Skip to main content
NCBI home page
BMC Gastroenterology logoLink to BMC Gastroenterology
. 2025 Dec 10;26:26. doi: 10.1186/s12876-025-04530-6

Changes of bone, adipose, and muscle-related body compositions in gastric cancers after gastrectomy using deep learning based automatic segmentation

Mengying Xu 1,#, Dan Liu 2,#, Mengze Zhang 3, Song Liu 1,✉, Zhengyang Zhou 1,✉
PMCID: PMC12801571  PMID: 41372761

Abstract

Background

To investigate bone, adipose, and muscle-related body compositions changes in gastric cancers (GCs) 12 months after gastrectomy utilizing an artificial intelligence (AI) based segmentation tool and to conduct subgroup analyses based on clinicopathological characteristics.

Methods

This retrospective study included 146 GCs who underwent gastrectomy. Body compositions of GCs at baseline and 12 months after surgery were automatically measured utilizing the AI-based segmentation tool. The differences in body compositions at baseline and 12 months after surgery were assessed. Subgroup analyses of body composition changes stratified by clinicopathological characteristics were conducted. Benjamini-Hochberg false discovery rate (FDR) correction was utilized for all subgroup analyses.

Results

All body composition parameters, including bone mineral density (BMD), adipose tissue, and muscle, decreased significantly 12 months after surgery (all p < 0.001). The greatest losses were observed in adipose-related compositions. The proportions of sarcopenia (from 37.7% to 55.5%) and osteoporosis (from 13.7% to 26.7%) showed significant increases. Subcutaneous and visceral adipose tissues, abdominal wall muscle, and skeletal muscle losses indicated significant differences in gender and body mass index (BMI) subgroups (all FDR p < 0.05). Adipose- and muscle-related losses differed significantly in GCs underwent different types of gastrectomy (all FDR p < 0.05), while BMD loss showed no significant difference. Subgroup analyses based on pathological stages showed no significant difference for body composition changes.

Conclusions

BMD, adipose-related, and muscle-related body compositions showed significant losses over 12 months in GCs underwent gastrectomy. The greatest losses were observed in adipose-related compositions. The proportions of sarcopenia and osteoporosis showed significant increases. Gender, baseline BMI, and gastrectomy differences affected adipose- and muscle-related body composition losses but showed no significant effect on BMD.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12876-025-04530-6.

Keywords: Stomach neoplasms; Bone mineral density; Body composition; Tomography, x-ray computed; Artificial intelligence

Introduction

Gastric cancer (GC) is the fifth most common cancer and one of the leading causes of cancer-related deaths globally [1]. Most patients with GC presented with advanced stages when initially diagnosed [2, 3]. Surgical resection remains the recommended curative treatment for locally advanced resectable GCs [4]. Increasing studies have demonstrated that the nutritional status of patients with GC, including body composition components, was associated with perioperative complications, tumor recurrence, and clinical outcomes [5–8]. These studies highlighted the important role of body composition analysis.

Due to anatomical and functional changes caused by gastrectomy, most patients with GC underwent body composition changes postoperatively, including bone mineral density (BMD), adipose-related, and muscle-related components [5, 7, 9, 10]. Recent studies have also reported that the changes in body composition components in GCs after surgery were related to long-term quality of life and prognosis [7–11]. Tao et al. found that BMD and visceral adipose tissue (VAT) changes were independent predictive factors for overall survival (OS) in GCs after surgery [7]. Therefore, comprehensive measurements of body composition components and their changes might indirectly reflect the body’s potential metabolism and nutritional status, which could benefit for individualized early intervention in GCs after surgery.

In clinical practice, BMD is commonly determined by dual-energy X-ray absorptiometry and quantitative computed tomography (CT). Part of the studies utilized routine CT images for BMD measurement [12, 13]. Adipose- and muscle-related body composition can be evaluated by dual-energy X-ray absorptiometry, CT images, bioelectrical impedance analysis, and magnetic resonance imaging [7, 8, 14–16]. However, CT examination is the most commonly utilized imaging tool for the diagnosis, staging, and postoperative follow-up in GCs [17–21]. Therefore, if the measurement of the above body composition components could be performed on routine CT images, the additional burden may not be added in patients with GC. In recent years, the application of automatic segmentation of body composition components based on CT images increased gradually with the rapid development of artificial intelligence (AI) algorithms [22–24]. It can facilitate the evaluation of body composition components and reduce the input of clinicians. To our knowledge, comprehensive analysis of BMD, adipose-related, and muscle-related body composition changes based on AI automated segmentation in GCs after surgery, as well as their distributions in clinicopathological variables, have not been well-studied.

This study aimed to evaluate the nutritional and metabolic changes in GCs 12 months after gastrectomy by automatic segmentation of BMD, adipose tissue, and muscle on CT images utilizing an AI-based body composition analysis tool. Additionally, subgroup analyses of body composition changes were further performed based on clinicopathological characteristics.

Methods

Patient

The retrospective study was approved by the Ethical Committee of our institution, the requirement for obtaining informed consent was waived because of the retrospective nature. From July 2022 to November 2023, 178 patients with GC confirmed by surgical pathology at our institution were consecutively enrolled by searching the electronic medical record systems. Patients were included in the analysis with the following criteria: (1) pathologically confirmed diagnosis of GC by gastrectomy; (2) qualified abdominal CT examinations performed before and one year after surgery. Patients were excluded with any of the following criteria: (1) prior endoscopic submucosal dissection or stomach surgery (n = 1); (2) received neoadjuvant therapy previously (n = 23); (3) accompanied with other malignant tumors (n = 5); (4) a history of abdomen surgery (n = 1); (5) lumbar modic changes that affect the measurement of BMD (n = 2). Finally, 146 patients with GC (male, 92; female, 54) were included in this study. Detailed information of sample size evaluation is listed in Supplemental Material. The flowchart with detailed patient selection is shown in Figure A1. Demographic and clinicopathological information, including gender, age, body mass index (BMI), comorbidities (hypertension and diabetes), approaches of gastrectomy, T stage, and N stage, were retrospectively collected.

CT image acquisition

CT examinations were performed within two weeks before surgery and one year after surgery. Detailed CT scan protocols are listed in Supplemental Material.

