Abstract
Aim: The aim of this study was to investigate the correlations between clinicopathological findings, laboratory data and survival outcome in patients undergoing curative laparoscopic surgery for colorectal cancer (CRC). Patients and Methods: Clinicopathological findings and laboratory data, including tumor markers carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) and systemic inflammatory response indicators, neutrophil-to-lymphocyte ratio (NLR) and modified Glasgow prognosis score (mGPS), for 204 patients (tumor stage I-III) undergoing laparoscopic curative surgery for CRC were collected. Results: Elevated CA19-9 and mGPS, and body mass index (BMI) <20 kg/m2 were significant indicators of poorer overall survival, while CA19-9 and BMI were validated as independent predictors of overall survival. In addition, BMI <20 kg/m2 was a significant independent factor predictive of poorer disease-free survival. BMI significantly negatively correlated with NLR, which reflects the patients' immune response. Conclusion: Lower BMI is a promising predictor of recurrence and poor prognosis in patients treated by laparoscopic surgery for CRC with curative intent.
Colorectal cancer (CRC) is one of the most common types of cancer worldwide (1). Surgical resection is the mainstay of CRC treatment, and laparoscopic surgery for CRC is an acceptable and safe alternative to open surgery, as it has clinical benefit – patients recover earlier from limited invasiveness, fewer postoperative analgesics are required, and the hospital stay is shorter (2, 3). In addition, because the long-term oncological outcome of laparoscopic surgery for CRC is reportedly comparable to that of open surgery (4, 5), laparoscopic surgery is slated to become a more common treatment for patients with CRC.
Routine laboratory data prior to surgery have been demonstrated as predictors of postoperative recurrence and prognosis in patients with CRC. Previous investigators have reported the importance of tumor markers carcinoembryonic antigen (CEA) (6), carbohydrate antigen 19-9 (CA19-9) (7, 8), and systemic inflammatory response (SIR) indicators [neutrophil/lymphocyte ratio (NLR) and modified Glasgow prognosis score (mGPS)] (9-14) as potential indicators for predicting oncological outcomes in CRC. However, the association between these indicators, clinicopathological findings, and oncological outcomes has not been fully evaluated in patients with CRC treated by laparoscopic surgery.
The aim of this study was to investigate the correlations between clinicopathological findings, laboratory data (including tumor markers and SIR indicators), and survival outcome in patients undergoing laparoscopic surgery for CRC with curative intent.
Patients and Methods
Overall, 204 patients who underwent potentially curative laparoscopic surgery for CRC at our institution between January 2004 and December 2013 were enrolled in this retrospective study. Curative resection was defined as the absence of any gross residual tumor in the surgical bed and a surgical resection margin that was pathologically negative for tumor invasion. No patient received chemotherapy or radiotherapy before surgery and no perioperative mortality was observed.
Patients were followed-up, using our standard protocol, every 12-16 weeks for at least 1 year. This protocol included tumor-marker studies, computed tomography, colorectal fiber examinations, ultrasonography, and chest radiography. Bone scans were performed when bone metastasis was indicated. The median follow-up time was 30.9 months [95% confidence interval (CI) for the mean was 33.6-41.1 months]. Data collected from inpatient and outpatient records included demographic data [age, sex, height, weight, and body mass index (BMI)], tumor-specific data [tumor location, size, pathologic data including T-classification, lymph node metastasis, differentiation, lymphatic and venous duct invasion, and levels of tumor markers CEA and CA19-9 at diagnosis], operative data (blood loss, operative time, anastomotic leakage, and postoperative complication), and survival data [disease-free survival (DFS) and overall survival (OS)]. Peripheral blood samples were collected from patients prior to surgery, and included the neutrophil and lymphocyte counts for calculating the neutrophil-to-lymphocyte ratio (NLR), albumin level, and C-reactive protein (CRP) level to define the mGPS (13). The mGPS was determined as below. Briefly, patients with elevated CRP levels (>0.5 mg/dl) plus hypoalbuminemia (<3.5 g/dl) were allocated a score of 2 (positive). Patients with only one of these factors were allocated a score of 1 (positive). Patients with neither of these factors were allocated a score of 0 (negative).
The cutoff values for CEA and CA19-9 were 5 ng/ml and 37 U/ml, respectively, according to the normal ranges used at our hospital. The cutoff values for NLR and mGPS were defined as 2.5 and 0. In addition, the cutoff value for BMI was defined as less than the 25th percentile (20 kg/m2; Figure 1A). Blood collection and subsequent analyses were approved by the Institutional Review Boards of Mie University Hospital in Japan (protocol number: 2216). All participants provided written informed consent and indicated their willingness to donate blood for research.
Statistical analysis. The data are presented as means±standard deviation (SD). Comparisons were made using the Mann–Whitney or Kruskal–Wallis tests, as appropriate. Correlations were analyzed by Spearman's coefficient analysis. Receiver-operating characteristic (ROC) curves were established for determining cutoff values for analyzing prediction of lymph node metastasis and prognosis by Youden's index. Survival curves were obtained using the Kaplan–Meier product limit method, and comparisons were made using the log-rank test. Prognostic factors were examined by univariate and multivariate analysis (Cox proportional hazards regression model). All p-values were two-sided and p<0.05 was considered statistically significant. All statistical analyses were carried out using Medcalc 7.2 for Windows (Mariakerke, Belgium).
