Sarcopenia as a prognostic factor in gastric cancer patients undergoing gastrectomy: Implications for postoperative and long-term outcomes

Abstract

Gastric cancer is the fifth most common cancer worldwide. The primary way to treat gastric cancer is surgical gastrectomy. Sarcopenia is being increasingly recognized as a crucial factor in determining post-gastrectomy outcomes. However, conventional models used to determine surgical risk have long focused on clinical-pathologic characteristics and the tumor involvement. The purpose of this review is to assess whether sarcopenia is an unaccounted factor of postoperative, short-term, long-term and survival outcomes. Current research shows that sarcopenia is linked to increased risks of postoperative morbidity including infectious complications, anastomotic leakage, longer periods of hospitalization, and readmission. Moreover, studies have shown that the functional measures of sarcopenia predict short-term postoperative outcomes. Additionally, when malnutrition and sarcopenia co-exist, they exacerbate the risk for complications during both the perioperative period and postoperative recovery. Several studies have demonstrated a significant decrease in disease-free survival and overall survival for patients with sarcopenia compared to non-sarcogenic patients, while muscle quality and functional parameters are stronger predictors than quantity of muscle alone. Perioperative changes in body composition including post-gastrectomy muscle loss and alterations in adipose tissue further enhance the ability to assess risk and may provide more accurate prediction. The co-existing phenotypes such as obesity with sarcopenia and myosteatosis also showed even larger negative effects on survival. New evidence also supports the use of multimodal models that incorporate sarcopenia, resulting in better discrimination between high vs low risk patients than using staging systems alone. To conclude, sarcopenia is an independent predictor, which can help to preoperatively identify patients at highest risk for postoperative complications and guide personalized interventions.

Key Words: Gastric cancer; Gastrectomy; Sarcopenia; Body composition; Skeletal muscle index; Postoperative complications; Overall survival; Sarcopenic obesity

Core Tip: Sarcopenia is increasingly recognized as a clinically important predictor of outcome in patients undergoing gastrectomy due to gastric cancer. Reduced skeletal mass, muscle quality, and functional performance contribute to postoperative morbidity and long-term prognosis. Systemic inflammation and adipose tissue distribution also play a significant role. This review highlights the growing need for multidimensional assessment models that incorporate computed tomography-derived body composition markers as well as functional tests and markers of nutritional-inflammatory activity. Incorporating sarcopenia into preoperative risk stratification may improve patient selection. Finally, it may guide perioperative optimization and support more personalized surgical and oncological management.

INTRODUCTION

Cancer develops from an abnormal accumulation of genetic and epigenetic changes disrupting normal cell cycle control, apoptosis (programmed death), and normal maintenance of tissue health, leading to uncontrolled growth, invasion into surrounding tissues, and metastases to remote sites. The past century has witnessed significant advancements in our basic scientific knowledge of how cancer develops; this is reflected by similar developments in the use of treatments for cancer. In early times, the primary means of treating cancer was through surgery and subsequent chemotherapy and/or radiotherapy which are non-specifically toxic to all rapidly dividing cells within the body. In more modern times, there has been a trend away from the use of broad spectrum therapy to target cancer based upon specific biological characteristics of the cancer. This includes immunotherapy which utilizes the body’s natural ability to fight off disease, specifically recognizing and eliminating cancerous cells. These changing trends in the treatment of cancer reflect not only advancements made in the field of molecular oncology but also increasing recognition that effectively managing patients with cancer requires treatment of not only the neoplastic process but also the patient’s overall physiological status[1].

Gastric cancer is still a huge problem for health worldwide, showing up as one of the most frequently diagnosed and deadliest forms of malignancies. Even with improvements in surgery, care around the time of surgery, chemotherapy, and immune therapy, nearly a million people are diagnosed with it and many hundreds of thousands die from it every year[2-6]. Gastrectomy is still the main way to cure gastric cancer if it has not spread, but there are many difficulties. After the surgery, patients often have problems, the cancer can come back, they lose their ability to get enough nourishment, and sadly, many will die in the long run. This is particularly true for older, weaker individuals or people having many treatments at once[7-10]. And within all these, doctors are now seeing that a loss of muscle—a decreasing of both the amount and strength of muscle in the body alongside a decline in how well someone can physically do things—is a key factor in how a patient is doing. More specifically, in more recent interpretative frameworks, not only the importance of muscle quantity but also of its quality is emphasized. Muscle strength and functional reserve are also considered important[2,11-13].

Sarcopenia is important in gastric cancer due to many factors that contribute to muscle loss and functional decline. Tumor-related inflammation, anorexia, and nausea play an important role. Dysphagia, malabsorption, and reduced food intake are equally important. Finally, it is worth mentioning the reduced peristaltic function and changes in the anatomy of the gastrointestinal tract after gastrectomy. Furthermore, the catabolic and immunometabolic disorders associated with cancer worsen muscle wasting. On the other hand, chemotherapy can further worsen malnutrition and reduce tolerance to treatment[3,14,15].

The reported incidence of sarcopenia varies considerably due to many differences in diagnostic criteria and patient populations. It also varies by tumor stage and assessment methods. According to the literature, there is a range from approximately 10% to over 50%. Specifically high rates are detected in elderly patients and in those with advanced disease. The above variations signal that in our bibliography, there is still some variation in the cutoffs and definitions[9,16-18].

In clinical terms, sarcopenia has been associated with increased postoperative complications, longer hospital stays, higher healthcare costs, reduced tolerance to chemotherapy, reduced quality of life, poorer functional recovery, higher risk of recurrence, and decreased overall survival (OS). Some reports have indicated that sarcopenia may be associated with different prognoses depending on tumor stage, patient age, surgical strategy, and presence or absence of systemic inflammation, malnutrition, frailty, or cachexia[19-22]. The most accurate and practical method of body composition assessment is computed tomography (CT). It is used in the staging of cancer and can be used to quantify skeletal muscle and adipose tissue at the L3 vertebra. The most commonly measured indexes are skeletal muscle index (SMI), total psoas area index (TPAI), visceral fat area (VFA), visceral fat index (VFI), skeletal muscle density (SMD), and myosteatosis. CT-based tests have been shown to measure muscle mass as a proxy for sarcopenia, but may not accurately reflect muscle function, contractile quality, or physiological reserve[2,19,23-25].

Related body-composition phenotypes such as visceral obesity, myosteatosis, and sarcopenic obesity, as well as combined nutritional-inflammatory parameters such as Geriatric Nutritional Risk Index (GNRI), cachexia index (CXI), albumin-bilirubin (ALBI) score, and deep learning/radiomics-derived CT signatures, are becoming more important because body mass index (BMI) alone does not take into account adiposity, ectopic fat mass, metabolic risk, and postoperative morbidity, and combined measures of muscle fat or muscle inflammation may provide a more biologically meaningful risk classifying variables[24,26-30]. Overall, the literature consensus is that sarcopenia in gastric cancer must not be interpreted as low muscle mass, but as a multi-faceted syndrome of malnutrition, frailty, inflammation, obesity, ineffective immunity, and decreased resiliency to treatment. There is still a gap in our understanding of how to manage this syndrome about diagnostic cut-offs, integrating functional and imaging assessment, the prediction value of muscle loss postoperatively, and the diet and exercise regimens to avoid or reverse body-composition abnormalities in this population[3,31-35].