Automated body composition segmentation and measurement

Body composition of all GCs at baseline and 12 months after surgery, including BMD, adipose-related, and muscle-related tissues, were automatically segmented utilizing an AI-based segmentation tool (uAI-Discover PLBMD). The body composition analysis tool is an AI-powered software solution based on a convolutional neural network called VB Net [25, 26]. The AI-based tool can automatically segment key tissues, such as VAT, subcutaneous adipose tissue (SAT), abdominal wall muscle (AWM), and paravertebral muscle (PM) using axial CT images at the central levels of L1-L4 vertebrae, and then provide quantitative statistics of tissue areas. In this study, non-contrast thin-slice CT images of all patients were loaded into the software. Subsequently, body composition parameters, including BMD, SAT, VAT, intermuscular adipose tissue (IMAT) in AWM, IMAT in PM, total IMAT, AWM, PM, and skeletal muscle (SM), were automatically segmented and measured at the L1-L4 levels (Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

Illustration of automated segmentation of body composition utilizing an artificial intelligence-based body composition analysis tool. a-b Automated measurement of bone mineral density at the L1 and L2 levels. c-d Automated segmentation of adipose- and muscle-related body composition. AWM, abdominal wall muscle; BMD, bone mineral density; CT, computed tomography; IMAT, intermuscular adipose tissue; PM, paravertebral muscle; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue

Then, the mean value of BMD at the L1-L2 levels was used as the final BMD value for further analysis, which is the generally applied method for BMD measurement in multi-slice CT scans [8, 27]. According to the American College of Radiology (ACR) practice guideline for bone densitometry, all patients were divided into three groups, including osteoporosis (BMD < 80 mg/cm3), low bone mass (80 mg/cm3 ≤ BMD ≤ 120 mg/cm3), and normal (BMD >120 mg/cm3) [27]. For adipose- and muscle-related body composition, the mean areas of regions of interest at the L1-L4 levels were calculated for further analysis. The SM index (SMI) was normalized by SM area for height squared (m2). Then, CT-determined sarcopenia was defined as SMI < 40.8 cm2/m2 for males and < 34.9 cm2/m2 for females, which is commonly utilized in Asian patients with gastrointestinal tumors [28, 29]. In addition, ΔBMD, ΔSAT, ΔVAT, ΔIMAT in AWM, ΔIMAT in PM, Δtotal IMAT, ΔAWM, ΔPM, and ΔSM of all patients were calculated and recorded.

To evaluate the accuracy and reliability of the AI-based segmentation tool for adipose, and muscle-related body compositions measurement on our patient cohort, a radiologist with five years of abdominal diagnostic experience manually segmented the body compositions. The manually segmentation was performed on L3-level CT images on the basis of predefined Hounsfield unit (HU) thresholds [15]. Firstly, preoperative unenhanced thin-slice CT images of 30 selected cases according to the time of surgery were loaded into an open-source software (3D Slicer, version 5.2.2; http://slicer.org). The detailed radiodensity criteria for manual segmentation of different body composition areas were as follows: an attenuation range of − 29 to + 150 HU was defined for SM, − 150 to − 30 HU for IMAT and VAT, and − 190 to − 50 HU for SAT [15]. After manual segmentation for the subset of images, the interobserver reproducibility between manual segmentation and the AI tool was evaluated via the intraclass correlation coefficient (ICC). In addition, detailed validation of the AI-based segmentation tool on BMD measurement are listed in Supplemental Material.

Statistical analysis

Categorical variables were presented as numbers (percentages). Continuous variables were presented as medians (interquartile ranges [IQRs]) after applying the Kolmogorov-Smirnov test for normality analysis. The differences in body composition parameters at baseline and 12 months after surgery were assessed using the Wilcoxon matched-pairs signed-ranks test. The effect sizes were reported as r for Wilcoxon matched-pairs signed-ranks test [30]. The differences in categorical variables were analyzed using the chi-square or Fisher’s exact test (n < 5). Subgroup analyses were compared with the Mann-Whitney U test or the Kruskal-Wallis test. All subgroup analyses were corrected by applying Benjamini-Hochberg false discovery rate (FDR) [31]. All statistical analyses were performed using SPSS (version 25.0). A two-sided p value < 0.05 indicates statistically significant.

Results

Patient characteristics

A total of 146 patients with GC who underwent surgery were enrolled in this study. There were 92 males (63.0%) and 54 females (37.0%) with a median age of 60 years (age range, 29–84 years). Among them, 12 (8.2%) patients underwent proximal gastrectomy, 74 (50.7%) patients underwent distal gastrectomy, and 60 (41.1%) patients underwent total gastrectomy. For all patients performed with 12-month CT examinations after surgery, the median interval between the date of the CT examination and the date of one year after surgery was 12 (IQR: 5–21) days. Table 1 summarizes the baseline demographic and clinicopathological characteristics of all patients.

Table 1.

Baseline demographic and clinicopathologic characteristics of included patients

Characteristics Male (n = 92) Female (n = 54) p value
Age, y 0.003*
 < 60 36 (39.1%) 35 (64.8%)
 ≥ 60 56 (60.9%) 19 (35.2%)
BMI, kg/m2 0.613
 < 18.5 1 (1.1%) 2 (3.7%)
 18.5–24 53 (57.6%) 32 (59.3%)
 ≥ 24 38 (41.3%) 20 (37.0%)
Hypertension 0.315
 Present 59 (64.1%) 39 (72.2%)
 Absent 33 (35.9%) 15 (27.8%)
Diabetes 0.109
 Present 79 (85.9%) 51 (94.4%)
 Absent 13 (14.1%) 3 (5.6%)
Approaches of gastrectomy 0.228
 Proximal 10 (10.9%) 2 (3.7%)
 Distal 43 (46.7%) 31 (57.4%)
 Total 39 (42.4%) 21 (38.9%)
T stage 0.787
 1 36 (39.1%) 24 (44.4%)
 2 10 (10.9%) 6 (11.1%)
 3 33 (35.9%) 15 (27.8%)
 4 13 (14.1%) 9 (16.7%)
N stage 0.445
 0 40 (43.5%) 20 (37.0%)
 1–3 52 (56.5%) 34 (63.0%)
Open in a new tab

