Introduction

Gastric cancers are the third leading cause of cancer death and the fifth most common cancer in the world, with half of all cases occurring in Eastern Asia, including Japan [1]. Solid tumours, including gastric cancers, are typically associated with a degree of anorexia as well as underlying metabolic alterations such as elevated energy expenditure, excess catabolism, and inflammation. In gastric cancers, these effects are worsened by the direct obstruction to the passage of food and combined effects of chemotherapy and major gastrectomy that results in a decline in nutritional intake [2, 3].

Recently, there has been growing interest relating to body composition assessment in patients diagnosed with gastric cancer for nutritional evaluation and prognostication. Measurement of skeletal muscle and/or adipose tissue have found success in a number of other areas of medicine for nutritional assessment [4,5,6,7], further supporting an argument for such measures to become part of clinical practice. There are, however, large variations in methods of assessment of body composition parameters. These include CT scans which are routinely used in clinical practice and have a high precision and specificity for muscle and fat distribution (visceral, intermuscular, and subcutaneous). Other measures less frequently used in cancer patients include dual-energy X-ray absorptiometry (DXA) and bioelectrical impedance (BIA).

To further delineate the clinical utility of body composition assessment in patients with gastric cancers, it is considered timely that existing evidence are reviewed. This systematic review sought to summarise current literature surrounding body composition assessment in patients with gastric cancer and assess its impact on peri-operative outcomes and long-term survival.

Methods

Search Strategy

A systematic search of PubMed, EMBASE and the Cochrane Library databases were conducted on the 22nd June 2018 by two independent investigators. The search terms used were ‘gastric cancer’ or ‘stomach cancer’, and ‘sarcopenia’ or ‘body composition’ or ‘muscular atrophy’ individually or in combination. Search terms used for this review is presented as shown in Supplementary Table 1. The ‘related articles’ function was used to broaden the search, and all citations were considered for relevance. A manual search of reference lists in recent reviews and eligible studies was also undertaken. This paper is reported according to the PRISMA guidelines [8].

Inclusion and exclusion criteria

Inclusion criteria were: (1) studies reporting the assessment of body composition (by any method) in human subjects with gastric cancer (receiving palliative or curative treatment); (2) published in the English language. Exclusion criteria were: (1) Conference abstracts, review articles, and case reports (< 5 patients); (2) publications with mixed populations where the outcomes of patients with either benign disease or cancers at another site could not be separated from those of patients with gastric cancers. After excluding duplicates, two researchers (SKK, BHT) independently reviewed the titles and abstracts of studies identified by the literature search. Where a study was considered to be potentially relevant to the research question a full copy of the publication was obtained for further review. The reference lists of all included studies were hand-searched in order to identify other potentially relevant studies. Any areas of disagreement between the two primary researchers were resolved through discussion.

Study outcomes

The primary outcome measure was post-operative complications such as overall complications (Grade I–V) and major complications (≥ Grade III) reported according to Clavien–Dindo Classification [9]. Secondary outcome measures were survival (overall, recurrence-free and cancer-specific) and surgery-specific complications (pulmonary, cardiac, surgical site infections).

Data extraction

One researcher (SKK) extracted data on study characteristics (author, year of publication, country of origin, study design, patient number), patient demographics (age, sex, AJCC T Classification, AJCC Stage), method and details of body and reported clinical outcomes. Body composition measures were abstracted according to method of assessment (CT, DXA, BIA), definition and cut-off points defining sarcopenia and obesity.

Definitions

Gastric cancer is a malignancy of any portion of the stomach. Body composition assessment is any method of reporting either the volume or characteristics of muscle and/or adipose compartments within the body. Sarcopenia is defined as severe depletion of skeletal muscle mass that has been defined by a range of criteria that are specific for the method of assessment. Cachexia is a multifactorial syndrome characterized by ongoing loss of skeletal muscle mass (with or without loss of fat), that is not fully reversible using conventional nutritional support and that eventually leads to functional impairment [10, 11].

Assessment of methodological quality

Methodological quality and standard of outcome reporting within included studies were assessed by two independent researchers (SKK, BT). Methodological quality was formally assessed using the Newcastle-Ottawa score for cohort studies (SKK, JB).