LITERATURE REVIEW

This study was designed as a structured narrative review of the current literature examining the clinical and prognostic significance of sarcopenia and related body-composition abnormalities in patients with gastric cancer undergoing gastrectomy. The review focused on the role of sarcopenia as a predictor of postoperative morbidity, long-term recovery, recurrence, disease-free survival (DFS), and OS. Particular emphasis was placed on CT-derived body-composition parameters, functional measures of sarcopenia, nutritional and inflammatory indices, and composite phenotypes such as sarcopenic obesity and myosteatosis.

Given the heterogeneity of the available literature in terms of study design, sarcopenia definitions, diagnostic thresholds, imaging methods, patient populations, surgical approaches, and reported outcomes, a formal meta-analysis was not planned. The evidence was therefore synthesized narratively, with the aim of identifying consistent patterns, methodological limitations, and clinically relevant implications for risk stratification and perioperative management.

A search of the literature in PubMed was performed to identify relevant studies investigating sarcopenia, body composition, and postoperative or oncological outcomes in patients with gastric cancer undergoing gastrectomy. The search was limited primarily to articles published in the last 5 years, but older landmark studies were also included when they provided important definitions, methodological frameworks, or clinically relevant findings.

The search strategy was developed using the following keywords, with Boolean operators AND/OR: Gastric cancer, gastric carcinoma, gastrectomy, gastric surgery, sarcopenia, skeletal muscle index, muscle mass, muscle strength, handgrip strength, gait speed, myosteatosis, sarcopenic obesity, body composition, CT-based sarcopenia, postoperative complications, disease-free survival, progression-free survival, and overall survival. Additional eligible articles were identified by manual screening of reference lists of relevant reviews, meta-analyses, and primary studies.

Eligibility was restricted to studies reporting on adult patients with gastric cancer assessing the association of sarcopenia or other parameters of body composition with surgical, postoperative, long-term, or survival outcomes. Eligible evidence consisted of retrospective and prospective cohort studies, observational studies, secondary analyses of clinical trials, systematic reviews, meta-analyses, and clinically relevant methodological studies.

Studies were included if they reported one of the following: Diagnostic assessment of sarcopenia, skeletal muscle measurements based on CT, muscle quality parameters, functional parameters (handgrip strength or gait speed), nutritional or inflammatory indices, postoperative complications, hospital stay, readmission, recurrence, DFS, progression-free survival, or OS.

Studies were excluded if they did not specifically address gastric cancer, did not assess sarcopenia or body composition, were focused solely on non-surgical cancer populations without relevance to gastrectomy, did not have clinically interpretable outcomes, or were not available in full text. Primary evidence was not from editorials, letters without original data, conference abstracts without sufficient methodological detail, and non-human studies, although selected expert reviews were used to contextualize definitions and clinical interpretation.

Relevant information was extracted from each included study using a structured approach. The extracted variables included study design, year of publication, patient population, sample size, cancer stage, treatment setting, type of gastrectomy, use of neoadjuvant or adjuvant therapy, method of sarcopenia assessment, diagnostic cut-off values, imaging level, functional assessment tools, nutritional or inflammatory markers, and reported clinical outcomes.

Particular attention was given to the method used to define sarcopenia. CT-based studies were reviewed for the use of SMI at the third lumbar vertebra (L3), TPAI, psoas muscle index (PMI), SMD, muscle attenuation (MA), and evidence of myosteatosis. Studies using functional definitions were reviewed for the inclusion of handgrip strength, gait speed, chair-rise test, SARC-F, Eastern Cooperative Oncology Group (ECOG) performance status, or other indicators of physical performance. Where available, the use of international criteria such as EWGSOP2, AWGS, GLIM, or ESPEN/EASO definitions was recorded.

The outcomes of interest were grouped into four main domains: Methods of sarcopenia assessment, postoperative outcomes, long-term oncological outcomes, and survival outcomes. Postoperative outcomes included overall complications, severe complications, infectious complications, anastomotic leakage, pancreatic fistula, pulmonary complications, length of hospital stay, intensive care unit (ICU) admission, and readmission. Long-term and survival outcomes included recurrence, DFS, progression-free survival, OS, cancer-specific survival, treatment tolerance, chemotherapy toxicity, and longitudinal changes in body composition.

The primary outcome of interest was the association between sarcopenia and postoperative morbidity after gastrectomy for gastric cancer. Secondary outcomes included severe postoperative complications, specific complication types, length of hospitalization, readmission, postoperative recovery, treatment tolerance, recurrence, DFS, progression-free survival, and OS. Additional exploratory outcomes included the prognostic role of muscle quality, sarcopenic obesity, visceral fat distribution, nutritional-inflammatory scores, and longitudinal changes in body composition before and after surgery or systemic therapy.

METHODS OF ASSESSMENT

The assessment of sarcopenia and more generally of body composition in gastric cancer is mainly based on multivariate parameters. The dominant tool is CT at the level of the L3, which is considered particularly useful in oncology. This is because it is already available from the preoperative staging control and is strongly correlated with total body muscle mass[3,35,36]. In most primary studies, muscle mass was estimated using CT-based SMI. Specifically, the cross-sectional area of skeletal muscle at L3 was measured and normalized to height in m². Most often, density thresholds ranging from -29 HU to +150 HU were used for skeletal muscle. In contrast, some studies used more specific indices such as the PMI or the TPAI as simpler and more reproducible surrogate indices of muscle mass[2,28,37,38]. The main CT-derived body composition parameters used in gastric cancer are summarized in Table 1[1]. Many studies at the same time were not limited to muscle quantity but also incorporated muscle quality indicators. Typical examples are the mean MA (MA or SMD). Myosteatosis, i.e., fatty infiltration of the muscle, is also important. This was assessed in HU at L3 and was used either autonomously or in combination with SMI for more complex risk classifications[23,39-41]. Muscle quality must be distinguished from muscle quantity. The SMI, PMI, and TPAI indices primarily quantify muscle surface area or mass. In contrast, SMD reflects the mean radiodensity of skeletal muscle on CT and is expressed in HU. A lower SMD value generally indicates increased fat infiltration and reduced muscle quality. The term myosteatosis refers to the pathological accumulation of intramuscular and intermuscular fat; on CT scans, it is identified by reduced radiation attenuation by the muscle tissue (muscle attenuation). Although SMD and myosteatosis are closely linked, they should not be considered identical concepts. SMD represents the measurable radiological parameter. On the other side, myosteatosis is the biological and morphological phenotype that this parameter seeks to capture[5,39,41-43]. The reported cut-off values in HU vary across studies depending on several factors. Specifically, they differ based on gender, BMI, and the CT scan protocol. They also vary according to the software used and population characteristics. This partly explains the heterogeneity in the definitions of sarcopenia and myosteatosis in gastric cancer[16,18,29].

Table 1 Computed tomography-based body composition parameters.

