PMMA, a Houns eld unit (HU) threshold range of − 29 to + 150 was applied.

Ethical considerations
This study protocol was approved by the Institutional Review Board of Kyung Hee University Hospital at Gangdong (IRB: 2019-12-039). Written consent was waived because this study was a retrospective study and the lack of it does not negatively affect the rights and well-being of the subject. This study was conducted in accordance with the ethical principles of the Declaration of Helsinki for medical research.

Statistical analysis
Clinical features were described as medians and interquartile ranges. Either the Mann-Whitney U test or the Chi-squared test was appropriately used for intergroup comparisons, depending on the nature of the variables. Variables without well-known cutoff points were analyzed as continuous variables without an intentional dichotomy. Disease-free survival (DFS) was de ned as the interval from the date of GC surgery to the date of recurrence or death from any cause, whichever came rst. Overall survival (OS) was de ned as the interval from the date of GC surgery to the date of death from any cause. Patients who did not experience cancer recurrence or death from any cause were censored at the last follow-up.
The survival rate was estimated using the Kaplan-Meier method, and the statistical signi cance between survival curves was tested using the log rank test. Statistical signi cance was declared at p <0.05. In addition, the Cox proportional hazards model was used to calculate hazard ratios, which was performed only on variables that met the proportional hazards assumption on the basis of graphic plots of Schoenfeld residuals. Only variables with p <0.05 in the univariate analysis were included in the multivariate Cox model. All statistical analyses were performed using R packages. To measure the model's discriminative capacity, Harrell's concordance index (C-index) for the Cox model was performed.
For the diagnosis of multicollinearity of the covariates, the variance in ation factors (VIF) was calculated.

Demographic and clinical characteristics of patients
The baseline demographic and clinical characteristics of the patients are presented in Table 1. The median age of the patients was 60 years, with more males (65.8%) than females. There were 207 (61.1%) patients in stage I, 66 (19.5%) in stage II, and 66 (19.5%) in stage III. Sixty-eight patients (20.1%) underwent total gastrectomy. The intestinal type, according to Lauren's classi cation, was the most common at 51.9%, and perineural invasion was found in 24 (7.1%) patients. Anemia was diagnosed in 122 (36.0%) patients, while 21 (6.2%) had hypoalbuminemia with serum albumin levels less than 3.5 g/dL.

Impact of muscle-related parameters on the Kaplan-Meier curves
Because there were signi cant differences in the medians of the four MRPs (i.e., SMI, SMMA, PMI, and PMMA) between sexes (p<0.001 in all variables), they were dichotomized for each sex, using thresholds determined using the receiver operating characteristic curve analysis ( Table 2). The number of patients below the SMI and SMMA thresholds were 143 (42.2%) and 111 (32.7%), respectively. In addition, the number of patients below the PMI and PMMA thresholds were 178 (52.5%) and 137 (40.4%).
When evaluating the four MRPs using the Kaplan-Meier method, there were signi cant differences in OS in variables, such as the SMMA (p<0.001), PMI (p=0.027), and PMMA (p<0.001). The ve-year OS rates were 70.2% in the low SMMA group and 85.5% in the high SMMA group. In addition, the ve-year OS rates were 71.5% in the low PMMA group and 86.6% in the high PMMA group. As for DFS, there were signi cant differences in DFS in variables, such as the SMMA (p<0.001) and PMMA (p<0.001). The veyear DFS rates were 64.8% in the low SMMA group and 83.7% in the high SMMA group. In addition, the ve-year DFS rates were 66.3% in the low PMMA group and 85.1% in the high PMMA group (Fig. 2 (Table 4).