Statistical analysis

This systematic review and meta-analysis were conducted in accordance with the recommendations of the Cochrane Library and MOOSE guidelines [12]. For categorical variables, analysis was performed by calculating the odds ratio (OR). For survival data, analysis was performed by calculating the logarithm of the hazard ratio (HR) with 95% confidence intervals (CI95%). HR and its variance were extracted directly from the published manuscript. Where these data were not available it was determined through additional calculations that were dependent on the data presented by the study: annual mortality rates, survival curves, number of deaths, or percentage freedom from death [13]. The random effects, the DerSimonian–Laird method was used for the meta-analysis of outcomes. Funnel plots were used to visually assess publication bias of included studies. Heterogeneity between studies was assessed using the I2 value to determine the degree of variation not attributable to chance alone. I2 values were considered to represent low, moderate, and high degrees of heterogeneity where values were < 25%, 25–75%, and > 75%, respectively. Funnel plot asymmetry was assessed using the Egger test. Statistical significance was considered when p < 0.05. Statistical analyses were performed using the RevMan 5.3 software (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2011) and Stata 15 (Version 15.1, StataCorp, College Station, Texas).

Results

Study Characteristics

The literature search identified 39 studies reported according to the PRISMA guidelines as shown in Fig. 1. Majority of studies identified were from Asia (n = 33) [14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46] followed by Europe (n = 5) [5, 47,48,49,50], and North America (n = 1) [51].

Fig. 1
figure 1

Flow diagram of included studies

Full size image

Reporting standards and methodological quality

Criteria for defining parameters of body composition varied considerably (Table 1). CT cut-points defining sarcopenia were reported in 21 studies, all of which adopting sex specific cut-off points. Of the 21 studies, only 18 studies reported normalisation of these measurements to height (n = 17) and body surface area (n = 1) [5, 17, 19, 21, 23, 29, 33, 34, 37, 38, 40, 41, 43,44,45, 49, 51]. Studies were also heterogeneous with respect to several factors known to influence body composition: race, sex, BMI, and tumour characteristics (Table 2). Variation in these factors was broadly reflective of known differences that exist between Asian and Western populations. Study quality was assessed using NOS and a breakdown of each components are presented under Supplementary Table 2. A summary of studies reporting impact of pre-operative sarcopenia on morbidity and mortality as Supplementary Table 3.

Table 1 Details of body composition assessment methodology
Full size table
Table 2 Details of included studies
Full size table

Body composition assessment in gastric cancer

Methods of assessment used in each study, including criteria for key descriptors of body composition are presented in Table 1. CT (n = 26 studies) [5, 17, 19,20,21,22,23,24, 26, 29,30,31, 33,34,35,36,37,38, 40, 42,43,44,45,46, 49, 51] and BIA (n = 9) [14,15,16, 18, 27, 28, 39, 41, 47] were the most commonly reported methods of body composition assessment, while DXA was utilized in three studies [32, 48, 50, 52]. Of the 25 studies published since January 1, 2016, CT was the chosen method of assessment in 19 (76%). A total of 3193 patients were evaluated (CT, n = 5999; BIA, n = 1366; DXA, n = 86) (Table 2). Assessment was predominantly performed in patients with resectable gastric cancer prior to surgery either before and/or after neoadjuvant therapy. The reported prevalence of sarcopenia in preoperative gastric cancer patients in 21 studies varied between 7% and 70%. Prevalence values can only be compared between studies that used identical criteria for sarcopenia in populations with similar characteristics; other comparisons are unreliable owing to differences in methodologies, patient populations, and criteria for defining this parameter (Tables 1, 2).

Postoperative morbidity and mortality

Impact of preoperative sarcopenia on overall complications was reported in 14 studies [18, 23, 25, 29, 31, 33, 34, 38,39,40,41, 44, 45, 51]. Preoperative sarcopenia was associated with significantly higher rates of overall postoperative complications (33% vs 23%, OR 2.18, CI95% 1.49–3.20, p < 0.001, I2 81%) (Fig. 2a). Impact of preoperative sarcopenia on major complications was reported in 13 studies [17, 18, 23, 29, 31, 33, 34, 38, 40, 41, 44, 45, 49]. Preoperative sarcopenia was associated with significantly higher rates of major postoperative complications (13% vs 7%, OR 1.67, CI95% 1.14–2.46, p = 0.009, I2 53%) (Fig. 2b). Impact of preoperative sarcopenia on pulmonary complications was reported in 9 studies [18, 23, 25, 29, 33, 39,40,41, 44, 52]. Preoperative sarcopenia was associated with significantly higher rates of postoperative pulmonary complications (13% vs 4%, OR 4.01, CI95% 2.23–7.21, p < 0.001, I2 47%) (Fig. 2c). Impact of preoperative sarcopenia on cardiac complications was reported in 7 studies [18, 23, 25, 33, 40, 41, 44]. Preoperative sarcopenia was associated with significantly higher rates of postoperative cardiac complications (4% vs 2%, OR 4.11, CI95% 1.54–11.02, p = 0.005, I2 26%) (Fig. 2d).