Parameter	Clinical meaning	Assessment method	Common diagnostic cut-off	Prognostic role	Strengths	Limitations
SMI	Muscle quantity, normalized for body size	Cross-sectional skeletal muscle area at L3 (cm2), indexed to height2 (cm2/m2)	Sarcopenia: < 52.4 cm2/m2 (men), < 38.5 cm2/m2 (women); BMI-adjusted cut-offs also used	Overall survival; predicts chemotherapy toxicity, postoperative complications	Well-validated; single L3 slice correlates strongly with whole-body muscle mass; widely used across cancer/surgical literature	No universal cut-off (varies by ethnicity, population, scanner protocol); requires dedicated software/manual segmentation; height2 normalization may under/overcorrect in obese or very short/tall patients
PMI	Surrogate for muscle mass using a single muscle	Cross-sectional area of bilateral psoas muscles at L3 or L4, indexed to height2 (cm2/m2)	No consensus; commonly cited thresholds approximately < 5.0 cm2/m2 (men), < 3.4 cm2/m2 (women), varies widely by study	Postoperative complications, morbidity	Fast to measure (small, discrete structure); minimal manual tracing needed; feasible even on lower-quality/Limited-slice scans	Poorly correlates with total skeletal muscle mass; high inter-study variability in cut-offs; psoas size affected by positioning and vertebral level chosen; less validated than SMI
SMD	Muscle quality (myosteatosis/fatty infiltration)	Mean CT attenuation (HU) of skeletal muscle at L3	Myosteatosis: < 41 HU (BMI < 25) or < 33 HU (BMI ≥ 25)	Overall survival, independent of muscle quantity	Captures muscle quality, not just size - adds prognostic information beyond SMI; low cost (same CT slice as SMI)	Attenuation affected by CT protocol/reconstruction settings and contrast phase; BMI-stratified cut-offs complicate use; less intuitive for clinicians
VFA	Adiposity, visceral fat burden	Cross-sectional visceral adipose tissue area at L3 (or umbilicus level), cm2	Visceral obesity: ≥ 100 cm2 (women), ≥ 130-163 cm2 (men), varies by study/ethnicity	Postoperative and infectious complications	Strong correlate of metabolic risk and surgical complication risk; reproducible measurement	Cut-offs not standardized across populations; requires accurate delineation from subcutaneous fat; single-slice VFA may not reflect total visceral fat volume
IMAC	Poor muscle quality (fat infiltration within/around muscle)	Ratio of CT attenuation of intramuscular/paraspinal fat to that of paraspinal muscle (using psoas or paraspinal muscles)	No universal threshold; typically defined by cohort-specific median or tertile split	Mortality	Complements SMD; useful when psoas is the only reliably visualized muscle group; correlates with frailty	Lacks standardized cut-off; formula/muscle group varies between studies, limiting comparability; less widely validated than SMI/SMD

SMI: Skeletal muscle index; BMI: Body mass index; CT: Computed tomography; PMI: Psoas muscle index; SMD: Skeletal muscle density; HU: Hounsfield units; VFA: Visceral fat area; IMAC: Intramuscular adipose content.

In newer assessment schemes, the diagnosis of sarcopenia is moving away from the exclusive dependence on axial measurements. It is approached based on combined criteria of the international groups EWGSOP2, AWGS, and GLIS. In other words, this is done through the coexistence of low muscle mass with low muscle strength and/or low physical performance. The most common functional tests are the handgrip strength, the 4-meter or 6-meter walk, as well as the chair rise test and the SARC-F. There are also newer indicators such as muscle-specific strength, i.e., the ratio of the grip to the muscle surface at L3. This aims to better reflect the actual contractile capacity of the muscle tissue[2,11-13,20,44]. Functional assessment appears to be particularly important in studies of elderly and surgical patients. There, physical performance was assessed by 6-m usual gait speed, 7-step stair climbing test, walking speed, ECOG performance status, or combinations thereof. This was done because simple muscle mass measurement does not adequately capture physiological reserve, frailty, and recovery capacity after gastrectomy[8,32,44,45]. In addition to axial CT assessment, other techniques used were bioelectrical impedance analysis (BIA). This is presented as an objective, non-invasive, repeatable, and cost-effective method for assessing muscle mass and fat percentage. There is also DXA. Several studies compare the simultaneous use of BIA, DXA, and CT or apply the AWGS criteria in postoperative populations through appendicular skeletal muscle mass index and handgrip strength[21,46,47]. At the same time, many studies have also incorporated fat compartments into their methodology. They used the measurement of VFA/VFI (VFA/VFI/VATI), subcutaneous fat area/index, as well as visceral-to-subcutaneous ratio and body fat ratio. However, they also used a combination of these, such as sarcopenic obesity. They usually had a visceral obesity threshold of 100 cm² and used specific sex-specific thresholds for men and women. This was done in order to capture the heterogeneity of body composition beyond BMI[24,27,48-50]. In more complex research protocols, body composition was combined with nutritional and inflammatory markers. Typical examples are prognostic nutritional index (PNI), GNRI, ALBI, neutrophil-to-lymphocyte ratio (NLR), and CXI (= SMI × albumin/NLR), as well as peripheral immune profile. In this way, they allowed the creation of complex prognostic indicators and nomograms that attempted to simultaneously integrate various parameters. These included muscle mass and muscle quality and nutritional status. Finally, systemic inflammation and immune function also played a decisive role[28,31,32,51,52]. In addition, newer technological applications have utilized quantitative CT as well as artificial intelligence (AI)-based segmentation and deep learning-based CT segmentation. They have also used radiomics as well as machine learning and transformer-based multimodal models for various purposes. Specifically, they have been used for automatic or semi-automatic segmentation of muscle and adipose tissue as well as calculation of morphological and textural characteristics. They have also been utilized for developing risk models for recurrence as well as OS or postoperative complications with high reproducibility and a clear attempt to limit the bias of manual measurement[6,30,53-56]. A significant recent development in body composition assessment is the shift from manual or semi-automated two-dimensional analysis of a single cross-section at the L3 vertebral level to fully automated three-dimensional volumetric assessment. While the L3-based skeletal muscle area remains practical and widely validated, it captures only a limited anatomical snapshot and may not fully represent the distribution of muscle and adipose tissue across the entire body. AI-assisted segmentation pipelines can rapidly and reproducibly extract muscle, visceral adipose tissue, subcutaneous adipose tissue, and intramuscular fat from routine preoperative staging CT scans. This enables the opportunistic assessment of body composition without an additional imaging burden. It can also reduce inter-observer variability associated with manual segmentation. It is anticipated that, in the future, three-dimensional volumetric approaches may enable a more comprehensive phenotypic analysis of sarcopenia, myosteatosis, sarcopenic obesity, and cachexia-related tissue loss. They are expected to yield particularly valuable insights when integrated with clinical, nutritional, inflammatory, and oncological variables[6,30,53,56-58].

Methodologically, the studies included various elements. These were prospective cohorts as well as retrospective single-center or multicenter series and secondary analyses of randomized controlled trials. These also included propensity score matching as well as scoping reviews and meta-analyses. However, the endpoints usually included postoperative complications within 30 days graded based on the Clavien-Dindo classification, as well as serious complications and length of stay. They also included readmission, OS, DFS, progression-free survival, or dynamic changes in body composition pre- and postoperatively or before and after neoadjuvant therapy[20,22,29,59,60]. Overall, the assessment methodology in gastric cancer is evolving from single-dimensional CT measurements of muscle surface area to multidimensional and increasingly standardized models that combine various factors. Typical examples are imaging body composition as well as muscle strength and physical performance. They also include nutritional/inflammatory markers and computational tools. However, the literature is still characterized by significant heterogeneity in cut-offs as well as in diagnostic frameworks and measurement techniques. This reinforces the need for further international standardization[3,4,16,18,61].

WHY MAY SARCOPENIA AFFECT SURGICAL OUTCOMES?

