Abstract
Background/Aim: Radical resection after preoperative adjuvant chemotherapy (NAC) is a standard treatment for patients with locally advanced esophageal squamous cell carcinoma (LAESCC), but its outcome remains unsatisfactory. In order to develop a personalized treatment program for LAES, we herein compared the survival prediction utility of five pre-NAC nutritional, inflammatory, and immune indexes in patients with LAESCC. Patients and Methods: We evaluated the survival of 203 patients with LAESCC who underwent radical resection after NAC from January 2011 to September 2019 for the following representative pre-NAC nutritional, inflammatory, and immune indices: modified Glasgow Prognostic Score, Prognostic Nutritional Index, C-reactive protein/albumin ratio, serum neutrophil/lymphocyte ratio, and Geriatric Nutrition Risk Index (GNRI) were evaluated for their impact on survival. Results: Of the five indices, GNRI was the best predictor of survival as determined by the area under the curve (p<0.05). When patients were divided into three groups according to the nutritional risk assessment of Bouillanne et al. using the pre-NAC GNRI, the 5-year overall survival (OS) and recurrence-free survival (RFS) were significantly stratified (p<0.001). On multivariate analysis, the GNRI independently identified a poor OS group [group 1: hazard ratio (HR)=2.598, p=0.002; group 2: HR=6.257, p<0.001] and a high recurrence risk group (group 1: HR=1.967, p=0.016; group 2: HR=4.467, p<0.001). Conclusion: In patients with LAESCC, GNRI may be the most accurate, reliable, and useful prognostic factor among the five major systemic inflammatory and nutritional indices.
Esophageal cancer (EC) is the seventh most common cancer worldwide and the sixth leading cause of cancer-related death, with more than 600,000 new diagnoses and 400,000 EC-related deaths reported annually (1). The standard treatment for patients with locally advanced esophageal squamous cell carcinoma (LAESCC) is multimodal, consisting of neoadjuvant chemotherapy (NAC) or neoadjuvant chemoradiotherapy (NACRT) plus curative resection±postoperative adjuvant chemotherapy (2, 3). However, the treatment results remain unsatisfactory. Therefore, personalized therapy based on predictors of preoperative therapy response, relapse, and survival could potentially improve outcomes for patients with LAESCC. Several potential markers have also been reported to date (4-16).
Patients with LAESCC often have poor performance status and quality of life due to malnutrition. Malnutrition is a risk factor for postoperative complications, as well as relapse and poor survival in these patients (17, 18). Pre-treatment malnutrition has also been associated with NAC and NARTC effects, and pre-treatment nutritional interventions have been reported to produce beneficial patient outcomes by improving perioperative treatment tolerability (19, 20). The European Society for Clinical Nutrition and Metabolism recommends that all cancer patients undergo nutritional risk screening early during care (21).
Herein, we calculated five representative nutritional, inflammatory, and immune indices: the modified Glasgow Prognostic Score (mGPS) (22), the Prognostic Nutritional Index (PNI) (23), the C-reactive protein/albumin ratio (CAR) (24), the serum neutrophil/lymphocyte ratio (NLR) (25), and the Geriatric Nutritional Risk Index (GNRI) (26). The pre-NAC values were compared to determine the usefulness of each index for predicting survival of patients with LAESCC.
Patients and Methods
Eligible patients. Consecutive patients who underwent esophagectomy for EC at Kanagawa Cancer Center from January 2011 to March 2020 were retrospectively reviewed. We included patients with histologically proven primary esophageal squamous cell carcinoma (ESCC) located in the thoracic esophagus, with clinical stage II or Stage III EC, and those who underwent complete resection (R0) of EC with radical lymph node dissection post-NAC. We excluded patients who received preoperative radiation therapy and chemotherapy other than cisplatin plus 5-fluorouracil, who underwent R2 or R1 resection, who underwent esophagectomy as salvage therapy, and patients who received postoperative chemotherapy.