Data in parentheses are presented with percentages; BMI, body mass index

*p < 0.05 with chi-square test or Fisher’s exact test (n < 5)

Baseline measurements of body composition and changes over 12 months

Bone-related body composition

Baseline and 12 months after surgery measurements of body composition parameters utilizing the AI-based body composition analysis tool are listed in Table 2. Significant BMD loss was observed in patients with GC over 12 months after surgery (p < 0.001, Fig. 2) with a large effect size of −0.569. The medians (IQRs) of BMD from 115.93 (87.74, 140.63) mg/cm³ decreased to 105.63 (78.28, 133.70) mg/cm³. In addition, the grade for bone densitometry based on ACR practice guideline differed significantly between before and 12 months after gastrectomy (p < 0.001, Fig. 3), and GCs with osteoporosis after surgery increased significantly (20/146, 13.7% to 39/146, 26.7%).

Table 2.

Baseline and 12-month after surgery measurements of body composition in patients with gastric cancer

Parameters Baseline measurement 12-month measurement 12-month % change p value
Bone-related body composition (L1-2)
 BMD, mg/cm³ 115.93 (87.74, 140.63) 105.63 (78.28, 133.70) −7.30 (−15.43, 0.64) < 0.001*
ACR practice guideline for bone densitometry < 0.001*
 Normal 69 (47.3%) 54 (37.0%) NA
 Low bone mass 57 (39.0%) 53 (36.3%) NA
 Osteoporosis 20 (13.7%) 39 (26.7%) NA
Adipose-related body composition (L1-4)
 SAT, cm2 97.36 (76.29, 136.05) 54.68 (35.55, 79.03) −44.56 (−60.89, −20.65) < 0.001*
 VAT, cm2 101.05 (60.04, 151.08) 23.64 (12.33, 44.87) −71.17 (−85.50, −51.00) < 0.001*
 IMAT in AWM, cm2 8.58 (5.84, 10.43) 2.53 (1.13, 5.03) −64.59 (−81.30, −35.42) < 0.001*
 IMAT in PM, cm2 3.21 (2.14, 5.54) 2.10 (1.00, 3.36) −35.46 (−56.12, −5.40) < 0.001*
 Total IMAT, cm2 11.65 (8.21, 15.11) 4.94 (2.58, 8.73) −55.03 (−72.52, −27.02) < 0.001*
Muscle-related body composition (L1-4)
 AWM, cm2 53.15 (42.26, 60.51) 47.11 (38.93, 54.71) −8.95 (−13.85, −2.77) < 0.001*
 PM, cm2 60.41 (52.74, 71.71) 57.81 (48.98, 67.81) −5.98 (−10.07, −0.78) < 0.001*
 SM, cm2 113.53 (97.38, 132.63) 104.71 (86.45, 121.53) −6.65 (−11.63, −2.65) < 0.001*
CT-determined sarcopenia < 0.001*
 Absent 91 (62.3%) 65 (44.5%) NA
 Present 55 (37.7%) 81 (55.5%) NA
Open in a new tab

Categorical variables are presented as numbers (percentages)

Quantitative variables are presented as median (25th and 75th percentiles)

ACR American College of Radiology, AWM abdominal wall muscle, BMD bone mineral density, IMAT intermuscular adipose tissue, PM paravertebral muscle, SAT subcutaneous adipose tissue, SM skeletal muscle, SMI skeletal muscle index, VAT visceral adipose tissue, NA not applicable

*p < 0.05 with Wilcoxon matched-pairs signed-ranks test in quantitative variables and with chi-square test or Fisher’s exact test (n < 5) in categorical variables

Fig. 2.

Fig. 2

Open in a new tab

Wilcoxon matched-pairs signed-ranks test for bone mineral density (a), adipose-related (b-f), and muscle-related (g-i) body composition changes in patients with gastric cancer at baseline (red box) and 12 months after surgery (blue box). AWM, abdominal wall muscle; BMD, bone mineral density; IMAT, intermuscular adipose tissue; PM, paravertebral muscle; SAT, subcutaneous adipose tissue; SM, skeletal muscle; VAT, visceral adipose tissue

Fig. 3.

Fig. 3

Open in a new tab

Variations of the proportions of sarcopenia and osteoporosis in patients with gastric cancer at baseline and 12 months after surgery. The proportion of osteoporosis (pink area) increased from 13.7% to 26.7% (a). The proportion of sarcopenia (pink area) increased from 37.7% to 55.5% (b)

Adipose-related body composition

Table 2 summarizes adipose-related body composition based on the AI-automated segmentation tool at baseline and 12 months after surgery in GCs. Medians (IQRs) were 97.36 (76.29, 136.05) cm2 for baseline SAT, 101.05 (60.04, 151.08) cm2 for baseline VAT, 8.58 (5.84, 10.43) cm2 for baseline IMAT in AWM, 3.21 (2.14, 5.54) cm2 for baseline IMAT in PM, and 11.65 (8.21, 15.11) cm2 for baseline total IMAT. At 12 months after gastrectomy, SAT decreased to 54.68 (35.55, 79.03) cm2, VAT decreased to 23.64 (12.33, 44.87) cm2, IMAT in AWM decreased to 2.53 (1.13, 5.03) cm2, IMAT in PM decreased to 2.10 (1.00, 3.36) cm2, and total IMAT decreased to 4.94 (2.58, 8.73) cm2. Statistical analyses showed that SAT and VAT decreased significantly over 12 months after surgery (both p < 0.001, Fig. 2) with large effect sizes of −0.833 and − 0.866, respectively. The IMAT in AWM, IMAT in PM, and total IMAT also showed significant losses (all p < 0.001) with effect sizes of −0.858, −0.646, and − 0.846, respectively.