At the level of pathophysiology, the adverse effects associated with sarcopenia, myosteostasis, and sarcopenic obesity, and more generally, body composition disorders in gastric cancer, appear to result from a multifactorial network of interactions. These include several mechanisms and not just one. Typical examples are systemic inflammation and metabolic dysregulation, as well as malnutrition and immune dysfunction. Also important are reduced physiological reserve and therapeutically induced catabolism[3,16,62]. The central inflammatory pathway mainly includes cytokines originating from the tumor or from the inflammatory microenvironment, such as interleukin (IL)-6, tumor necrosis factor (TNF)-α, and IL-1β. These cytokines not only activate the ubiquitin-proteasome pathway and enhance protein catabolism but also accelerate skeletal muscle loss. At the same time, the hypermetabolic state and insulin resistance, as well as the suppression of the insulin-like growth factor 1/protein kinase B/mammalian target of rapamycin axis, reduce muscle protein synthesis and impede anabolic regeneration of muscle tissue[3,41,48]. In gastric cancer, these mechanisms are exacerbated by the particular nutritional burdens of the disease and treatment. This is observed because anorexia, early satiety, nausea, vomiting, dysphagia, and disturbed peristalsis, as well as upper gastrointestinal obstruction and malabsorption, create unfavorable conditions. Specifically, they lead to low protein and calorie intake and characteristic deficiencies. These include deficiencies of branched-chain amino acid and iron, as well as vitamin B12 and fat-soluble vitamins. Thus, they ultimately cause a state of negative energy and nitrogen balance, which undermines both muscle mass and postoperative healing[59,63,64]. Gastrectomy itself is an independent pathophysiological stressor mainly because it alters gastrointestinal hormonal regulation and reduces ghrelin and leptin. It also disrupts nutrient utilization and causes dumping syndrome as well as prolonged anorexia and chronic caloric protein deficiency. Meanwhile, even the type of anastomosis can influence the course of sarcopenia through differences in nutrient absorption and intestinal flora, as well as bile acids and enteroendocrine responses[3,7,65,66]. In parallel, sarcopenia seems to reflect not only weight loss but also an overall decrease in physiological reserve. Skeletal muscle functions as a major reservoir of amino acids and proteins. Specifically, it is essential for the synthesis of redox enzymes and acute-phase proteins, as well as immune cells and healing molecules after major surgical stress. Therefore, its depletion is associated with impaired wound healing, increased vulnerability to infections, and reduced resistance to anemia, as well as reduced cardiorespiratory fitness and slower mobilization after surgery[6,16,28,29,52]. This connection with reduced physiologic reserve explains why many studies identify a stronger association of sarcopenia with systemic rather than purely technical complications. Typical examples are respiratory and cardiovascular burdens and pneumonia, as well as prolonged stay and reduced recovery. This is particularly important when there is low physical performance, low grip, or reduced stair-climbing/gait performance[20,44,45,51]. Equally critical is the immunological axis. Muscle tissue loss is associated with impaired production of myokines that normally have anti-inflammatory and possibly anti-tumor effects. Conversely, lower muscle mass and quality have been associated with reduced CD3+, CD4+, and CD8+ lymphocyte counts, compromised natural killer-cell activity, and worse adaptive immunity, as well as impaired tumor immunosurveillance. These observations suggest that sarcopenia may affect not only surgical recovery but also the very biology of recurrence and OS[32,61,67]. In fact, it has been suggested that sarcopenia reflects a chronic biological vulnerability that affects immunological surveillance and the tumor-host microenvironmental interaction, as well as long-term oncological resistance more than short-term surgical resistance. This explains why some studies find a stronger effect on OS than on immediate postoperative complications[19,22]. Muscle quality also appears to have distinct pathophysiological significance. Myosteostasis and low muscle radiodensity represent fatty infiltration and possibly muscle fibrosis. This results in poorer contractile capacity and metabolic dysfunction, lower mobility, and reduced resistance to surgical stress, even when total muscle surface area is not extremely low. This finding suggests that muscle quantity and muscle quality contribute independently and additively to the postoperative risk[5,39,42,43,68]. At the other end of the spectrum, the pathophysiology of adipose tissue is equally complex. Visceral fat is not a passive energy store but an endocrine and immunologically active organ that promotes chronic low-grade inflammation and insulin resistance as well as perturbed adipokine signaling. It also promotes IL-6 and TNF-α production and macrophage and T-cell recruitment, as well as angiogenesis and metabolic support of the tumor. Conversely, excessive visceral deposition can make surgery technically difficult, increase intra-abdominal pressure, and burden the microcirculation of the anastomosis, contributing to fistulas, infections, and greater surgical stress[24,48,54,69]. At the same time, however, fat mass is not uniformly harmful. Low visceral or subcutaneous fat may reflect advanced cachexia, early fat depletion, reduced energy reserves, and poorer immunometabolic support. This explains why some studies describe an “obesity paradox” or at least a dual pathophysiological reality where excess VAT increases short-term complications but very low fat mass is associated with poor survival and reduced tolerance to treatment[70-72]. Sarcopenic obesity precisely encapsulates this double damage. More specifically, it is characterized by the coexistence of reduced muscle reserve with increased visceral or intramuscular fat deposition. This is considered the worst clinical scenario because it combines reduced amino-acid reserves and antioxidant capacity with chronic adipose-driven inflammation, lipotoxicity, and insulin resistance, altered adipokines, and possible disruption of chemotherapy pharmacokinetics. In this way, it leads to more serious complications and a worse oncological course[27,33,49]. This pathophysiology extends to the dynamic change in body composition under treatment. During neoadjuvant or adjuvant therapy, fat and muscle losses are part of a cachexia-like catabolic cycle. More specifically, fat stores are often depleted before muscle, and maintaining muscle mass requires not only nutritional support but also overcoming cancer-related anabolic resistance through physical activity and especially progressive resistance training[14,59,60]. Finally, several studies support that adverse body-composition phenotypes do not simply correlate with poor prognosis; instead, they serve as biological mediators linking tumor biology, host resilience, and various markers. Typical examples of these are psoas muscle density, CXI, cholinesterase change, GNRI, PNI, NLR, and complex CT- or AI-based models that reflect in different ways the combination of malnutrition, inflammation, immune fatigue, metabolic dysregulation, and functional deterioration. Ultimately, this determines healing and complications, as well as resistance to treatment, relapse, and survival[30,31,43,55,66]. Overall, therefore, the pathophysiology of outcomes in gastric cancer must be seen as a result of the interaction of various factors, specifically tumor, inflammation, and metabolism, as well as immunity, nutrition, and physical functionality. Under such conditions, sarcopenia, myosteostasis, and adipose tissue disorders are not simply concomitant phenomena but active indicators and potential accelerators of the adverse surgical and oncological course[3,16,73].

POSTOPERATIVE OUTCOMES

Sarcopenia and changes to body composition can now be viewed as key factors determining immediate postoperative outcomes following gastrectomy in patients with gastric cancer. Although there are improvements being made in the area of surgery and pre/post-surgical care, complications arising immediately after surgery remain significant. Complications occurring immediately postoperatively range anywhere from around 20% up to around 50%. These include but are not limited to pulmonary complications, anastomotic leakages, intra-abdominal infection, pancreatic fistulas, delayed gastric emptying, and wound site related issues. Postoperative complications occur at a significantly higher rate among those diagnosed with sarcopenia. Patients with sarcopenia-related complications face longer hospital stays, increased health-care expenditures, greater rates of ICU admission, and increased chances of unplanned re-admissions[20,43,54]. According to several cohort studies and meta-analyses, sarcopenia is a strong independent indicator of total and serious postoperative complications, with odds ratios generally reported in the range of 2 to 4.5 once adjusted for potential confounders, such as patient age, tumor stage, and comorbid conditions[74,75]. When considering the predictive capability of sarcopenia in combination with indicators of malnutrition and/or systemic inflammation, it may appear that sarcopenia is indicative of a larger scope of physiologic vulnerability, beyond its identification as merely a musculoskeletal disorder. It should be noted that functional indicators such as low grip strength and slow walking speeds have been found to possess improved predictive capabilities relative to skeletal muscle mass alone, especially when utilizing radiological measures of muscle quantity or quality to support recent multi-dimensional definitions of sarcopenia[25,76]. Additionally, fat mass-related metrics add predictive specificity by demonstrating associations with increased incidence rates of infectious complications and pancreatic fistulae, thus increasing overall morbidity risk[71,77]. In addition to an increased incidence of postoperative readmissions and altered recovery trajectories, sarcopenia is also considered to affect the entire process of convalescence, as opposed to being confined to direct effects on postoperative complications[78]. On the other hand, there have been reports that a decline in SMI alone does not independently increase the occurrence of complications when accounting for other factors (e.g., different evaluation techniques and varying levels for diagnosis), highlighting the underlying heterogeneity across studies[19,49]. Thus, these results indicate that when assessing sarcopenia by using models that include muscle mass, function, and adiposity, it provides clinically useful risk stratification and can provide improved pre- and/or intra-operative decision-making for patients undergoing surgery beyond traditional cancer-related or surgical-based criteria.

