This scoping review aimed to map the application of symptom network analysis among patients with gastric cancer, summarizing study characteristics, assessment instruments, identified core symptoms and bridge symptoms, temporal symptom relationships, and factors associated with core symptom identification. The goal was to inform precision and stage-specific symptom management by synthesizing available evidence.
The review followed the Arksey and O’Malley framework and reporting aligned with PRISMA-ScR. Nine databases (including PubMed, Web of Science, CINAHL, Cochrane Library, Embase, CNKI, Wanfang, VIP, and CBM) were searched from inception to April 1, 2026. Eligible studies were original research in Chinese or English applying symptom network analysis to gastric cancer populations (including mixed cancer samples where gastric cancer patients were explicit). Studies were eligible for perioperative, chemotherapy, or postoperative chemotherapy contexts. Reviews, abstracts, non-original research, and records without full text were excluded.
Nine studies met inclusion criteria: five cross-sectional and four longitudinal. Seven studies were conducted in China and two in South Korea. Study populations primarily comprised patients receiving chemotherapy, those undergoing postoperative chemotherapy, and perioperative patients. Two studies included mixed cancer cohorts with gastric cancer patients; these were treated as indirect evidence because gastric-cancer-specific networks could not be isolated. A total of 669 records were initially identified, with nine ultimately included after screening.
Six different symptom assessment instruments were used across the included studies. The Chinese version of the MDASI-GI-C (MD Anderson Symptom Inventory–Gastrointestinal Cancer Module) was the most frequently applied instrument (reported in 7 studies). Other instruments included the MDASI, Hospital Anxiety and Depression Scale (HADS), Functional Assessment of Cancer Therapy–Gastric (FACT-Ga), a 20-item Symptom Severity Scale, and the Classification and Determination of Traditional Chinese Medicine Constitution (used in three studies).
Assessment timing varied by treatment context. Perioperative studies measured symptoms at multiple pre- and postoperative points (for example, 1 week before surgery, 1 week after surgery, 7 days postoperatively, and 3–6 months postoperatively). Chemotherapy studies assessed symptoms before and after chemotherapy cycles or across cycles (examples included after the first through sixth cycles or from chemotherapy initiation to discharge). Two cross-sectional studies did not specify assessment time points.
Seven studies explicitly identified core symptoms in gastric cancer symptom networks. Across studies, nine core symptoms were reported. Fatigue was the most frequently reported core symptom (identified in 5 studies). Pain and vomiting were each reported as core symptoms in three studies. Sadness, altered taste, and loss of appetite were each identified in two studies; shortness of breath and some quality-of-life–related symptoms appeared in one study each.
Core symptoms varied by treatment stage and symptom-burden subgroup. For example, on postoperative day 7, vomiting, fatigue, and pain were primary core symptoms in one study, whereas altered taste, vomiting, and loss of appetite were core symptoms among patients receiving postoperative chemotherapy in another. Differences were also observed according to overall symptom burden levels.
Three studies reported bridge symptoms linking distinct symptom communities. Ten bridge symptoms were identified across these reports, including sleep disturbances, loss of appetite, somnolence, altered taste, dry mouth, abdominal distension, anxiety, depression, emotional well-being, and physical health. Findings varied considerably across studies: one study identified sleep disturbances, loss of appetite, somnolence, and altered taste during chemotherapy; another found dry mouth as a bridge in early postoperative chemotherapy and abdominal distension in later stages; a third identified anxiety, depression, and emotional well-being as bridge roles across perioperative time points. The heterogeneity in bridge symptom identification highlights variability by stage, sample, and measurement.
Two longitudinal studies examined time-ordered associations using cross-lagged network analysis. Reported temporal paths included sadness → fatigue between the second and fourth chemotherapy cycles, and altered taste → loss of appetite between the fourth and sixth cycles in one study. Another study reported vomiting → loss of appetite from the first to third postoperative chemotherapy sessions, and abdominal distension → nausea from the third to sixth sessions, with altered taste → loss of appetite also prominent. The authors note these cross-lagged associations represent time-ordered statistical links and should not be interpreted as causal effects because the data were observational.
Several studies reported pairs of symptoms with strong within-network associations. Recurrently identified strongly connected pairs included altered taste–loss of appetite and nausea–vomiting. Strong associations were most common among gastrointestinal symptoms, eating-related symptoms, and psychological/emotional symptoms, reflecting clinically relevant symptom interactions in gastric cancer.
All nine included studies used R for network estimation and visualization. Five studies used strength centrality to identify core symptoms. Reporting of network stability and accuracy was inconsistent: three studies explicitly reported stability and accuracy metrics. Methodological features charted included network model, software/packages, centrality and bridge indicators, temporal indicators, and whether stability/accuracy were reported. The review highlighted methodological heterogeneity across studies in network estimation approaches, indicator selection, assessment time points, and reporting practices.
The reviewed evidence indicates that symptom profiles in gastric cancer are complex, dynamic across treatment stages, and heterogeneous across populations. The authors suggest symptom management may benefit from individualized, stage-specific approaches targeting identified core symptoms, bridge symptoms, and key symptom pathways (for example, altered taste → loss of appetite). Recommendations for future research included standardizing symptom assessment tools and time points, conducting multicenter longitudinal studies to clarify temporal symptom dynamics, and developing and evaluating interventions informed by symptom network findings for clinical practice.
This scoping review intentionally did not perform a formal risk-of-bias appraisal because its aim was to map the breadth and methods of existing studies rather than synthesize effect sizes. Two included studies used mixed cancer samples and were treated as indirect evidence. Details on specific network models, exact centrality values, and full methodological parameters were charted in the source review’s tables; where such granular reporting varied across studies, the review noted inconsistency. The cross-lagged temporal findings are observational and should be interpreted as time-ordered statistical associations rather than causal pathways.