Muscle-related body composition

Significant muscle-related body composition losses, including AWM, PM, and SM, were found in patients with GC at 12 months after surgery (all p < 0.001, Table 2) with large effect sizes of −0.696, −0.636, and − 0.720, respectively. Medians (IQRs) were 53.15 (42.26, 60.51) cm2 for baseline AWM, 60.41 (52.74, 71.71) cm2 for baseline PM, and 113.53 (97.38, 132.63) cm2 for baseline SM. At 12 months after gastrectomy, AWM decreased to 47.11 (38.93, 54.71) cm2, PM decreased to 57.81 (48.98, 67.81) cm2, and SM decreased to 104.71 (86.45, 121.53) cm2. Medians (IQRs) for the loss % over 12 months in AWM (−9.0%, [−13.9%, −2.8%]) was the highest, followed by SM (−6.7%, [−11.6%, −2.7%]), and PM (−6.0%, [−10.1%, −0.8%]). In addition, CT-determined sarcopenia differed significantly between before and 12 months after gastrectomy (p < 0.001), and patients with sarcopenia over 12 months increased significantly (55/146, 37.7% to 81/146, 55.5%).

Subgroup analysis of body composition changes over 12 months

Bone-related body composition changes

Table A1 shows statistical descriptions of ΔBMD in subgroups stratified by clinicopathological characteristics. ΔBMD in patients with GC between before and 12 months after surgery indicated no significant difference in subgroups, including gender, age, BMI, approaches of gastrectomy, T stage, and N stage (all FDR p > 0.05). Subgroup analysis based on approaches of gastrectomy showed that BMD loss in GCs who underwent distal gastrectomy (−9.05, IQR: [−20.11, 0.15]) was the highest, followed by total gastrectomy (−7.63, IQR: [−16.23, 1.88]) and proximal gastrectomy (−2.53, IQR: [−17.15, 5.35]).

Adipose-related body composition changes

Table A2 lists the results of subgroup analysis in adipose-related body composition changes over 12 months stratified by clinicopathological characteristics. Significant SAT loss was observed in female patients, while male patients presented with significant VAT loss (both FDR p < 0.05, Fig. 4). However, ΔIMAT in AWM, ΔIMAT in PM, and Δtotal IMAT showed no significant difference in gender subgroup (all FDR p > 0.05). As for subgroup analysis of baseline BMI, ΔSAT and ΔVAT in GCs with baseline BMI (≥ 24 kg/m2) were the highest, followed by baseline BMI (18.5–24 kg/m2) and baseline BMI (< 18.5 kg/m2), with all losses statistically significant (FDR p < 0.05, Figs. 4 and Table A2). However, ΔIMAT in AWM, ΔIMAT in PM, and Δtotal IMAT demonstrated no significant difference between baseline BMI subgroup (FDR p = 0.051, 0.714, and 0.150, Table A2).

Fig. 4.

Fig. 4

Open in a new tab

Scatter plots for subgroup analyses of adipose- and muscle-related body composition losses based on clinicopathological characteristics. a-d Subgroup analysis based on gender. The brown scatters represent male patients, while the blue scatters represent female patients. e-h Subgroup analysis based on baseline body mass index (BMI). The brown, blue, and purple scatters represent patients with baseline BMI < 18.5 kg/m2, 18.5–24 kg/m2, and ≥ 24 kg/m2, respectively. i-l Subgroup analysis based on approaches of gastrectomy. The brown, blue, and purple scatters represent patients underwent proximal, distal, and total gastrectomy, respectively. AWM, abdominal wall muscle; BMI, body mass index; IMAT, intermuscular adipose tissue; SAT, subcutaneous adipose tissue; SM, skeletal muscle; VAT, visceral adipose tissue

Subgroup analysis based on approaches of gastrectomy showed that SAT, VAT, IMAT in AWM, IMAT in PM, and total IMAT losses in GCs who underwent proximal gastrectomy were the largest, followed by total gastrectomy and distal gastrectomy (all FDR p < 0.05). However, there was no significant difference in ΔSAT, ΔVAT, ΔIMAT in AWM, ΔIMAT in PM, or Δtotal IMAT between subgroups of age, T stage, or N stage (all FDR p > 0.05).

Muscle-related body composition changes

Subgroup analysis in muscle-related body composition changes over 12 months stratified by clinicopathological characteristics is summarized in Table A3. Significant AWM and SM losses were found in male patients (FDR p = 0.006 and 0.036). ΔAWM and ΔSM in GCs with baseline BMI (≥ 24 kg/m2) were the highest, followed by baseline BMI (18.5–24 kg/m2) and baseline BMI (< 18.5 kg/m2) (FDR p = 0.015 and 0.034). However, ΔPM showed no significant difference in gender and baseline BMI subgroups (FDR p > 0.05, Table A3).

Subgroup analysis based on approaches of gastrectomy showed that AWM, PM, and SM losses in GCs who underwent proximal gastrectomy were the largest, followed by total gastrectomy and distal gastrectomy (all FDR p < 0.05). It demonstrated similar trends with adipose-related body composition losses. However, there was no significant difference in ΔAWM, ΔPM, or ΔSM between subgroups of age, T stage, or N stage (all FDR p > 0.05).

Interobserver agreement assessment

Our results showed that body compositions, including SM, SAT, VAT, and IMAT-related parameters demonstrated excellent interobserver agreement (ICC = 0.831–0.998, Table A4). It indicated the excellent accuracy and reliability of AI-based segmentation tool in our study cohort. In addition, detailed validation results of the AI-based segmentation tool on BMD measurement are listed in Supplemental Material.

Discussion

As a systemic and complex disease, cancers were reported to interact with metabolism and immune macro-environment [32]. Many studies indicated that the physical status of GCs, including nutrition, sarcopenia, and osteoporosis, affected the long-term quality of life and OS [6–10]. Our study is the first comprehensive study to evaluate the postoperative changes of BMD, adipose-related, and muscle-related body composition in GCs based on AI-automated segmentation tool. In addition, subgroup analyses based on clinicopathological characteristics were further conducted in our study, which might be helpful for individualized management and nutritional support strategies.