LONG-TERM OUTCOMES

The long-term effects of gastrectomy are greatly affected by several patient-related physical characteristics that can change over time. The most obvious are changes in adiposity (nutritional status), physical performance (functional reserve), and skeletal muscle mass (body composition)[79]. Although historically, many researchers considered traditional cancer treatment-related variables to be the best predictors for cancer treatment outcome, current research indicates that these new variables will become increasingly important in predicting how likely it is that patients will live longer[80].

Skeletal muscle atrophy commonly follows surgery and has been identified as one of the strongest indicators of a decreased likelihood of remaining free from cancer recurrence, an increased chance of developing recurrent disease, and an increased chance of death due to cancer. Postoperative muscle wasting was found to almost double the risk of death, and resulted in significant negative impacts on recurrence-specific end points[66,81]. Also, while longitudinal changes in body composition were determined to have a greater impact on prognosis than baseline measures, changes in visceral and subcutaneous adipose tissue that occurred as a result of neoadjuvant chemotherapy were also determined to be predictive of poor progression-free and OS. Specifically, studies demonstrated that changes in visceral and subcutaneous adipose tissue during neoadjuvant chemotherapy had a stronger correlation with poor progression-free survival and OS when compared to sarcopenia alone[60]. Therefore, this research suggests that dynamic metabolic decline represents both tumor behavior and the ability of the host to withstand the aggressiveness of their tumor; furthermore, it appears to capture aspects of disease progression that would otherwise go undetected by using static preoperative evaluations. Sarcopenia has also been shown to impair recovery potential and reduce tolerance to systemic therapy, leading to increased incidence of chemotherapy-related toxicities that often necessitate dose reduction or premature termination, which may ultimately contribute to worse long-term oncological outcomes[3]. Emerging prognostic models using sarcopenia in conjunction with visceral obesity, inflammatory markers, and nutritionally-related measures outperform the traditional TNM-staging classification system for predicting patients at high-risk of cancer recurrence and poor long term-survival[52]. Integrated prognostic models and composite indices relevant to gastric cancer body-composition assessment are summarized in Table 2. Moreover, studies have shown that the collective body composition phenotypes (osteosarcopenia, myosteatosis, and sarcopenic-obesity) are directly correlated to decreased disease free-survival and poorer prognosis, indicating that both muscle quality and fat distribution contain additional prognostic information relative to muscle mass alone[68,82,83]. These composite phenotypes presumably represent a more advanced state of systemic metabolic dysregulation, along with chronic inflammation, thus contributing to tumor progression as well as impaired host-defense mechanisms. Therefore, it is evident from this data that long-term outcome following gastrectomy is the result of an interaction among muscle wasting, changes in adiposity, and systemic metabolic status. Therefore, there is a strong rationale for longitudinal assessment and incorporation of body composition parameters into post-surgical monitoring and predictive modeling.

Table 2 Integrated prognostic models and composite indices in gastric cancer body-composition assessment.

Model/index	Main components	Clinical meaning	Prognostic value	Main limitation
GLIM criteria	Phenotypic criteria, including weight loss, low BMI, reduced muscle mass; etiologic criteria, including reduced intake, inflammation, or disease burden	Standardized assessment of malnutrition with body-composition relevance	Identifies patients at higher risk of postoperative morbidity, poor recovery, and worse survival	Requires nutritional and clinical assessment; cut-offs and implementation may vary across populations
GNRI	Serum albumin and body weight relative to ideal body weight	Nutritional risk and systemic reserve	Associated with postoperative complications, treatment tolerance, and long-term survival	Does not directly measure muscle mass or muscle quality
PNI	Serum albumin and lymphocyte count	Combined nutritional and immune status	Predicts postoperative morbidity, recurrence risk, and overall survival	Influenced by inflammation, infection, hydration status, and non-cancer-related conditions
CXI	Skeletal muscle index × serum albumin/neutrophil-to-lymphocyte ratio	Integrates muscle mass, nutrition, and systemic inflammation	May improve risk stratification beyond isolated CT-derived or laboratory markers	Requires standardized CT assessment and laboratory timing
ALBI-related models	Serum albumin and bilirubin, alone or combined with sarcopenia/body-composition variables	Reflects hepatic and nutritional reserve	May help identify patients with impaired physiological reserve and poorer prognosis	More relevant when liver function has prognostic impact; should not be confused with liver-disease scores such as MELD or CTP
Radiomics models	CT texture features from tumor and/or body-composition compartments combined with clinical variables	Imaging-based prediction of recurrence and survival	Can improve individualized prediction beyond conventional staging alone	Requires software, technical standardization, and external validation
AI-based nomograms	Machine-learning or deep-learning features combined with clinicopathological, CT-derived, nutritional, and inflammatory variables	Multivariable personalized risk prediction	May predict postoperative complications, recurrence, and overall survival more accurately than single markers	Limited by retrospective datasets, black-box interpretability, and generalizability across centers
Integrated body-composition scores	SMI, SMD/myosteatosis, visceral fat, sarcopenic obesity, functional tests, and inflammatory/nutritional markers	Multidimensional assessment of host reserve	Captures the combined prognostic impact of muscle quantity, muscle quality, adiposity, and inflammation	Lack of universally accepted thresholds and standardized scoring systems

GLIM: Global leadership initiative in malnutrition; BMI: Body mass index; GNRI: Geriatric Nutritional Risk Index; PNI: Prognostic nutritional index; CXI: Cachexia index; CT: Computed tomography; ALBI: Albumin-bilirubin; MELD: Model for end-stage liver disease; CTP: Child-Turcotte-Pugh; AI: Artificial intelligence; SMI: Skeletal muscle index; SMD: Skeletal muscle density.

SURVIVAL OUTCOMES

Sarcopenia has been proven by multiple studies to be an independent factor that predicts both lower OS and lower DFS among patients who are treated for gastric cancer through gastrectomy. The evidence clearly demonstrates that those patients classified as having sarcopenia have a substantially shorter duration of survival than their nonsarcopenic counterparts. Furthermore, hazard ratios for death from gastric cancer are often greater than 2-3 when adjusted for other variables, thus strongly indicating that sarcopenia is a strong, independent predictor of prognosis in many different groups of patients[41,84,85].

