Cross-Lagged Panel Network Analysis of Symptoms in Patients with Gastric Cancer Undergoing Postoperative Chemotherapy

To develop temporal symptom networks at three postoperative chemotherapy time points and investigate the longitudinal relationships between 18 symptoms in patients with gastric cancer undergoing postoperative chemotherapy in China.

Symptom prevalence and severity were measured using the M. D. Anderson Symptom Inventory (MDASI) and Gastrointestinal Cancer Module of the MDASI (MDASI-GI) at T1, T2, and T3 on the day after the first, third, and sixth chemotherapy sessions, respectively. Cross-lagged panel network (CLPN) models were employed to examine the temporal dynamics of the 18 symptoms and their interrelationships.

In total, 379 participants were included. Dry mouth had the highest out-prediction (r = 0.101) and out-strength (r = 0.863) values during T1→T2. The strongest direct effect was observed for the change in taste → lack of appetite (β = 0.38) during T2→T3. Feeling bloated had the highest values for out-strength (r = 0.910), out-prediction (r = 0.215), and bridge strength (r = 1.010) during the T2→T3 period. The two CLPNs showed medium to high stability based on the centrality stability coefficients of out-strength and in-strength.

Lack of appetite can be improved during chemotherapy by managing vomiting and taste changes. Attention should be paid to the dry mouth and feeling bloated, as they are the strongest predictors in the early and middle stages of treatment.

Understanding the relationship between symptoms during chemotherapy in patients with postoperative gastric cancer can help clinicians identify targets for intervention at different times.