Discussion
The purpose of the current study was to evaluate the clinical signi cance of the four MRPs (i.e., SMI, SMMA, PMI, and PMMA) at the L3 level in stage I through III GC patients who underwent R0 resection. In this study, using a multivariate Cox model, the PMMA was the only MRP that proved to be a signi cant determinant of OS and DFS.
In this study, the SMA at the L3 level was correlated with height, as in Yoshizumi's previous study. [43]. Therefore, we calculated the SMI, the SMA normalized for height squared, for use as an indicator of muscle mass. Although the most frequent cutoffs of SMI were 52-55 cm 2 /m 2 for men and 39-41 cm 2 /m 2 for women in the literature [19], we determined the cutoffs based on our patient cohort. As the SMI was sex-dependent, it was dichotomized as low and high groups with sex-speci c cutoffs (i.e., 46 [29]. Thus, the results of the current study were somewhat inconsistent with previous studies that found that the SMI is an important determinant of OS in GC patients who underwent gastrectomy [22][23][24][25][26]. However, in a recent study by Hacker et al., encompassing 761 advanced stage GC patients, the SMI was also not a prognostic factor for OS [31]. In addition, in Zhuang et al.'s study, encompassing 937 stage I GC patients, it was not associated with OS [25]. The reason for the discrepancies between studies is unclear, but the differences in cutoffs, pathologic stage, ethnicity, and variation in CT acquisition parameters (e.g., type of CT examination, slice thickness, and administration of contrast media) and image analysis software may have possibly caused these [19]. Because only a limited number of studies have been performed on CT-based SMI measurements in GC patients, further studies are needed to validate the prognostic role of the SMI.
In addition to the SMA, we measured the PMA at the L3 level as another marker of muscle mass. In our study, the group of muscles, including the erector spinae, multi dus, psoas, and quadratus lumborum, was de ned as the paraspinal muscle, as had been reported previously [29,34,44,45]. Since the PMA correlated with height, we calculated the PMI, the PMA normalized for height squared for statistical analysis. As PMI was sex-dependent, PMI was dichotomized as low and high groups with sex-speci c  [31]. Therefore, our ndings are in line with previous studies in terms of DFS, but not OS. The reason for the inconsistent ndings between studies is not certain; however, the differences in de nition of paraspinal muscles (i.e., with or without the psoas muscle), cutoffs, stage of the tumor (i.e., stage I to III vs. stage IV), tumor location, and ethnicity (i.e., Asian vs. international) may have affected the differences in outcomes between studies.
Fat in the skeletal muscles is present in the form of intermuscular adipose tissue, intramuscular adipose tissue, or intramyocellular lipids [46]. The MA, which is calculated using CT images, is a radiologic index of muscle fat content; the values of the mean MA are inversely correlated with muscle fat content [27]. Under physiological conditions, the values of the MA are considered dependent on sex, age, and ethnicity [47]. Excess fat deposition in skeletal muscle, a pathological variation in the MA, is observed in patients with malignant tumors [27].
In previous studies on gastrointestinal malignancies, the SMMA was usually measured within the total abdominal wall musculature at the L3 level with prede ned HU ranges of -29 HU to +150 HU. Therefore, in our study, the SMMA was dichotomized using sex-speci c cutoffs (i.e., 40.56 HU for men and 26.39 HU for women), and the adjustment for age was performed by including age as a covariate at entry [48]. In this study, while the SMMA was found to be an important determinant of OS and DFS in the univariate Cox model, but not in the multivariate model. Similarly, in Tamandl et al.'s study, the SMMA was a determinant of OS in patients with resected esophageal or gastroesophageal junction cancer by applying the univariate Cox model, but was no longer true in the multivariate Cox model; therefore, their ndings are consistent with our ndings [30]. On the other hand, in other studies, using the multivariate Cox model, the SMMA was a signi cant prognostic factor for determining OS in patients with GC who underwent gastrectomy, patients with various gastrointestinal malignancies, and patients with gastric or gastroesophageal junction cancer. Therefore, the results of these studies were inconsistent with our ndings [28,29,31].
In our study, the PMMA was measured in a group of skeletal muscles (i.e., erector spinae, multi dus, quadratus lumborum, and psoas muscle) at the L3 level with prede ned HU ranges of -29 HU to +150 HU. Before statistical evaluation, this was dichotomized with sex-speci c cutoffs (48.12 HU for men and 34.24 HU for women), and the adjustment for age was performed by including age as a covariate at entry.