The size of the negative effects associated with sarcopenia also appear to be influenced by the stage at which gastric cancer was diagnosed. For example, sarcopenia may have a greater impact on prognosis in earlier stages of the disease; however, it is likely that cachexia and advanced systemic wasting will be more predominant factors in determining prognosis in later stages of the disease[86]. Additionally, researchers have found gender-based differences in the relationship between low skeletal muscle mass and long-term survival among male vs female gastric cancer patients. Low skeletal muscle mass (sarcopenia) has been shown to predict poorer long-term survival among male patients, while no such association has been demonstrated among female patients. This suggests that there may exist some form of biological or hormonal modifier(s) of the adverse effects associated with sarcopenia on prognosis risk[36]. These studies have demonstrated that muscle quality parameters (skeletal muscle radiodensity and myosteatosis) could be superior to existing muscle mass metrics for the prediction of survival; in addition, by being included with clinical-pathologic variables, they enhance the discriminative capability of prognostic models[36,39]. Sarcopenia before and/or after surgery is an independent predictor of survival and can increase both overall mortality and disease recurrence rates in those who are non-sarcopenic at baseline[21,87]. Furthermore, adverse composite phenotypes (e.g., sarcopenic obesity and malnutrition-sarcopenia syndrome) appear to be associated with particularly poor prognosis and likely represent additive deleterious effects from loss of muscle mass, metabolic disturbances, and inflammation[33,75]. These results demonstrate a strong rationale for inclusion of sarcopenia and a complete body composition analysis into predictive models because this would offer significant added value over current staging systems and could potentially guide individualized therapies.

SARCOPENIC OBESITY

Sarcopenic obesity is increasingly emerging as one of the most complex and clinically burdensome body composition phenotypes in gastric cancer, as it combines reduced muscle mass and/or low muscle quality with excessive fat deposition, especially visceral and heterotopic fat infiltration, creating a high-risk metabolic and inflammatory environment that often remains invisible when only conventional markers such as BMI are used[3,24]. This concept is of particular importance in gastric cancer because BMI fails to distinguish between lean mass and visceral fat as well as subcutaneous fat. This results in patients with a seemingly “acceptable” or even normal BMI actually having an unfavorable profile with low muscle mass and increased metabolically active fat mass[11,12,16,24]. The literature to date indicates that sarcopenic obesity is not a rare phenomenon. A recent meta-analysis revealed an overall prevalence of 16.3% in patients with gastric cancer. More specifically, values of 18.2% in men and 14.7% in women were observed. Variations were also observed between European and Asian populations, highlighting both the frequency and geographical heterogeneity of the phenotype[4]. In addition, other series have shown that the prevalence varies significantly depending on the diagnostic definition and in particular whether the handgrip is included in the ESPEN/EASO algorithms. In particular, body-composition-only models identify many more cases than models that also require low muscle strength, suggesting that some forms of clinically significant sarcopenic obesity may be missed when strict functional criteria are used[49]. At the definitional level, most imaging studies identify sarcopenic obesity as the combination of low SMI with high visceral fat and/or visceral obesity often with a VFA threshold ≥ 100 cm². In contrast, newer approaches expand the phenotype to include myosteostasis or low muscle density. They present as forms of “hidden” obesity that are not reflected in BMI but are strongly associated with inflammation and insulin resistance as well as poor prognosis[6,27,73]. Indeed, one of the most important contributions of the modern literature is that poor prognosis does not seem to be explained by reduced muscle mass alone or by obesity alone. Rather, it is explained by the interaction between muscle loss and intramuscular fat infiltration and visceral fat accumulation. More specifically, Xu et al[48] showed that the more of these adverse elements coexist, the worse the long-term survival. Especially in studies that classified patients into phenotypes such as non-sarcopenic/non-obese, sarcopenic only, obese only, and sarcopenic obese, sarcopenic obesity emerged as the most aggravating combination, particularly for OS and recurrence-free survival. This association remained significant after adjustment for clinicopathological factors and postoperative complications, supporting the argument for its independent prognostic value[27,33]. Similarly, there have been deeper body composition analyses in which sarcopenic obesity was examined together with myosteatosis and sarcopenic visceral obesity. In these, it was found that phenotypes that include pathological fat deposition, such as SMVO, are associated with even worse survival and worse cachexia-related survival. On the other hand, BMI alone dramatically underestimates these phenomena. This is because only a small percentage of patients were classified as obese by BMI, even though almost half had visceral obesity and the majority presented sarcopenia or myosteatosis[73]. From a pathophysiological point of view, sarcopenic obesity is considered particularly harmful, as it combines the adverse consequences of two metabolically active states. On the one hand, low muscle mass/quality is associated with reduced physiological reserve, weakened immune response, and worse tolerance to treatment. On the other hand, visceral or intramuscular fatty infiltration is associated with chronic inflammation, dysregulation of adipokines, and insulin resistance, as well as lipotoxicity and pro-oncogenic metabolic changes[11,16,73]. This may explain why many studies point out that simply maintaining overall strength or even an “acceptable” BMI can be misleading, particularly because overall grip strength may appear maintained in obese individual whereas muscle-specific strength is significantly lower. Thus, the hidden functional insufficiency of muscle that is consistent with obesity-related sarcopenia is revealed[11,12]. At the same time, the literature shows that the components of fat mass do not always function uniformly. In some contexts, increased visceral fat is associated with more postoperative infectious complications or pancreatic fistulas. In contrast, in others, low visceral fat mass seems to be associated with poor chemotherapy compliance, recurrence, or reduced survival. This suggests that the prognostic significance of fat is dependent not only on the therapeutic context and fat distribution, but also on whether the patient is moving more towards a sarcopenic obesity phenotype or a cachectic/wasted phenotype with low muscle and low fat[71,72,88,89]. Additionally, studies that incorporated systemic inflammation showed that sarcopenic obesity can coexist with low PNI and high platelet-to-lymphocyte ratio as well as other markers of inflammatory activation. These findings reinforce the idea that sarcopenic obesity is a phenotype that is not simply morphological but biologically burdened, associated with worse survival, greater treatment toxicity, and a more unfavorable course after gastrectomy or chemoradiotherapy[90]. Therefore, current evidence supports that sarcopenic obesity should be treated as a distinct and high-risk phenotype in the surgical and oncological management of gastric cancer. In particular, it should not be adequately identified with BMI alone. In particular, it requires a combined assessment of muscle quantity and quality, functionality, and fat distribution. Particular attention is required to visceral and intramuscular fat deposition because these parameters appear to add essential and non-redundant prognostic information beyond classical oncological markers[16,27,30,49,73].

PREHABILITATION STRATEGIES

Recognizing sarcopenia as a modifiable risk factor has significant clinical implications. Prehabilitation aims to enhance physiological reserves prior to gastrectomy through structured exercise, nutritional optimization, and multidisciplinary perioperative support. This is particularly important in the context of gastric cancer, a disease in which patients frequently present with anorexia, reduced oral intake, systemic inflammation, and catabolism. These factors are significant due to their association with treatment outcomes and postoperative nutritional dysfunction. Resistance exercise is central to preparation because it directly targets muscle strength and anabolic signaling. On the other hand, aerobic training can improve cardiopulmonary reserve and functional recovery. Nutritional interventions should prioritize several key elements, including adequate energy intake, high-protein supplementation, correction of micronutrient deficiencies, and personalized support. These can be achieved through oral nutritional supplements or enteral feeding when necessary[14,46,59].

Current evidence suggests that multimodal prehabilitation can improve functional status, as well as treatment tolerance and short-term recovery, particularly in elderly or frail patients. However, it remains uncertain whether such preparation can fully reverse the risks associated with sarcopenia, such as anastomotic leakage, severe postoperative complications, recurrence, or poor OS. Available studies are heterogeneous regarding the type of intervention, duration, timing, adherence, and definitions of sarcopenia. Therefore, prehabilitation should currently be considered a rational and clinically promising strategy rather than a fully validated corrective therapy. Future prospective trials should determine which patients derive the greatest benefit from nutritional prehabilitation, the optimal duration of prehabilitation prior to gastrectomy, and whether improvements in muscle strength or quality translate into better surgical and oncological outcomes[16,18,29,35].

