Abstract
Background: This study was designed to assess the prognostic role of body weight loss (BWL) after curative gastrectomy for gastric cancer patients. Patients and Methods: We analyzed a total of 102 consecutive patients who underwent curative gastrectomy for stage II or III gastric cancer between 2005 and 2011. Postoperative BWL during the first 6 months in patients was calculated from their hospital records. Results: Postoperative BWL was significantly higher in patients with tumours in the upper stomach, total gastrectomy, or subsequent recurrence. A postoperative BWL of 12% was the optimal cut-off value, which was strongly associated with subsequent recurrence. Patients with a postoperative BWL of 12% or higher did not have poorer overall survival but did have significantly poorer disease-free survival than patients with BWL of less than 12%. Cox proportional hazard analyses revealed that a postoperative BWL of 12% or higher was an independent prognostic factor of poor disease-free survival. Conclusion: High BWL after curative gastrectomy is related to recurrence in patients with gastric cancer.
Gastric cancer is one of the most common causes of death from cancer worldwide (1). Because of recent advances in diagnostic techniques, less invasive treatment, and perioperative management, earlier detection of gastric cancer has been achieved and mortality and morbidity have decreased (2-6). Patients with advanced-stage disease, however, still have a poor prognostic outcome and a high incidence of recurrence. There is no doubt that tumour-related pathological factors, such as lymph node metastasis, depth of tumour invasion and lymphatic invasion, are recognized as pivotal prognostic factors. In addition, several recent studies demonstrated that such factors have a great influence on the prognosis of gastric cancer based on inflammation status (7), age and comorbidity (8), malnutrition (9), and complications (10). As a result of gastrectomy that hampers food intake, patients with gastric cancer experience loss of appetite and are predominantly at risk for body weight loss (BWL) leading to malnutrition. BWL may be caused by multiple factors. It was reported that BWL after other cancer surgery had a negative impact on the prognosis of patients (11). However, only a few studies have reported on BWL in patients with gastric cancer who underwent curative surgery (12, 13), and clinicopathological factors related to BWL and the prognostic effect of BWL in the patients have not yet been fully elucidated. To clarity this issue, as well as whether BWL should be proactively treated after surgery, this study was performed retrospectively from patients' hospital records.
Patients and Methods
Study population and treatments. A total of 102 consecutive patients with pathological stage (pStage) II or III who underwent curative gastrectomy for gastric cancer at the Division of Digestive Surgery, Department of Surgery, Kyoto Prefectural University of Medicine, Japan, between January 2005 and December 2011, and who did not experience a recurrence for the first 6 months were retrospectively analysed from their hospital records. We excluded patients in pStage I because they rarely have recurrences, and patients in pStage IV because evaluation of the clinical and prognostic effects of postoperative BWL might be difficult after non-curative resection.
Almost all patients without gastric stenosis took food until dinner the day before surgery. Curative gastrectomy with lymphadenectomy was performed based on the Japanese gastric cancer treatment guidelines (14, 15). Surgical procedures comprised distal gastrectomy, total gastrectomy, and proximal gastrectomy according to the preoperative stage and tumour location. Prophylactically, one or two drainage tubes for subsequent leakage or pancreatic fistulae were placed for radical gastrectomy. The nasogastric tube was removed on postoperative day 1. Oral intake of rice gruel was started on postoperative day 3 or 4. At the time of discharge, almost all patients were taking regular food. Postoperative complications were assessed by Clavien-Dindo classification (16, 17).
Resected specimens were examined and evaluated by pathologists based on the 14th Japanese Classification of Gastric Carcinoma (18) and the 7th Tumour-Node-Metastasis classification (19). All patients were examined in the outpatient clinic, where abdominal ultrasound, computed tomography, and measurement of levels of carcinoembryonic antigen and carbohydrate antigen 19-9 were performed every 3 to 6 months after surgery.
Assessment of postoperative BWL. For all patients examined, body weight, height, body mass index (BMI), and BWL before and 6 months after surgery were analysed. Postoperative BWL was calculated by the following formula: BWL (%)=(preoperative BMI–postoperative BMI at 6 months)/preoperative BMI×100. The body weight of obese patients was reduced more substantially than was that of patients with normal BMI after gastrectomy (data not shown) because the total food intake of obese patients became similar to that of non-obese patients after surgery. In other words, the preoperative food intake of obese patients was relatively higher than that of non-obese patients. Therefore, to examine the clinical effect of critical BWL in patients with a preoperative BMI of 25 kg/m2 or higher, we modified the formula as follows: BWL (%)=(25–postoperative BMI at 6 months)/preoperative BMI×100.