Treatment procedure. Preoperative NAC was administered to patients diagnosed with EC clinical stage II and III, as assessed using the Union for International Cancer Control (UICC)-TNM staging system, 8th edition (27), based on the Japanese Clinical Oncology Research Group 9907 study (2). Patients received two courses of cisplatin and 5-fluorouracil. Cisplatin was administered at 80 mg/m2 intravenously (IV) on day 1, and 5-fluorouracil was administered at 800 mg/m2 as a continuous IV infusion on days 1-5. Cisplatin and 5-fluorouracil were administered every 3 or 4 weeks.
Surgical resection was performed 4-6 weeks post-NAC. Surgery consisted of open subtotal esophagectomy through a right thoracotomy, reconstruction with a gastric tube through the posterior mediastinal or retrosternal route, and anastomosis through a cervical incision. Lymph node dissection was indicated with two- or three-field lymph node dissection if the tumor was in the middle to lower thoracic esophagus, and with three-field dissection if the tumor was in the upper thoracic esophagus. An enteral feeding tube was placed during surgery, and enteral nutrition was combined with intravenous fluids and food post-surgery.
Postoperative follow-up included physical examination, blood chemistry (every 3 months for the first year and every 6 months thereafter), and computed tomography of the neck, chest, and abdomen every 6 months. Disease recurrence was diagnosed based on radiographic evidence. Unclear masses were examined using positron-emission tomography.
Calculation of pre-NAC nutritional, inflammatory, and immune indices. For each patient, pre-NAC data were used to calculate mGPS (cut-off values for CRP and Alb were 1.0 mg/dl and 3.5 g/dl, respectively) (22), PNI (10×Alb value+0.005×total lymphocyte count) (23), CAR (24), NLR (25), and GNRI (1.489×Alb, g/l)+(41.7×current/ideal weight) (26). Ideal weight, kg was calculated from the Lorenz equation: for men, (height, cm − 100) − [(height, cm − 150)/4]; for women, (height, cm − 100) − [(height, cm − 150)/2.5] (26).
Postoperative infectious complications. Complications were defined as grade II or higher in the Clavien-Dindo classification (28). Postoperative infectious complications included anastomotic leak, pneumonia, abdominal abscess, surgical site infection, and/or pyothorax occurring during hospitalization within 30 days post-surgery.
Pathological response to preoperative chemotherapy. Based on the histopathological criteria of the Japanese Esophageal Association (29), tumor regression was classified into the following five categories according to the percentage of degeneration or necrosis in the primary tumor. 0 degree: No therapeutic effect on cancerous tissue or cells. Grade Ia: Residual tumor is at least two-thirds of the size of the original tumor. Grade Ib: Residual tumor is greater than one-third but less than two-thirds of the size of the original tumor. Grade II: Residual tumor is less than one-third of the size of the original tumor. Grade III: No viable cancer cells are observed.
Statistical analysis. Survival data were obtained from hospital records. Overall survival (OS) was defined as the interval from the start of treatment to death. Recurrence-free survival (RFS) was defined as the interval from the start of treatment to the discovery of recurrence or death, whichever occurred first. Data for patients who did not experience an event were censored at the date of last follow-up. OS and RFS curves were calculated using the Kaplan-Meier method and were compared using the log-rank test. Univariate and multivariate analyses using the Cox proportional hazards model were used to identify survival-related prognostic factors. Receiver operating characteristic curves were generated and differences in the areas under the curves (AUCs) were compared to assess the discriminatory power of the prognostic scoring system. Fisher’s exact test and the Kruskal-Wallis test were used for group comparisons. A p-value <0.05 was considered statistically significant. All statistical analyses were performed using EZR version 1.61 (Saitama Medical Center, Jichi Medical University, Saitama, Japan) (30) and R version 3.6.2 (The R Foundation for Statistical Computing, Vienna, Austria).
Ethical approval. This study was conducted in compliance with the ethical guidelines for clinical research and the 1975 Declaration of Helsinki, as revised in 2013, and was approved by the Institutional Review Board of Kanagawa Cancer Center (Approval No. 2023 Epidemiological Study-129). Prior to surgery, written informed consent was obtained from each patient for the use of anonymized clinical data.