DISCUSSION

The evidence from both surgery and oncology has shown that CT-derived body-composition indices—including SMI, total psoas area, myosteatosis, VFA, and composite sarcopenia scores—provide independent and usually better predictive power than traditional anthropometric measures (e.g., BMI) for outcomes after gastric cancer surgery or multimodality treatment. Notably, multiple studies have indicated that low or “underweight” BMI is associated with poorer prognoses, whereas higher BMI per se does not systematically correlate with adverse outcomes; these data support the concept that “weight” and “fat topology” are not equivalent constructs[33,38,39,57,91].

In addition to cytokine-mediated catabolic reorganization, iron-dependent lipid peroxidation/ferroptosis has recently gained recognition as an important functional pathway contributing to the sarcopenia/adiposity axis in cancer. Ferroptosis is a type of regulated programmed cell death that occurs independent of apoptosis, necrosis, and pyroptosis through disruption in normal iron regulation and failure of glutathione peroxidase 4 (GPX4), a key antioxidant enzyme that prevents the formation of toxic lipid peroxides. Since both skeletal muscle and adipose tissue have the ability to handle iron and are actively engaged in metabolism, impaired ferroptotic function in these tissues may contribute to the previously mentioned catabolic effects of IL-6/TNF-α on muscle mass and quality, while at the same time modifying the rate of adipocyte turnover and overall fat distribution. The dual nature of ferroptosis also illustrates how this process can behave differently depending upon the specific cellular/tissue environment. For example, ferroptosis can be considered a suppressive force against tumors due to its ability to eliminate malignant or drug resistant cancer cells; however, there is evidence suggesting that dysregulation of ferroptosis can lead to enhanced tumor growth and development of drug resistance[92]. In a similar manner to the obesity paradox phenomenon noted above regarding visceral adiposity, the dual nature of ferroptosis suggests that mechanisms related to ferroptosis may cross-talk with other body composition based prognostic models, providing another potential mechanistic relationship between systemic metabolic-inflammatory dysfunction, altered tissue level iron/Lipid handling processes, and overall oncological outcome. Future studies using established inflammatory/body composition measures combined with newly developed ferroptosis associated biomarkers (i.e., GPX4, SLC7A11, and ACSL4) will likely represent a new area for mechanistic and prognostic research in gastric cancer[92].

Although there are many types of assessments regarding the tumor-immune-muscle axis, there has been a greater focus on using functional and molecular imaging techniques to study how the tumor-immune-muscle axis changes during therapy as opposed to assessing it at a single time point. With regard to chimeric antigen receptor-T cell therapy, which utilizes chimeric antigen receptors to treat patients who have hematological malignancies, ¹⁸F-fluorodeoxyglucose positron emission tomography/CT has been utilized to assess early metabolic responses associated with therapy, identify potential immune related side effects such as cytokine release syndrome, and provide information to assist clinicians in identifying responders vs non-responders. This illustrates how traditional methods utilizing CT based imaging can be converted into dynamic, immunotherapy relevant outcomes[93]. Utilizing similar reasoning in treating patients with solid tumors, such as those diagnosed with gastric cancer, by integrating body composition metrics derived from CT images with circulating or tissue based immune biomarkers—including markers that are reflective of immune checkpoints (such as programmed death 1/programmed death-ligand 1 or cytotoxic T lymphocyte-associated antigen-4), or newer targets such as TIM-3, LAG-3, or TIGIT—may potentially improve risk assessment compared to either metric individually, since sarcopenia and myosteatosis are indicative of both malnutrition and a systemic immunosuppressed environment that promotes tumor evasion of the host’s immune system[94]. However, few prospective studies have assessed the relationship between body composition metrics obtained through CT imaging and targeted immune profiles in gastric cancer populations. Therefore, this represents an area of potential for future translational research as opposed to a well-established finding.

The context-dependent nature of the prognostic value of visceral adiposity is an additional layer of complexity[23,71]. This implies an “obesity-paradox” type effect and raises the question of whether optimal visceral fat cut-offs can vary depending on the BMI stratum. Therefore, the combination of these findings identifies a “dual-edged” fat phenotype where accumulation of ectopic and visceral fat increases risk but preserves of visceral stores in non-overweight cohorts indicates potential critical energy storage mechanisms supporting patient resilience during surgical and oncologic stresses[8,9]. In the larger body-composition context, CT-defined sarcopenia (i.e., low skeletal muscle mass, decreased muscle density, and/or composite measures such as skeletal muscle gauge) has emerged as an independent predictor of complication, readmission, and survival in all stages of disease even in the modern era of minimal invasiveness and robotics.

Both cachexia and GLIM-defined malnutrition define high-risk groups with some overlap but also with considerable differences between them. The prognostic value of cachexia vs malnutrition varies over the course of an illness, and both could be further defined by adding strength tests, radiomic assessments, and inflammation biomarkers. Post-gastrectomy patients who suffer from both malnutrition and sarcopenia are at special risk of compromised recovery and poor long-term prognosis[46,52].

Therefore, composite indices which simultaneously assess nutrition, liver function, and muscle mass such as the s-ALBI-index, and composite indexes that include CT-SMI with nutritional scores (GNRI/PNI) provide a more precise and accurate evaluation of physiologic resilience than do individual parameter-based markers. Additionally, composite score indexes that incorporate muscle mass into either inflammatory/immunoglobulin-based markers (SMI-IgM scores/APR-sarcopenia/SFD/LMR), typically yield prognostic capabilities similar to or superior to those provided by TNM staging using routine CT scans and standard lab results[6,13,35,40,57]. Furthermore, interest in assessing longitudinal changes in body composition has grown. Although quantitative CT-assessed longitudinal changes in bone mineral density, visceral adiposity, and intramuscular fat after gastrectomy and/or during/neoadjuvant therapy demonstrated that the trajectory of these measures provides additional prognostic value beyond initial clinicopathological parameters, their application is currently limited due to: Retrospective data collection, single-center designs, relatively small sample size, lack of control of diet and physical activity, and variability associated with whether changes at the L3-level accurately reflect whole-body composition[5,13,33,40,52,57,58].

The potential future direction of this area of study will continue far past a one-dimensional approach using single task segmentation networks. The fully automated nature of today’s increasingly common, end-to-end segmentation pipelines will eventually replace manual methods (L3 level) for tracing semi-manually and eliminate much of the inter-observer variability associated with body composition measurements, allowing for scalable body composition phenotyping to occur at non-specialized research centers. With foundation models being developed to train on large amounts of diverse image data, it seems possible that there may be generalized body-composition and tumor radiomic feature extraction from images, and that these models may have the ability to be “fine-tuned” to use different institution or scanners without requiring the substantial retraining needed to develop specific convolutional networks. Additionally, multimodal AI architectures are starting to demonstrate that integrating CT-derived body composition/radiomic information along with genomic/transcriptomic/proteomic data, can potentially lead to better diagnostics and prognostics than using single modality architecture in oncology, and this same type of paradigm could also improve pre-operative risk assessment in patients undergoing surgery for gastric cancer[95]. In addition to advancements mentioned above, automatic extraction/structuring tools utilizing large language models are emerging as tools to assist in automating the reporting process related to body-composition/radiomic findings from routine radiology reports/images. These tools will help reduce the barriers to workflow limiting the use of these metrics in daily clinical practice[96].