Compared with body weight and BMI, body composition measurements could reflect the body’s nutritional status more comprehensively and objectively. Our study found that BMD, adipose-related, and muscle-related body composition parameters decreased significantly 12 months after surgery in patients with GC. The greatest losses were observed in adipose-related compositions, followed by BMD and muscle-related body compositions. Recently, the important role of adipose tissue in coordinating body metabolism and endocrine has been increasingly recognized [33–35]. VAT and SAT are two main locations of body adipose tissue, and VAT is indicated to present with relatively higher metabolic activity as well as inflammatory reaction compared with SAT [36, 37]. In this study, VAT losses over 12 months in GCs reached 71.2%, which contained the effect of omentectomy during gastrectomy. SAT losses over 12 months were approximately 50.0%, though less than VAT. Many previous studies demonstrated that adipose-related body compositions were associated with recurrence and prognosis in many malignant tumors [15, 38, 39]. Song et al. found that VAT change after surgery was the independent risk factor for recurrence-free survival in advanced GCs [5]. Therefore, measurements of SAT and VAT changes could indirectly reflect the body’s macroscopic metabolism and immune status, which could benefit for individualized early intervention in GCs after surgery. As for subgroup analyses, significant SAT loss was observed in female patients, while male patients presented with significant VAT loss. The gender differences in adipose tissue losses may be related to the distribution differences of body fat between man and female [40]. Therefore, postoperative monitoring of VAT and SAT based on gender differences might be helpful for reflecting individualized body’s nutritional status. In addition, GCs with baseline BMI (≥ 24 kg/m2) or underwent proximal gastrectomy showed the most significant losses of both VAT and SAT. It emphasized the need for early nutritional support and metabolic surveillance for GCs with higher baseline BMI or underwent proximal gastrectomy.

Skeletal muscle health involves two different aspects, including muscle mass and muscle quality [39]. Sarcopenia, characterized by decreased muscle mass, is one of the features of chronic disease, which can be assessed by CT images. It has been well-studied that sarcopenia is associated with worse prognosis and OS in many malignancies [29, 41, 42]. Our study showed significant muscle mass losses, especially AWM, in patients with GC at 12 months after surgery. The proportions of patients with sarcopenia increased from 37.7% to 55.5%. In addition, larger SM and AWM losses were detected in male patients or patients with higher baseline BMI, while PM changes showed no significant difference between the subgroups. Accordingly, appropriately supplementing nutrients for protein metabolism in muscles and appropriately increasing physical activity in male patients or patients with higher baseline BMI might help reduce muscle losses. Previous studies analyzing muscle-related body composition mainly focused on the areas of SM or PM [29, 41, 43]. For GCs with gender or baseline BMI differences, detailed evaluations of skeletal muscle located in different regions might help for optimal nutritional support and surveillance. Myosteatosis, defined as muscular fat deposition, is characterized by decreased muscle quality and dysfunction and is associated with poor prognosis in many malignant tumors [28, 39]. Our study found that IMAT-related body composition also decreased significantly in patients with GC at 12 months after surgery. Additionally, muscle- and IMAT-related body composition changes demonstrated with similar trends with adipose-related body composition losses in GCs who underwent different types of gastrectomy. It indicated that postoperative dietary management and nutritional monitoring should be paid more attention in GCs underwent proximal gastrectomy.

Osteoporosis is one of the complications in GCs after surgery, which impacts the long-term quality of life postoperatively. Our study suggested significant BMD losses in GCs at 12 months after surgery, with the proportion of osteoporosis assessed according to ACR guideline increasing from 13.7% to 26.7%. Hence, early monitoring of BMD and serum markers such as calcium might be beneficial for early intervention and thus improving bone metabolism for GCs. Subgroup analysis showed that BMD loss in female patients was slightly higher in male patients, with no significant difference. Moreover, the types of gastrectomy were not associated with BMD losses, which was consistent with previous studies [9]. It indicated that BMD losses were related to the gastrectomy itself. Subgroup analyses based on pathological stages of GCs showed no significant difference for BMD, adipose-related, and muscle-related body composition changes, suggesting that the changes in body composition might not be associated with the tumor lesion itself.

In clinical practice, some potential factors, including postoperative diet differences, physical activity levels, and medication use such as chemotherapy, could influence postoperative body composition changes. The procedure of gastrectomy in GCs induces significant anatomical and functional changes, which would lead to postoperative reduced food intake, low gastric acid levels, and decreased intestinal absorption [44–46]. Those changes could induce postoperative low body weight, reduced muscle mass, and bone mineral loss [9, 45, 47]. Accordingly, administration differences of postoperative diet could influence body composition status [9, 48, 49]. As for physical activity levels, Sakurai et al. indicated that low physical activity after surgery was an independent risk factor for BMD losses 12 months after gastrectomy [50]. In addition, poor physical activity in postoperative GC patients might lead to declined muscle mass and function. Previous studies also indicated thar postoperative treatments such as adjuvant chemotherapy might affect the nutritional status and bone metabolism of GC patients, thereby leading to body composition changes [9, 44, 51].

Our study had some limitations. First, the single-center retrospective design may limit the generalizability of our results to other patient populations and healthcare settings. Our results need further verification in future multi-center studies and other patient populations. Second, body composition changes in GCs after surgery may be affected by potential factors, including postoperative diet differences, physical activity levels, and medication use such as chemotherapy. Due to the retrospective nature of our study, it was difficult to control these potential factors. Third, the manual measurement of BMD on CT is commonly based on quantitative CT in clinical practice. In this retrospective study, all CT images were performed on conventional CT scanners. The manual segmentation of BMD for evaluating the reliability of AI-based segmentation tool was limited. Therefore, our findings need further verification in future multi-center studies. Forth, preoperative and postoperative CT examinations were not performed with the same scanner because of the retrospective design of this study. Due to the differences in CT scanners and scanning protocols, body composition measurements might show potential variability. Therefore, our results need to be validated in future multi-center studies with standardized scanning protocols and image processing methods. Fifth, our study evaluated the changes in body composition in GCs at 12 months after surgery for early monitoring, and the long-term changes in body composition need further studies.

In conclusion, BMD, adipose-related, and muscle-related body compositions showed significant losses over 12 months in GCs underwent gastrectomy. In addition, the proportions of sarcopenia and osteoporosis in patients with GC increased significantly. Gender, baseline BMI, and gastrectomy differences affected adipose- and muscle-related body composition losses but showed no significant effect on BMD. These findings evaluated by the AI-based body composition analysis tool could help establish individualized management and nutritional support strategies in clinical practice. 