Statistical analysis. The Mann–Whitney U-test was used for unpaired continuous variables of postoperative BWL. Receiver-operating characteristic (ROC) curves and the area under the ROC curve (AUC) were used to assess the feasibility of using postoperative BWL at 6 months after gastrectomy as a predictive marker for subsequent recurrence. The Youden index was used to determine the cut-off value for postoperative BWL (20). The chi-square test or Fisher's exact probability test was used for categorical variables.
For the analysis of overall (OS) and disease-free (DFS) survival rates, survival curves were estimated using the Kaplan–Meier method, and statistical differences were examined using the log-rank test. Univariate and multivariate survival analyses were performed using the likelihood ratio test of the stratified Cox proportional hazards model. A value of p<0.05 was considered statistically significant. All statistical analyses were performed using JMP 10.0 software program (SAS Institute, Cary, NC, USA).
Results
Patients' characteristics. Surgical procedures comprised of distal gastrectomy in 55 patients, total gastrectomy in 44, and proximal gastrectomy in three. D2 or D2+ lymphadenectomy was performed in 74 patients and D0-1+ lymphadenectomy was performed in 28 patients, based on the Japanese guidelines (15). Disease in eight patients was staged as pT1, in 17 as pT2, in 35 as pT3, and in 42 as pT4. Of 102 patients, disease in 49 was categorized as pStage II, whereas in 53 patients it was categorized as pStage III. Eight patients had postoperative complication assessed by the Clavien-Dindo classification as grade II or higher, and 94 patients had no complications or Clavien-Dindo grade I complication. There were 25 patients whose preoperative BMI was 25 kg/m2 or higher, and their BWLRs were calculated by the modified formula.
Association between body weight loss (BWL) and clinicopathological variables.
A total of 71 patients (70%) received adjuvant chemotherapy, but 31 patients (30%) did not. The duration of adjuvant chemotherapy was as follows: 0-3 months for four patients, 3-6 months for eight and 6 months or more for 59 patients. As adjuvant chemotherapy, the following drugs were administered: S-1 alone to 55 patients, tegafur-uracil to three patients, tegafur-uracil after S-1 to two patients, 5-fluorouracil after S-1 to one patient, 5-fluorouracil and cisplatin to one patient, S-1 and paclitaxel to two patients, or S-1 and cisplatin to seven patients. None of the patients received adjuvant radiotherapy or chemoradiotherapy. The mean follow-up period was 44.4 months.
Curve for tumor recurrence and body weight loss at 6 months after surgery for patients with stage II and III gastric cancer. The area under the curve was 0.634, with a sensitivity of 59.0%, a specificity of 69.8%, and an accuracy of 65.7%.
Clinicopathological characteristics and postoperative BWL after curative gastrectomy in patients with pStage II or III gastric cancer. We analyzed the clinicopathological characteristics of patients and compared their postoperative BWL (Table I). Patients with tumours in the upper stomach (p=0.014), total gastrectomy (p<0.001), or subsequent recurrences (p=0.010) had a significantly higher BWL, whereas the staging factors, such as depth of invasion and lymph node metastasis, were not associated with BWLR. A higher BWL also tended to be observed after D2 or D2+ lymphadenectomy (p=0.073) and be related to the duration of adjuvant chemotherapy (3-6 months, p=0.090; 6 months or more, p=0.081) and severe complications after surgery (p=0.062).
Cut-off value of BWL associated with subsequent recurrence. To find a cut-off point of BWL that could predict a subsequent recurrence in patients, we used the AUC with the Youden index (Figure 1). The value of the AUC was 0.634 and the optimal cut-off point of BWL was observed at 12% with a sensitivity of 59.0%, a specificity of 69.8%, and an accuracy of 65.7%.
Comparisons of Kaplan–Meier curves of five-year overall survival (OS) (a) and disease-free survival (DFS) (b) between patients based on body weight loss (BWL) during the first 6 months after surgery. The log-rank test was used for statistical analysis. All 102 patients analysed in this study were divided into two groups by BWL: BWL≥12% (n=42) and BWL<12% (n=60).
Comparison of clinicopathological factors between patients with and without 12% BWL. We next compared the clinicopathological features between patients with a BWL of 12% or more and those without (Table II). Long duration of adjuvant chemotherapy and subsequent recurrence were significantly associated with BWL of 12% or higher (p=0.001 and p=0.004, respectively).
Survival outcomes. The OS and DFS curves are represented in Figure 2. There was no difference in OS between patients with a postoperative BWL of 12% or higher and those with a BWL of less than 12% (Figure 2a), whereas those with a postoperative BWL of 12% or higher had a significantly poor DFS (p=0.009). The 5-year DFS rates of patients with a BWL of 12% or higher and less than 12% was 44% and 67%, respectively (Figure 2b).