Results
Patients. Between January 2011 and March 2020, 239 patients were diagnosed with clinical stage II or III EC and received preoperative treatment. Of these, nine patients who received NACRT, 11 patients who received NAC in triplet regimens, eight patients for whom sufficient data were not available, and eight patients who received non-therapeutic resection were excluded. The final analysis included 203 patients.
Comparison of AUCs for each inflammatory marker. AUCs generated for 5-year postoperative survival status are given in Table I. The AUC value for GNRI was significantly greater than those for other inflammatory indices (Table I).
Area under the curve of the inflammatory indicators generated for survival status at 5 years after surgery.
Clinicopathological characteristics based on three GNRI groups divided according to nutritional risk. The GNRI pre-NAC was used to define three groups: high-risk (GNRI: <92, group 2), low-risk (GNRI: 92-98, group 1), and no-risk (GNRI: >98, group 0). Body mass index (BMI; p<0.001), clinical T-factor (p<0.001), clinical N-factor (p=0.029), clinical stage (p<0.001), pathological T-factor (p=0.001), and histopathologic response (p=0.03) differed significantly among the three groups (Table II). Pathological N-factors and pathological stage showed a trend for differences among the three groups.
Association between the Geriatric Nutritional Risk Index and clinicopathological factors.
OS and RFS differences according to three GNRI groups. The median follow-up was 34.13 [interquartile range (IQR)=17.22-64.49] months. OS and PFS were significantly different among the three groups (both p<0.001; Figure 1 and Figure 2).
Kaplan-Meier OS curves according to the Geriatric Nutritional Risk Index. The median follow-up period was 34.13 (interquartile range=17.22-64.49) months. The 5-year OS rates were 23.3%, 40.2%, and 77.7% in Groups 2, 1, and 0, respectively, which were significantly different among the three groups (p<0.001). OS: Overall survival.
Kaplan-Meier RFS) curves according to the Geriatric Nutritional Risk Index. The 5-year RFS rates were 14.0%, 34.5%, and 68.4% in Groups 2, 1, and 0, respectively, which were significantly different among the three groups (p<0.001). RFS: Relapse-free survival.
Univariate and multivariate analysis of factors related to OS and RFS in the three GNRI groups. Univariate analysis of OS showed that sex, clinical T factor, clinical N factor, and GNRI score were significant predictors of OS (Table III). Multivariate analysis further showed that male sex and GNRI group 1/2 were significant independent predictors of poor OS (Table III).
Univariate and multivariate Cox proportional hazard analyses of the clinicopathological factors associated with overall survival.
On univariate analysis of RFS, sex, clinical T-factor, clinical N-factor, and GNRI score were significantly associated with RFS (Table IV). On multivariate analysis, male sex, clinical T3-4, clinical N1-3 stage, and GNRI group 1/2 were significant independent predictors of poor RFS in our cohort (Table IV).
Univariate and multivariate Cox proportional hazards analyses of the clinicopathological factors associated with recurrence-free survival.
Discussion
In this study, we evaluated five representative reported inflammatory and nutritional indices: mGPS (22), PNI (23), CAR (24), NLR (25), and GNRI (26) in LAESCC patients and we compared the usefulness of the pre-NAC data of each index for predicting survival. The GNRI performed best among these indices. Therefore, we examined the relationship between the GNRI and clinicopathologic factors, RFS, and OS by dividing pre-NAC GNRI into three groups according to the nutritional risk defined by Bouillanne et al. (26). The results showed that BMI, T factor, N factor, and pathological response (p=0.03) were associated with GNRI. Moreover, significant differences were observed among the three GNRI groups in terms of both RFS and OS. Furthermore, GNRI was an independent adverse factor for RFS and OS in multivariate analysis.
The GNRI was published by Bouillanne et al. in 2005 as a modified version of the NRI, which is used to estimate the risk of postoperative complications related to nutritional disorders for older patients, by adding serum albumin and ideal body weight according to sex. It is expected to become an increasingly important indicator as cancer patients age.