Fig. 2
figure 2

Summary meta-analysis of studies reporting impact of pre-operative sarcopenia on post-operative outcomes in patients undergoing gastrectomy. a Overall, b major, c pulmonary complications, d cardiac complications, e in-hospital mortality

Full size image

Impact of preoperative sarcopenia on mortality was reported in ten studies [18, 23, 25, 29, 33, 38, 40, 44, 45, 49]. Preoperative sarcopenia was associated with significantly higher rates of mortality (2% vs 1%, OR 2.17, CI95% 1.06–4.43, p = 0.03, I2 0%) (Fig. 2e). Egger regression testing suggested publication biases were minimal for reporting of overall complications (p = 0.219), major complications (p = 0.655), pulmonary complications (p = 0.503), and cardiac complications (p = 0.220) and mortality (p = 0.612) (Supplementary Fig. 1).

There was no increased risk of non-pulmonary postoperative complications in patients with sarcopenia, including: anastomotic leak (nine studies: OR 1.09, CI95% 0.62–1.94, p = 0.76, I2 0%) (Supplementary Fig. 2) [18, 23, 25, 29, 33, 37, 39, 40, 44]; surgical site infections (nine studies: OR 0.97, CI95% 0.69–1.37, p = 0.87, I2 0%) (Supplementary Fig. 3) [18, 23, 25, 33, 37, 38, 40, 41, 44]. These relationships remained non-significant following reanalysis of only those studies utilizing CT for the assessment of sarcopenia. Egger regression testing suggested publication biases were minimal for reporting of anastomotic leak (p = 0.514) and surgical site infections (p = 0.409) (Supplementary Fig. 1).

Long-term survival

The influence of preoperative sarcopenia on the overall survival of patients following treatment for gastric cancer was reported in nine studies [19, 21, 25, 29, 31, 34, 38, 43, 45]. Meta-analysis of outcomes presented within these studies identified that sarcopenia was associated with lower overall survival (HR 2.12, CI95% 1.89–2.38, p < 0.001) (Fig. 3a). Heterogeneity between studies was moderate (I2 37%). The influence of preoperative sarcopenia on the recurrence-free [34, 43, 45] and cancer-specific survival [25, 34, 38] of patients following treatment for gastric cancer was reported in three studies respectively. Meta-analysis of outcomes presented within these studies identified that sarcopenia was associated with lower recurrence-free survival (HR 2.12, CI95% 1.82–2.47, p < 0.001, I2 40%) (Fig. 3b) and cancer-specific survival (HR 2.00, CI95% 1.54–2.59, p < 0.001, I2 0%) (Fig. 3c). Egger regression testing suggested publication biases were minimal for reporting of overall survival (p = 0.315), recurrence free survival (p = 0.716) and cancer specific survival (p = 0.800) (Supplementary Fig. 1).

Fig. 3
figure 3

Summary meta-analysis of studies reporting impact of pre-operative sarcopenia on a overall, b recurrence-free, c cancer-specific survival in patients undergoing gastrectomy

Full size image

Sensitivity analysis on adjusted outcomes

Sensitivity analysis of studies reporting multivariable analysis on the influence of pre-operative sarcopenia on overall complications, major complications and overall survival were performed (Supplementary Fig. 4). Three studies reported adjusted odds ratio for overall [17, 21, 44] and major [18, 40, 45] complications. Meta-analysis of odds ratios derived from multivariable analysis found pre-operative sarcopenia to be independently associated with poorer overall complications (HR 2.50, CI95% 1.68–3.72; p < 0.001, I2 3%; Fig. 3a) and major complications (HR 3.62, CI95% 2.34–5.59; p < 0.001, I2 0%, Fig. 3b).

Seven studies reported adjusted hazard ratios for overall survival [14, 19, 25, 29, 31, 34, 38, 43, 45]. Meta-analysis of hazard ratios derived from multivariable analysis found pre-operative sarcopenia to be independently associated with poorer overall survival (HR 1.76, CI95% 1.53–2.03; p < 0.001, I2 14%, Fig. 3c). Other factors controlled for by multivariable analysis included age, gender, American Society Anaesthesiology (ASA) Grade, Charlson co-morbidities, diabetes, tumour histology, operation type and American Joint Comission on Cancer stage (T, N, M and overall).