Supplementary Information

Supplementary Material 1. (396.9KB, docx)

Acknowledgements

None.

Abbreviations

AI

Artificial intelligence

ACR

American College of Radiology

AWM

Abdominal wall muscle

BMD

Bone mineral density

BMI

Body mass index

CT

Computed tomography

GC

Gastric cancer

IMAT

Intermuscular adipose tissue

IQR

Interquartile range

PM

Paravertebral muscle

SAT

Subcutaneous adipose tissue

SM

Skeletal muscle

SMI

Skeletal muscle index

VAT

Visceral adipose tissue

Authors’ contributions

Conceptualization: SL and ZYZ; Data curation: MYX and DL; Formal analysis: MYX and DL; Methodology and Project administration: SL and ZYZ; software: MZZ; Resources and Supervision: MYX and SL; Visualization: MYX and SL; Original draft: MYX, SL, and ZYZ; Review & editing: all authors. All authors have read and agreed to the published version of the manuscript.

Funding

None.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This retrospective study was conducted according to the principles in the Declaration of Helsinki and was approved by the Ethical Committee of Nanjing Drum Tower Hospital (No. 2024-081-02). The requirement for obtaining informed consent was waived because of its retrospective nature.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Abbreviations

AI, Artificial intelligence; ACR, American College of Radiology; AWM, Abdominal wall muscle; BMD, Bone mineral density; BMI, Body mass index; CT, Computed tomography; GC, Gastric cancer; IMAT, Intermuscular adipose tissue; IQR, Interquartile range; PM, Paravertebral muscle; SAT, Subcutaneous adipose tissue; SM, Skeletal muscle; SMI, Skeletal muscle index; VAT, Visceral adipose tissue.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Mengying Xu and Dan Liu contributed equally to this work.

Contributor Information

Song Liu, Email: liusongnj@126.com.

Zhengyang Zhou, Email: zyzhou@nju.edu.cn.