Using the multivariate Cox model, the PMMA was a signi cant determinant of both OS and DFS. The results of this study were consistent with our group's previous study that enrolled only stage I and II GC without including SMMA as a covariate [34]. Similarly, in Dohzono's study of patients with gastrointestinal cancers, including GC at 25%, the PMMA was a signi cant determinant of survival. However, discrepancies in terms of the cutoffs, paraspinal muscle de nition, tumor stage, and tumor histology limit the exact comparison between studies [33].
As a subgroup analysis, the multivariate Cox model with or without the PMMA was applied. When applying multivariate Cox analysis while excluding the PMMA as a covariate in this study, the SMMA was an independent determinant of OS (p=0.008) and DFS (p=0.008  [31]. On the other hand, in studies by Zhuang [28] and Martin [29], both the SMI and SMMA were signi cant determinants of OS. Therefore, the results of these studies were inconsistent with our nding, which suggests that muscle quality may be a stronger determinant than muscle mass. Since studies on MA, including PMMA along with SMMA, are rare, more studies are needed to verify the clinical signi cance of MA as a survival determinant.
In our study, multivariate Cox analysis revealed that total gastrectomy was an important determinant of OS and DFS. Compared with subtotal gastrectomy, total gastrectomy is a technically di cult surgery associated with poor prognosis due to high postoperative complications, infection rates, and hospitalization rates [49,50]. Therefore, our ndings con rmed the current trend. In this study, we also found that perineural invasion was an independent determinant of survival by applying a multivariate Cox model. The prognostic value of perineural invasion has been reported previously [51]. In the present study, the PNI, which is calculated based on serum albumin levels and peripheral ALC, was an important determinant of OS and DFS in the multivariate Cox model [40]. The clinical role of PNI as an indicator of nutritional and immunological status in cancer patients has been validated. Several previous studies on GC have con rmed that the PNI is associated with the depth of invasion, lymph node metastasis, pathological stage, and lymph and vascular in ltration and that low PNI was an adverse prognosticator for survival [51][52][53].
The strengths of our study are as follows: First, this was the rst study to assess the prognostic signi cance of the PMMA, along with the SMMA as a covariate, in patients with stage I to III GC undergoing curative gastrectomy. In this study, we found that the PMMA was the only MRP capable of predicting OS and DFS by applying a multivariate Cox model. Thus, skeletal muscle quality (i.e., PMMA) compared to skeletal muscle mass (i.e., SMI or PMI) appears to be a more important determinant of survival. Second, for MRP measurements, we used CT images taken as a part of standard cancer staging to avoid additional costs and radiation exposure. Finally, for consistency, we included only those patients who underwent curative gastrectomy by a gastroenterology expert.
However, since this study had several limitations, the results of the study should be interpreted carefully. First, this study was performed retrospectively; therefore, omission of data including CT images is inevitable and may have affected the results. Second, although random errors and potential biases were controlled from the study design to implementation, the lack of external validation was another limitation of our study. For veri cation, public databases were investigated in domestic and global databases, but there were no CT images suitable for measuring the MRPs. However, based on the results of this study, it is possible to conduct a prospective study with an independent external validation group at the next stage of this study, which could be an essential step for the clinical use of PMMA. Third, since it was a retrospective study, there was no opportunity to provide special interventions to improve postoperative outcomes in patients with PMMA below the threshold. Finally, the cutoffs for dichotomization of MRPs in our cohort were determined using receiver operating characteristic curve analysis. As such, the cutoffs presented in this study may not be applicable to patients with malignant tumors other than GC.
In conclusion, we found that the PMMA was a signi cant determinant of OS and DFS. Since the PMMA is a newly characterized determinant of survival in GC, its prognostic importance requires further validation prior to clinical application. In addition, measurements of muscle mass and MA using CT require standardization for comparison between studies. . * The cutoff points are 12 g/dL in female patients and 13 g/dL in male patients.   *** The cutoff points are 12 g/dL in female patients and 13 g/dL in male patients.