Previous reports have suggested that malnutrition, systemic inflammation, and the immune system in the tumor microenvironment induce tumor cell growth, metastasis, angiogenesis, resistance to anticancer therapy, and disruption of antitumor immunity (31, 32). The serum albumin level is one of the most important blood count parameters for predicting perioperative risk and oncological outcome in ESCC patients (33, 34). However, cutoff values are arbitrarily set, and these levels may thus not adequately reflect the patient’s general condition, including factors such as fluid levels.
Higher BMI correlates with improved survival outcomes in patients with lung cancer (35), bladder cancer (36), and ESCC (37). Thus, a multidimensional prognostic evaluation system incorporating multiple factors may be more accurate than prediction based on a single prognostic factor (38). Nevertheless, few studies have evaluated combinations of these markers, particularly in ESCC patients.
After Bouillanne et al. first used the GNRI to assess nutritional outcomes in older patients (26), a small number of reports on the utility of GNRI as a prognostic factor in ESCC have been published. In a study of ESCC patients limited to Stage I/II, without NAC, GNRI independently predicted poor cancer-specific survival (HR=1.94, 95%CI=1.12-3.37, p=0.02) (39). In a retrospective analysis limited to stage III ESCC patients, 5-year OS and cancer-specific survival were significantly worse in the low-GNRI group than in the high-GNRI group (34.4% vs. 52.1%, p=0.049 and 36.1% vs. 57.2%, p=0.041, respectively) (40). However, in a study of patients with LAESCC who had undergone NAC or NACRT, GNRI was not an independent predictor of poor RFS or OS in multivariate analysis (40). Therefore, no previous studies have examined the relationship of OS with pre-NAC GNRI in LAESCC patients.
GNRI, consisting of blood albumin level and BMI, may be of value as a prognostic indicator for patients with stage II/III ESCC because blood albumin levels may better reflect the nutritional and inflammatory status of the host (41, 42). Cancer-related hypoalbuminemia enhances malnutrition associated with cancer cachexia, a multifactorial syndrome characterized by progressive skeletal muscle wasting (43). In addition, high albumin levels are associated with reduced mortality rates in patients with cachexia (44). BMI, which is calculated from body weight and height, can be used to assess body shape. Nutritional assessment using BMI can provide some insight into the balance between muscle mass and fat mass, allowing nutritional status to be assessed based on body shape as well as weight changes (45-47). Thus, including BMI in GNRI calculations may allow a more comprehensive assessment of nutritional risk. Furthermore, the ability to perform a detailed nutritional assessment facilitates planning of appropriate nutritional support and interventions (31).
Study limitations. First, the study was retrospective, with a relatively small sample size from a single center and a short follow-up period. Therefore, this finding should be validated in a multicenter study with a larger population. Second, although docetaxel, cisplatin, and fluorouracil as NAC are currently the standard of care for stage II/III ESCC patients in Japan (48), this study evaluated the efficacy of GNRI in patients receiving 5-fluorouracil and cisplatin chemotherapy. Thus, validation in patients with stage II/III ESCC receiving docetaxel, cisplatin, and fluorouracil therapy is warranted.
Conclusion
In conclusion, in LAESCC patients, GNRI may be the most accurate, reliable, and useful prognostic factor among five representative systemic inflammatory and nutritional indices. Early detection of malnutrition using the GNRI and providing nutritional supportive care before NAC may improve OS and RFS in LAESCC patients.
Acknowledgements
The Authors thank the patients, their families, and the staff at Kanagawa Cancer Center for their participation in this study.
Footnotes
Authors’ Contributions
IH, KK, TO, and TO had full access to all data in the study and take responsibility for the data’s integrity and the accuracy of data analysis. Concept and design: IH, KK, HS, and TO. Acquisition, analysis, or interpretation of data: IH, KK, and TO. Drafting of the manuscript: IH, KK, and TO. Critical revision of the manuscript for important intellectual content: IH, KK, HS, TY, HW, KK, SN, TA, TO, YR, AS, and TO. All the Authors actively participated in this study and read and approved the final version of the paper for publication.
Conflicts of Interest
The Authors declare that there are no conflicts of interest regarding this study.
- Received October 14, 2023.
- Revision received November 21, 2023.
- Accepted December 6, 2023.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