Although internal and external validation methods have a number of benefits, they do come with significant limitations which include mono-ethnic cohorts, retrospective study design, analyses limited to the third level (L3), manual segmentation, and “black-box” interpretability issues. In addition to the above tools, more traditional CT-derived markers such as psoas muscle density or broader body-composition metrics may now be readily identified through access to standard Picture Archiving and Communication Systems networks with no requirement for specialized software. Such markers offer potential to provide accessible prognostic information that enhances the utility of existing nutritional assessments including PNI and GNRI[56].

These restrictions justify much greater evaluation than the simple listing would imply. The significant variation in reported SMI cut-points likely stems from multiple interacting factors instead of a single error in methodology. A recently re-evaluated analysis that applied all 14 previously defined SMI cut-point definitions to a single standardized oncology data set resulted in varying levels of sarcopenia prevalence ranging from 8.9% to 69.8%, and only 3 of the 14 SMI cut-points had a statistically significant association with survival[97]. This illustrates that different reference populations and measurement techniques can result in both varied prevalence estimates and different prognostic implications; similar large variations have been documented in meta-analyses examining CT-defined sarcopenia in other gastrointestinal cancers (I² > 0.85)[98]. Additionally, some of the variability may represent real biologic differences in muscle mass and fat distribution among ethnic groups rather than solely due to artifacts related to measurements. For example, SMI thresholds established based upon a population-level derivation in Asian cohorts are substantially lower than those of their Western counterparts, and better predictive value for mortality in Asian patients with cancer is demonstrated by these same thresholds[99]. A 2025 systematic review of the genetic basis of susceptibility to sarcopenia indicates that polymorphisms in muscle regulatory genes such as ACTN3 and MSTN (which signal via the activin receptor pathway) demonstrate different allelic frequency distributions and different relationships with decreased muscle mass in cohorts of Koreans/Japanese vs primarily Caucasians. While the authors suggest that there may be a genetic component contributing to cross-population differences in muscle phenotype, they do not indicate whether or not this genetic component contributes to differential body composition phenotypes in gastric cancer[100]. In addition, while segmenting images is a source of potential errors given the inherent assumptions regarding the reliability of using single-slice measurements at the L3 to accurately estimate total body muscle and fat content, such assumptions have been largely validated using reference populations who were healthy but less well understood in patients with sarcopenic obesity whose subcutaneous and visceral compartments undergo disproportionate changes. The majority of the supporting evidence for this field is also retrospective and based upon studies conducted at a single center with no standardization for a CT-based definition of muscle mass, an issue specifically identified in previous reviews of body-composition metrics in gastrointestinal oncology which allows for the possibility of immortal time-bias when body-composition status is measured at one pre-operative time point and then used to forecast subsequent outcome times. Immortal time-bias is a widely recognized distortion in observational studies generally that has been rarely identified within the literature on body composition until now. Finally, although AI-based segmentation algorithms report high intra-class correlation coefficients against manually-traced expert tracings, such algorithms require additional separate and independent external validation datasets to validate that performance persists beyond their initial development setting; therefore, it is important to evaluate not just how accurate an algorithm performs on its internal testing datasets, but also if generalizability will occur in the clinical environment[101].

The combination of systematic preoperative evaluation of sarcopenia utilizing standard CT imaging along with basic functional measurements (hand-grip strength, etc.) and frailty tools offers a comprehensive framework by which clinicians can develop prognostic models and treatment plans. Studies have shown that this multi-dimensional model can assist clinicians in determining the appropriate regimen or dose of neoadjuvant therapy for at-risk patients. Furthermore, it supports a cautious approach to administering highly toxic treatments in patients with lower levels of myo-hepatic reserves, particularly among older adults where both sarcopenia and frailty are present as overlapping yet separate entities, each of which have an independent relationship to patient prognosis. Structured multidisciplinary perioperative care pathways that incorporate standardized evaluations of sarcopenia and frailty into preoperative workflows demonstrate that interventional strategies [i.e., targeted rehabilitation, optimized nutrition (frequently achieved through oral supplementation or enteral feeding), dental/oral health (“oral frailty”) management, and emotional/social support] can lead to improved short-term outcomes in elderly patients who are at higher risk of experiencing adverse events during surgery, while also identifying that sarcopenia contributes to delayed recovery of complex complications such as esophageal/jejunal anastomosis leak.

The proposed framework for prehabilitation warrants further expansion, as it remains notably brief despite being one of the few factors related to surgery which can be changed. The benefits of structured resistance training, performed 2-3 times per week for 2-4 weeks prior to gastrectomy, in preserving skeletal muscle and improving function at the time of surgery is well established, while aerobic plus resistance training may provide additional benefit with respect to improving cardiovascular reserves relevant to postoperative recovery. Pre-operative nutrition—often defined as either orally consumed, enriched proteins, or oral or parenteral administration of immunonutrition supplements containing arginine, omega-3 fatty acids, and nucleotides—has been demonstrated to both prevent loss of skeletal muscle and reduce risk of infections when used in conjunction with exercise and psychologic support as part of “trimodal” pre-habilitation programs. Postoperative rehabilitation pathways that incorporate the above interventions into the post-surgical period along with early mobilization and standardized discharge planning have also shown to add to the previously noted benefits, but are limited due to an extremely small number of participants and heterogeneity of protocols. Evaluating body composition within the broader context of gastric cancer treatment offers valuable insight into two complementary paradigms: One directed at tumor eradication through cytotoxic therapy (chemotherapy, targeted agents, and cellular/immunebased approaches), and the other focused on the host and its microenvironment—encompassing nutrition, muscle stores, systemic inflammation, and the tumor stroma—where prehabilitation and bodycompositionguided treatment intensity adjustments represent the latter approach[102].

However, the overall quality of the current literature is generally poor to fair, owing to several limitations: Wide variability in established SMI cutoffs, a predominance of single-center retrospective designs, lack of representation of non-Asian and Western population groups, and potential residual confounding by BMI, tumor stage, patient age, and comorbidities. Consequently, it will be necessary for researchers to conduct larger-scale prospective multi-institutional studies[16,22,36,41,60,74,76,103,104]. These types of studies will help determine whether corrective actions taken to address the presence of preoperative sarcopenia and other negative body-composition-related conditions can lead to improved outcomes including survival. In addition, these studies may provide an opportunity to develop international consensus regarding standardized diagnostic criteria—potentially with use of AI informed algorithms—for the assessment of sarcopenia, sarcopenic obesity, and body composition-related factors[5,6,13,18,33,35,38-40,46,52,57,58,63,87,91].

CONCLUSION

Sarcopenia has been identified as an independent and clinically relevant factor influencing patient outcomes following gastrectomy. Moreover, sarcopenia appears superior to the more traditional clinical predictors, such as cancer stage and operative variables, in predicting outcomes from gastrectomy. Indeed, the development of sarcopenia as a reliable predictor of poor prognosis in cancer is evidenced by a number of recent studies, which demonstrate a correlation between sarcopenia and poorer postsurgical outcomes and tolerance to multi-modality treatment strategies.

It is important to note that sarcopenia is no longer perceived as a uni-dimensional condition but rather a complex process, the impact of which on outcomes from cancer surgery and multimodality therapy cannot be underestimated. Indeed, muscle function and muscle quality can be argued as being more informative than skeletal muscle mass itself. In addition, body composition changes dynamically in relation to perioperative events and has a direct effect on patient outcomes. New classifications, including sarcopenic obesity and myosteatosis, can contribute to the identification of at-risk patients.