References

  • 1.Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74(3):229–63. 10.3322/caac.21834. [DOI] [PubMed] [Google Scholar]
  • 2.Smyth EC, Nilsson M, Grabsch HI, van Grieken NC, Lordick F. Gastric cancer. Lancet. 2020;396(10251):635–48. 10.1016/S0140-6736(20)31288-5. [DOI] [PubMed] [Google Scholar]
  • 3.Chen X, Xu H, Chen X, Xu T, Tian Y, Wang D, et al. First-line camrelizumab (a PD-1 inhibitor) plus apatinib (an VEGFR-2 inhibitor) and chemotherapy for advanced gastric cancer (SPACE): a phase 1 study. Signal Transduct Target Ther. 2024;9(1):73. 10.1038/s41392-024-01773-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Ajani JA, D’Amico TA, Almhanna K, Bentrem DJ, Chao J, Das P et al. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology. Gastric Cancer, Version 1.2019. Accessed 14 May, 2019. Available at: https://www.nccn.org/professionals/physician_gls/PDF/gastric.pdf [DOI] [PubMed]
  • 5.Song GJ, Ahn H, Son MW, Yun JH, Lee MS, Lee SM. Adipose tissue quantification improves the prognostic value of GLIM criteria in advanced gastric cancer patients. Nutrients. 2024;16(5):728. 10.3390/nu16050728. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Kim M, Lee CM, Kang BK, Ha TK, Choi YY, Lee SJ. Sarcopenia assessed with DXA and CT increases the risk of perioperative complications in patients with gastrectomy. Eur Radiol. 2023;33(7):5150–8. 10.1007/s00330-023-09401-w. [DOI] [PubMed] [Google Scholar]
  • 7.Tao C, Hong W, Yin P, Wu S, Fan L, Lei Z, et al. Nomogram based on body composition and prognostic nutritional index predicts survival after curative resection of gastric cancer. Acad Radiol. 2024;31(5):1940–9. 10.1016/j.acra.2023.10.057. [DOI] [PubMed] [Google Scholar]
  • 8.Wei L, Fu B, Bo J, Jia H, Sun M, Jiang X, et al. Development of a nomogram based on body composition analysis of quantitative computed tomography combined with clinical prognostic factors to predict disease-free survival after surgery and adjuvant chemotherapy in patients with gastric cancer. Quant Imaging Med Surg. 2023;13(12):8489–503. 10.21037/qims-23-309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Oh HJ, Yoon BH, Ha YC, Suh DC, Lee SM, Koo KH, et al. The change of bone mineral density and bone metabolism after gastrectomy for gastric cancer: a meta-analysis. Osteoporos Int. 2020;31(2):267–75. 10.1007/s00198-019-05220-2. [DOI] [PubMed] [Google Scholar]
  • 10.Park HS, Kim HS, Beom SH, Rha SY, Chung HC, Kim JH, et al. Marked loss of muscle, visceral fat, or subcutaneous fat after gastrectomy predicts poor survival in advanced gastric cancer: single-center study from the CLASSIC trial. Ann Surg Oncol. 2018;25(11):3222–30. 10.1245/s10434-018-6624-1. [DOI] [PubMed] [Google Scholar]
  • 11.Hu Y, Vos EL, Baser RE, Schattner MA, Nishimura M, Coit DG, et al. Longitudinal analysis of quality-of-life recovery after gastrectomy for cancer. Ann Surg Oncol. 2021;28(1):48–56. 10.1245/s10434-020-09274-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Lou J, Gong B, Li Y, Guo Y, Li L, Wang J, et al. Bone mineral density as an individual prognostic biomarker in NSCLC patients treated with immune checkpoint inhibitors. Front Immunol. 2024;15:1332303. 10.3389/fimmu.2024.1332303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Chen H, Zhu X, Zhou Q, Pu X, Wang B, Lin H, et al. Utility of MRI-based vertebral bone quality scores and CT-based Hounsfield unit values in vertebral bone mineral density assessment for patients with diffuse idiopathic skeletal hyperostosis. Osteoporos Int. 2024;35(4):705–15. 10.1007/s00198-023-06999-x. [DOI] [PubMed] [Google Scholar]
  • 14.Kim KE, Bae SU, Jeong WK, Baek SK. Impact of preoperative visceral fat area measured by bioelectrical impedance analysis on clinical and oncologic outcomes of colorectal cancer. Nutrients. 2022;14(19):3971. 10.3390/nu14193971. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Shi S, Zhao YX, Fan JL, Chang LY, Yu DX. Development and external validation of a nomogram including body composition parameters for predicting early recurrence of hepatocellular carcinoma after hepatectomy. Acad Radiol. 2023;30(12):2940–53. 10.1016/j.acra.2023.05.022. [DOI] [PubMed] [Google Scholar]
  • 16.Cui B, Duan J, Zhu L, Wang G, Sun X, Su Z, et al. Effect of laparoscopic sleeve gastrectomy on mobilization of site-specific body adipose depots: a prospective cohort study. Int J Surg. 2023;109(10):3013–20. 10.1097/JS9.0000000000000573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Li Q, Xu WY, Sun NN, Feng QX, Zhu ZN, Hou YJ, et al. MRI versus dual-energy CT in local-regional staging of gastric cancer. Radiology. 2024;312(1):e232387. 10.1148/radiol.232387. [DOI] [PubMed] [Google Scholar]
  • 18.Xu M, Liu D, Wang L, Sun S, Liu S, Zhou Z. Clinical implications of CT-detected ascites in gastric cancer: association with peritoneal metastasis and systemic inflammatory response. Insights Imaging. 2024;15(1):237. 10.1186/s13244-024-01818-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Dong D, Fang MJ, Tang L, Shan XH, Gao JB, Giganti F, et al. Deep learning radiomic nomogram can predict the number of lymph node metastasis in locally advanced gastric cancer: an international multicenter study. Ann Oncol. 2020;31(7):912–20. 10.1016/j.annonc.2020.04.003. [DOI] [PubMed] [Google Scholar]
  • 20.Bae JS, Chang W, Kim SH, Choi Y, Kong SH, Lee HJ, et al. Development of a predictive model for extragastric recurrence after curative resection for early gastric cancer. Gastric Cancer. 2022;25(1):255–64. 10.1007/s10120-021-01217-1. [DOI] [PubMed] [Google Scholar]
  • 21.Liu S, Qiao X, Xu M, Ji C, Li L, Zhou Z. Development and validation of multivariate models integrating preoperative clinicopathological parameters and radiographic findings based on late arterial phase CT images for predicting lymph node metastasis in gastric cancer. Acad Radiol. 2021;28(Suppl 1):S167–78. 10.1016/j.acra.2021.01.011. [DOI] [PubMed] [Google Scholar]
  • 22.Just IA, Schoenrath F, Roehrich L, Heil E, Stein J, Auer TA, et al. Artificial intelligence-based analysis of body composition predicts outcome in patients receiving long-term mechanical circulatory support. J Cachexia Sarcopenia Muscle. 2024;15(1):270–80. 10.1002/jcsm.13402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.He M, Chen ZF, Zhang L, Gao X, Chong X, Li HS, et al. Associations of subcutaneous fat area and systemic immune-inflammation index with survival in patients with advanced gastric cancer receiving dual PD-1 and HER2 blockade. J Immunother Cancer. 2023;11(6):e007054. 10.1136/jitc-2023-007054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Oh J, Kim B, Oh G, Hwangbo Y, Ye JC. End-to-end semi-supervised opportunistic osteoporosis screening using computed tomography. Endocrinol Metab (Seoul). 2024;39(3):500–10. 10.3803/EnM.2023.1860. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Han M, Yao G, Zhang W, Mu G, Zhan Y, Zhou X et al. Segmentation of CT Thoracic Organs by Multi-resolution VB-nets. SegTHOR@ ISBI. 2019; 2019:1–4.
  • 26.Que Y, Xia X, Luo X, Deng Y, Chen M. Diagnostic performance of an artificial intelligence based software for opportunistic osteoporosis detection using chest CT. European Journal of Radiology Artificial Intelligence. 2025;3:100025. 10.1016/j.ejrai.2025.100025. [Google Scholar]