Comparison of DFS in patients with gastric cancer after curative gastrectomy. Univariate and multivariate analysis using the Cox proportional hazard model demonstrated that a BWL of 12% or higher was an independent prognostic factor of poor DFS (hazard ratio=2.69 (95% confidence interval=1.42-5.21; p=0.002) (Table III).
Discussion
The aim of the study was to assess the prognostic effect of BWL after gastrectomy in patients with gastric cancer. Recent studies in some malignant diseases identified that BWL after esophagectomy (11), after allogenic hematopoietic stem cell transplantation (21) or during radiotherapy (22) were reported to have a negative impact on prognosis. In gastric cancer, however, there exist only a few reports on poor outcomes associated with BWL. Yu et al. reported that BWL between 6 and 12 months after surgery was an independent poor prognostic factor (12). In addition, Lee et al. revealed that a BWL of more than 30% at 6 months after surgery was significantly associated with an early recurrence (13). In this study, we recorded a potential risk of subsequent recurrences associated with BWL even in the first 6 months after surgery. Moreover, we verified that the BWL cut-off value of 12%, which is lower than that in a previous report (13), was crucial for predicting patients with subsequent recurrence. Our results imply that nutritional insufficiency at an early stage after surgery and higher BWL could affect recurrence in patients after curative surgery for gastric cancer. We, therefore, should reconsider how patients with BWL should be treated after gastrectomy.
The immune system is highly dependent upon the adequate availability of nutrients, such as amino acids (23), specific vitamins, minerals and trace elements (24, 25). BWL may be closely linked to the deficiency of these nutrients. In this study, we hypothesized that poor outcomes after curative gastrectomy might be due to an impaired immune response as a consequence of BWL in patients with gastric cancer. To verify this hypothesis, we investigated which clinicopathological factors are related to BWL, and whether BWL is indeed associated with poor outcomes in patients. Consequently, we clearly demonstrated that BWL during the first 6 months after surgery is significantly associated with clinical factors, such as tumour in the upper stomach, total gastrectomy, and subsequent recurrence. In addition, we suggest that a postoperative BWL of 12% or higher is an independent risk factor for subsequent recurrence in patients with stage II or III gastric cancer.
Risk factors associated with recurrence in 102 patients after curative gastrectomy.
Patients with gastric cancer are predominantly at-risk for malnutrition due to loss of appetite, which is a sequelae of gastrectomy that hampers food intake. Malnutrition is a sub-acute or chronic state in which a combination of varying degrees of undernutrition and inflammation leads to a change in body composition and diminished function (26). As shown in a previous study related to BWL (27), we found that patients with tumours in the upper stomach or after total gastrectomy had a significantly higher BWL as a result of functional impairment after gastrectomy, regardless of the staging factors, such as the depth of invasion and lymph node metastasis. In addition, as an inflammatory effect leading to malnutrition, D2 or D2+ lymphadenectomy and severe complications after surgery tended to lead to a higher BWL. Therefore, these risk factors could be important in predicting severe postoperative BWL and making a decision about meticulous nutritional support.
Univariate and multivariate cox proportional hazards analysis of clinicopathological factors associated with disease-free survival (DFS).
How BWL affects tumour recurrence remains unclear. BWL in patients with cancer is not caused only by diminished nutrient intake and increased nutrient loss but also by tumour-induced dysregulation of metabolism (28, 29). Recently, a high BWL after gastrectomy was reported to be an important factor preventing patients from continuing with S-1 adjuvant chemotherapy (30). Therefore, our first hypothesis was that this poor compliance might be a risk factor for subsequent recurrence. However, in the present study, we did not find such a correlation. On the contrary, long duration of adjuvant chemotherapy was an independent risk factor for patients with BWL of 12% or higher. A similar result was also reported by Yamaoka et al. who found that adjuvant chemotherapy after total gastrectomy was a risk factor for skeletal muscle loss (31). The correlation of BWL and skeletal muscle loss is still unclear, as is whether perioperative nutritional intervention with sufficient caloric intake can prevent subsequent recurrence after curative gastrectomy for patients with gastric cancer. To elucidate these issues, further studies are warranted.
Our study has limitations because it is retrospective in its design and based on a single Institution. Therefore, a prospective study with a large number of patients is needed to validate issues associated with BWL. In conclusion, our study clearly demonstrates that a high BWL after curative gastrectomy contributes to subsequent recurrence in patients with gastric cancer.
Footnotes
Conflicts of Interest
The Authors declare no conflicts of interest.
- Received October 22, 2015.
- Revision received December 12, 2015.
- Accepted December 18, 2015.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
