In this cohort study of gastric cancer patients, we found that high scores of PG-SGA and NUTRISCORE assessed at 2 or 3 months after gastrectomy were associated with poor survival. However, high PG-SGA and NUTRISCORE scores at 1 month after gastrectomy were not associated with poor survival, suggesting that an early evaluation of the PG-SGA and NUTRISCORE right after surgery may not predict prognostic outcomes among gastric cancer patients who underwent surgery. In our study, nutritional assessment at 1 month after surgery was mostly performed before the start of chemotherapy, and assessment at 2 or 3 months after surgery was done during chemotherapy. We found that the risk of mortality in 9–11 scores of PG-SGA was significantly higher, but not that in 12 or more at the second consultation. Whether PG-SGA scores have a non-linear association with mortality risk needs further investigation.
Gastric cancer patients undergoing postoperative chemotherapy often experience chemotherapy-related adverse effects, including nausea/vomiting, anorexia, diarrhea, and stomatitis, leading to reduced food intake, weight loss, and malnutrition . Postoperative gastric stasis, known as delayed gastric emptying, may worsen these symptoms in gastric cancer patients [26, 27]. Therefore, a proper diagnosis of malnutrition is essential to provide adequate nutrition support and intervention for cancer patients . In particular, it emphasizes the importance of nutritional intervention to improve eating after gastrectomy and prevent malnutrition and excessive weight loss .
The PG-SGA comprises clinician and patient-centered assessments . The PG-SGA has been recommended by the Oncology Nutrition Dietetic Practice Group of the Academy of Nutrition and Dietetics as the standard for nutrition assessment for cancer patients [21, 28]. A few cohort studies have reported on the prediction capacity of the PG-SGA for survival among gastric cancer patients. A retrospective cohort study of 256 Taiwanese patients with metastatic gastric cancer showed that PG-SGA assessed within one week before chemotherapy was independently associated with survival. In that study, HR (95% CI) for severely malnourished patients (PG-SGA C) compared to patients with PG-SGA A/B was 2.73 (95% CI 1.73–4.29) . Another cohort study of 120 palliative care patients in Brazil, including the patients with gastrointestinal tumors (25.8%, n = 31), evaluated the PG-SGA within 24h of hospitalization and found that one score increase in the PG-SGA scores was associated with a 4% increase in mortality . In a Brazilian prospective study of 178 patients with gastric and colorectal cancer, malnourished status assessed by PG-SGA was associated with a 2.9 times higher risk of overall mortality . Those studies included patients with metastatic gastric cancer or palliative care, whereas we included the patients who underwent curative gastrectomy. Our current study of patients who underwent curative gastrectomy also showed a higher risk of mortality among patients with high PG-SGA scores assessed at 2 months after surgery.
The NUTRISCORE, a newly developed cancer-specific nutrition assessment tool, has not been previously evaluated regarding cancer survival. However, a few studies recently reported comparing the NUTRISCORE with other nutritional assessment tools [11, 16, 33]. To the best of our knowledge, no studies have reported the association with survival through multiple nutritional assessments using PG-SGA or NUTRISCORE during the post-gastrectomy follow-up period. The reason why we observed a lower risk of mortality with high NUTRISCORE scores assessed at 1 month after surgery is unclear. However, it is possible that scoring early after surgery may not reflect nutritional status associated with survival, or scoring based on weight loss, appetite, and treatment may not be desired near the start of anticancer therapy. We observed that increase in NUTRSCORE scores at 3 months after surgery was associated with high mortality among gastric patients who underwent curative gastrectomy and received at least 1 cycle of chemotherapy. Our findings may warrant further research.
Regarding the timing of the nutritional assessment in relation to survival, only a few previous studies evaluated nutritional status more than once. Cancer survival studies that used PNI or NRI calculated by a nutrition-related biochemical marker such as albumin and total lymphocyte count (TLC) or weight loss [17, 18] reported that PNI or NRI might be a potential prognostic factor for cancer prognosis. Results of the nutritional evaluation with the NRI after gastrectomy in 760 Japanese patients with stage I-III gastric cancer showed that the overall survival rate was lower among the malnourished group (NRI ≤ 97.5) before the gastrectomy (HR = 1.68; 9% CI: 1.14–2.48), at 1 month (HR = 1.77; 9% CI: 1.22–2.56), at 3 months (HR = 2.18; 9% CI: 1.49–3.21), at 6 months (HR = 1.81; 9% CI: 1.23–2.65), and at 12 months (HR = 2.17; 9% CI: 1.43–3.29) after gastrectomy. Also, malnutrition at 1 month (HR = 1.73, 9% CI: 1.06–2.83) and at 3 months (HR = 1.98, 9% CI: 1.20–3.28) after gastrectomy was significantly associated with poor cause-specific survival . A retrospective study of 1,415 Korean gastric cancer patients who underwent gastrectomy in our hospital found that a decline in PNI scores from preoperative to postoperative 3 months was associated with a 1.53 times higher risk of death, but the PNI scores assessed before surgery or 6 or 12 months after surgery were not associated with risk of death .
In the present study, we found that malnutrition at a median of 2 and 3 months after surgery was associated with a poor survival. There was no significant association for the PG-SGA or an inverse association for NUTRISCORE assessed at 1 month. Although the reasons were not clear, nutritional status right after surgery could be affected by gastrointestinal symptoms and dietary intake reduction due to the surgical procedure and acute surgical stress and short-term conditions. Because nutritional status immediately after surgery may reflect surgical stress, it may not sufficiently predict the long-term survival of patients with stage II and III gastric cancer. Further cohort studies are needed on whether there is a difference in the association between malnutrition and cancer survival according to the nutritional evaluation timing. Also, it is possible that it takes time to detect progressive malnutrition status associated with survival given continued weight loss .
Our study has several limitations. First, because the 2nd and 3rd nutritional consultations were conducted only when the patient wanted, we did not have all the initial group of patients at the 2nd and 3rd consultations compared to the 1st consultation. However, we found similar clinical characteristics, including disease progression and stage, between patients who participated in the 2nd and 3rd consultations and those who did not. Also, when we included only patients who received all three nutritional consultations, we observed an increase in mortality with high scores. Second, we cannot rule out the possibility of potential unknown or residual confounding factors. Third, our results may not be generalizable to all gastric cancer patients; nevertheless, given our study center receives referral patients from all over the country, the generalizability may not be an issue. The strength of our study is examining the association between mortality and malnutrition evaluated according to the timing of at least one postoperative nutritional consultation in the same group of gastric cancer patients. Our study provides insight that efforts should be made to improve survival by intensive nutritional intervention for malnutrition patients during initial 2–3 months of chemotherapy after gastrectomy.