  • 27.American College of Radiology. ACR–SPR–SSR practice parameter for the performance of quantitative computed tomography (QCT) bone densitometry. Accessed March 31. 2022. Available at: http://www.acr.org/~/media/ACR/Documents/PGTS/guidelines/QCT.pdf
  • 28.Chen BB, Liang PC, Shih TT, Liu TH, Shen YC, Lu LC, et al. Sarcopenia and myosteatosis are associated with survival in patients receiving immunotherapy for advanced hepatocellular carcinoma. Eur Radiol. 2023;33(1):512–22. 10.1007/s00330-022-08980-4. [DOI] [PubMed] [Google Scholar]
  • 29.Su H, Ruan J, Chen T, Lin E, Shi L. CT-assessed sarcopenia is a predictive factor for both long-term and short-term outcomes in gastrointestinal oncology patients: a systematic review and meta-analysis. Cancer Imaging. 2019;19(1):82. 10.1186/s40644-019-0270-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Lawrance EL, Jennings N, Kioupi V, Thompson R, Diffey J, Vercammen A. Psychological responses, mental health, and sense of agency for the dual challenges of climate change and the COVID-19 pandemic in young people in the UK: an online survey study. Lancet Planet Health. 2022;6(9):e726–38. 10.1016/S2542-5196(22)00172-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Long MB, Gilmour A, Hull RC, Giam YH, Richardson H, Hughes C, et al. Investigating the impact of dipeptidyl peptidase-1 inhibition in humans using multi-omics. J Allergy Clin Immunol. 2025. 10.1016/j.jaci.2025.07.016. [DOI] [PubMed] [Google Scholar]
  • 32.Swanton C, Bernard E, Abbosh C, Andre F, Auwerx J, Balmain A, et al. Embracing cancer complexity: hallmarks of systemic disease. Cell. 2024;187(7):1589–616. 10.1016/j.cell.2024.02.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Ghaben AL, Scherer PE. Adipogenesis and metabolic health. Nat Rev Mol Cell Biol. 2019;20(4):242–58. 10.1038/s41580-018-0093-z. [DOI] [PubMed] [Google Scholar]
  • 34.Burak MF, Stanley TL, Lawson EA, Campbell SL, Lynch L, Hasty AH, et al. Adiposity, immunity, and inflammation: interrelationships in health and disease: a report from 24th Annual Harvard Nutrition Obesity Symposium, June 2023. Am J Clin Nutr. 2024;120(1):257–68. 10.1016/j.ajcnut.2024.04.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Whitehead A, Krause FN, Moran A, MacCannell ADV, Scragg JL, McNally BD, et al. Brown and beige adipose tissue regulate systemic metabolism through a metabolite interorgan signaling axis. Nat Commun. 2021;12(1):1905. 10.1038/s41467-021-22272-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Doyle SL, Donohoe CL, Lysaght J, Reynolds JV. Visceral obesity, metabolic syndrome, insulin resistance and cancer. Proc Nutr Soc. 2012;71(1):181–9. 10.1017/S002966511100320X. [DOI] [PubMed] [Google Scholar]
  • 37.Lu Y, Tang H, Huang P, Wang J, Deng P, Li Y, et al. Assessment of causal effects of visceral adipose tissue on risk of cancers: a Mendelian randomization study. Int J Epidemiol. 2022;51(4):1204–18. 10.1093/ije/dyac025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Bimurzayeva A, Kim MJ, Ahn JS, Ku GY, Moon D, Choi J, et al. Three-dimensional body composition parameters using automatic volumetric segmentation allow accurate prediction of colorectal cancer outcomes. J Cachexia Sarcopenia Muscle. 2024;15(1):281–91. 10.1002/jcsm.13404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Lee PC, Cheng TY, Ho CT, Huang KW, Chau GY, Huang YH, et al. Gender different impacts of muscle mass and adipose tissue on patients with hepatocellular carcinoma undergoing surgical resection. Liver Int. 2025;45(2):e16237. 10.1111/liv.16237. [DOI] [PubMed] [Google Scholar]
  • 40.Blanc-Durand P, Campedel L, Mule S, Jegou S, Luciani A, Pigneur F, et al. Prognostic value of anthropometric measures extracted from whole-body CT using deep learning in patients with non-small-cell lung cancer. Eur Radiol. 2020;30(6):3528–37. 10.1007/s00330-019-06630-w. [DOI] [PubMed] [Google Scholar]
  • 41.Mantz L, Mercaldo ND, Simon J, Kaess P, Yang K, Dietrich AW, et al. Preoperative chest CT myosteatosis indicates worse postoperative survival in stage 0-IIB Non-Small cell lung cancer. Radiology. 2025;314(2):e240282. 10.1148/radiol.240282. [DOI] [PubMed] [Google Scholar]
  • 42.Surov A, Meyer HJ, Hinnerichs M, Ferraro V, Zeremski V, Mougiakakos D, et al. CT-defined sarcopenia predicts treatment response in primary central nervous system lymphomas. Eur Radiol. 2024;34(2):790–6. 10.1007/s00330-023-09712-y. [DOI] [PubMed] [Google Scholar]
  • 43.Wu JM, Tsai HH, Tseng SM, Liu KL, Lin MT. Perioperative glutamine supplementation may restore atrophy of Psoas muscles in gastric adenocarcinoma patients undergoing gastrectomy. Nutrients. 2024;16(14):2301. 10.3390/nu16142301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Wu CH, Ho MC, Chen CH, Liang JD, Huang KW, Cheng MF, et al. Computed tomography-defined sarcopenia in outcomes of patients with unresectable hepatocellular carcinoma undergoing radioembolization: assessment with total abdominal, psoas, and paraspinal muscles. Liver Cancer. 2023;12(6):550–64. 10.1159/000529676. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Akad F, Stan CI, Zugun-Eloae F, Peiu SN, Akad N, Crauciuc DV, et al. Nutritional and biochemical outcomes after total versus subtotal gastrectomy: insights into early postoperative prognosis. Nutrients. 2025;17(13):2146. 10.3390/nu17132146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Ueno K, Nishigori T, Tokoro Y, Hirai K, Miki A, Yamashita Y, et al. Efficacy of simple oral nutritional supplements versus usual care in postoperative patients with gastric cancer: A Multicenter, Open-label, Parallel, randomized controlled trial. Ann Surg Published Online August. 2025;7. 10.1097/SLA.0000000000006888. [DOI] [PubMed]
  • 47.Climent M, Pera M, Aymar I, Ramon JM, Grande L, Nogues X. Bone health in long-term gastric cancer survivors: a prospective study of high-dose vitamin D supplementation using an easy administration scheme. J Bone Miner Metab. 2018;36(4):462–9. 10.1007/s00774-017-0856-1. [DOI] [PubMed] [Google Scholar]
  • 48.Yoo SH, Lee JA, Kang SY, Kim YS, Sunwoo S, Kim BS, et al. Risk of osteoporosis after gastrectomy in long-term gastric cancer survivors. Gastric Cancer. 2018;21(4):720–7. 10.1007/s10120-017-0777-7. [DOI] [PubMed] [Google Scholar]
  • 49.Sakurai Y, Honda M, Kawamura H, Kobayashi H, Toshiyama S, Yamamoto R, et al. Relationship between physical activity and bone mineral density loss after gastrectomy in gastric cancer patients. Support Care Cancer. 2022;31(1):19. 10.1007/s00520-022-07500-w. [DOI] [PubMed] [Google Scholar]
  • 50.Rejeski K, Cordas Dos Santos DM, Parker NH, Bucklein VL, Winkelmann M, Jhaveri KS, et al. Influence of adipose tissue distribution, sarcopenia, and nutritional status on clinical outcomes after CD19 CAR T-cell therapy. Cancer Immunol Res. 2023;11(6):707–19. 10.1158/2326-6066.CIR-22-0487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Atsumi Y, Rino Y, Aoyama T, Okuda N, Kawahara S, Kazama K, et al. A gender comparison of bone metabolic changes after gastric cancer surgery: a prospective observational study. In Vivo. 2021;35(4):2341–8. 10.21873/invivo.12510. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1. (396.9KB, docx)

Data Availability Statement

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Articles from BMC Gastroenterology are provided here courtesy of BMC

RESOURCES