Discussion

Weight loss and malnutrition remain a central concern for patients at all stages of treatment for gastric cancer. At the time of diagnosis more than half of the patients with gastric cancer present with some degree of weight loss due to underlying tumourigenesis [53]. The inability to maintain body weight is recognised as a poor prognostic factor in sustaining therapeutic response to neoadjuvant or adjuvant treatment and long-term survival [54, 55]. A wide range of approaches have been utilised to assess nutrition in patients with gastric cancers such as anthropometric measures, blood markers, measures of energy expenditure, validated nutritional risk scores, and patient reported dietary history. However, the current review focuses on the assessment of body composition due to the promise that this approach has demonstrated in other areas of medicine and surgery as well as the relative ease with which it might form part of future routine clinical practice [56].

Although four techniques for body composition assessment in patients with gastric cancer have been described, CT and BIA were most commonly used within studies identified in the review. However, none of the studies included in this review discussed the cost of prospectively assessing body composition in patients with gastric cancer. Also, variation in assessment and lack of standardized definition for body composition parameters remain a barrier to adoption into routine clinical practice. DXA is widely considered to be the gold standard non-invasive method of measuring fat mass (FM) and fat free mass (FFM) [57]. The use of DXA is limited in some situations due to the cost of equipment, the need for trained operators, the lack of portability, and the need for exposure to ionizing radiation. CT also requires expensive equipment and trained operators and exposes the patient to a high dose of radiation. The analysis of body composition from CT necessitates specialist software and training. BIA is practical, non-invasive, and easy to perform. However, studies evaluating BIA have provided inconsistent findings, with some reporting good accuracy [58, 59], whereas others have shown poor results, particularly in obese populations [60, 61].

CT scans form a regular part of the standard management of gastric cancer patients and hence offers an opportune method for the assessment of body composition at specific and relevant time-points in the patient treatment pathway. This avoids exposing the patient to repeated imaging and high radiation doses; and it is the authors’ recommended method for body composition analysis. A recent study has found excellent correlation between results obtained from DXA and analysis of CT scans [62]. The definition of sarcopenia, a state of severe of depletion of skeletal muscle mass (and function), have been largely established using CT measures and is associated with an increased risk of mortality in many malignant conditions [63, 64]. This review has highlighted that many studies fail to account for variation introduced by factors such as sex, race, and tumour stage when interpreting and reporting parameters of body composition. Where such controls do not exist it becomes impossible to compare individual studies and to draw meaning that is generalizable.

As sarcopenia is associated with higher rates post-operative morbidity and mortality and poor long-term survival, perioperative interventions are important to reduce these risks. One study in this review reported benefits of oral nutrition supplement to reduce weight loss in the post-operative setting [28]. However, this study did not evaluate its impact on surgical outcomes and long-term survival. Pre-operative rehabilitation is a widely growing concept and adopted into Enhanced Recovery After Surgery (ERAS) protocols across a variety of major surgeries [65]. In gastrectomy for gastric cancer, several studies have reported the improved post-operative outcomes and shorter hospital stay in patients receiving ERAS management as compared to those receiving conventional care [66,67,68].

The overall strength of conclusions from this review is limited by the heterogeneity of the studies with regards to methods and techniques of assessment of body composition. Subgroup analyses of outcomes, specifically survival, by cancer stage was not possible due to variability in reporting, which is an important confounding factor to note. Furthermore, the absence of studies assessing patients with advanced metastatic gastric cancer who are receiving palliative therapy is also recognized. Finally, a number of other studies were not suitable for inclusion in this review as they reported body composition assessment in a mixed cohort including patients with oesophageal cancer. Collaborative international efforts should now seek to better establish patterns of variation in body composition in gastric cancer patient from disparate racial and geographical backgrounds to establish consensus guidelines. Specifically, the adoption of either a continuous or graded scale for determining sarcopenia and features of adiposity should also be considered.

Conclusion

This systematic review and meta-analysis have shown higher rates of morbidity and in-hospital mortality following gastrectomy in sarcopenic patients prior to surgery; and an association between sarcopenia and reduced overall, recurrence-free and cancer-specific survival in patients undergoing gastrectomy. Similar findings have been reported for other solid organ tumors [69,70,71]. This review highlights the need for standardised assessment of body composition as it has the potential to support future decision-making in patients with gastric cancer. It also offers fresh impetus for future studies to further understand underlying mechanisms driving malnutrition with poor outcomes in this patient population. With lack of consensus in regard to optimal methodology and reporting standards, future efforts should be focused at establishing consensus guidelines for body composition assessment in gastric cancer.