Results
Patients' characteristics. The study group comprised of 101 males and 103 females aged 37-89 years (median=69 years; interquartile range=62-77 years). Staging was principally based on the Union for International Cancer Control /TNM classification of colorectal cancer (15). Overall, 78 patients (39%) had stage I disease, 70 (34%) had stage II, and 56 (27%) had stage III. Starting 4 weeks after curative surgery, pyrimidine fluoride-based regimens were used for 6 months to 1 year in patients classified with stage III disease. Out of the 204 registered patients, 159 (78%) had tumors located in the colon and 45 (22%) had tumors in the rectum. Pathological tumor diameter was the maximum microscopic length of the tumor, irrespective of depth. Histologically-differentiated tumors were noted in 196 patients (96%) and undifferentiated tumors in eight patients (4%). Lymph node metastasis was identified in 56 (27%) patients. In addition, lymphatic invasion was observed in 109 patients (53%) and vascular invasion in 64 patients (31%).
Associations between preoperative tumor markers, SIR markers, BMI, and clinicopathological features. Firstly, we examined the correlation between preoperative tumor markers CEA and CA19-9, SIR markers NLR and mGPS, BMI before surgery, and clinicopathological characteristics. There were significant positive correlations between CEA and CA19-9 (rho=0.301, p<0.0001; Figure 1B), CEA and NLR (rho=0.156, p=0.03; Figure 1C), and CEA and mGPS (rho=0.197, p=0.0058; Figure 1D). In contrast, there were significant negative correlations between BMI and CEA (rho=−0.141, p=0.0498; Figure 1E) and BMI and NLR (rho=−0.213, p=0.0027; Figure 1F). Table I shows that the preoperative serum CEA level was significantly associated with tumor progression, including high pathological T-stage (p=0.0002), lymphatic (p=0.008) and vascular (p=0.0013) duct invasion, lymph node metastasis (p=0.0005), and presence of recurrence (p=0.017). In addition, a high mGPS score was significantly associated with high pathological T-stage (p=0.006). Furthermore, a low BMI was significantly associated with postoperative recurrence (p=0.03).
Identifying independent predictors of poor prognosis in patients undergoing laparoscopic surgery for CRC. OS curves for patients with CRC classified on the basis of CEA, CA19-9, mGPS, NLR, and BMI values prior to laparoscopic surgery are shown in Figure 2. Patients with elevated CEA, CA19-9, and mGPS had significantly poorer OS than patients with levels below the cutoff values (log-rank test: CEA, p=0.05; CA19-9, p=0.0024; mGPS, p=0.0166; Figure 2A-C). In addition, Figure 2E reveals the interesting finding that patients with a BMI <20 kg/m2 had a significantly inferior prognosis compared to those with BMI >20 kg/m2. Univariate analysis identified pathological stage T3-4 (p=0.00049), poorly-differentiated adenocarcinoma (p<0.0001), venous invasion (p=0.0157), elevated CA19-9 (p=0.007), elevated mGPS (p=0.0267), and BMI <20 kg/m2 as significant prognostic factors for poor OS (Table II). Multivariate analysis using a Cox proportional hazards model showed that pathological stage T3-4 [hazard ratio (HR)=29.8598, 95% confidence interval (CI)=2.0999-424.5951; p=0.0126], poorly differentiated adenocarcinoma (HR=24.4823, 95% CI=2.9137-205.7101; p=0.0034), elevated CA19-9 (HR=15.5553, 95% CI=2.288-105.7524; p=0.0052), and BMI less than 20 (HR=0.0642, 95% CI=0.0126-0.3269, p=0.001) were independent predictors of poor prognosis in patients treated with laparoscopic surgery for CRC (Table II).
Distribution of body mass index (BMI) in our cohort and correlation between preoperative tumor markers, systemic inflammatory response (SIR) status, and BMI before laparoscopic surgery for colorectal cancer. A: Distribution of BMI values. The cut-off value for BMI was at the 25th percentile (20 kg/m2). B: Positive correlation between carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9). C: Positive correlation between CEA and neutrophil-to-lymphocyte ratio (NLR). D: Positive correlation between CEA and modified Glasgow prognosis score (mGPS). E: Negative correlation between BMI and CEA. F: Negative correlation between BMI and NLR.
Overall survival (OS) curves for patients with colorectal cancer classified on the basis of several indicators prior to laparoscopic surgery. OS curves for patients when partitioned by carcinoembryonic antigen (CEA) (cut-off=5 ng/ml) (A), carbohydrate antigen 19-9 (CA19-9) (cut-off=37 U/ml) (B), modified Glasgow prognosis score (mGPS) (cut-off=0) (C), neutrophil-to-lymphocyte ratio (NLR) (cut-off=2.5) (D), and body mass index (BMI) (cut-off=20 kg/m2) (E).
Disease-free survival (DFS) curves for colorectal cancer patients classified on the basis of several indicators prior to laparoscopic surgery. DFS curve when partitioned by carcinoembryonic antigen (CEA) (cut-off=5 ng/ml) (A), carbohydrate antigen 19-9 (CA19-9) (cut-off=37 U/ml) (B), modified Glasgow prognosis score (mGPS) (cut-off=0) (C), neutrophil-to-lymphocyte ratio (NLR) (cut-off=2.5) (D), and body mass index (BMI) (cut-off=20 kg/m2) (E).
Association between tumor markers (carcinoembryonic antigen [CEA] and carbohydrate antigen 19-9 [CA19-9]), systemic inflammatory response status (neutrophil-to-lymphocyte ratio [NLR] and modified Glasgow prognosis score [mGPS], and body mass index [BMI] and clinicopathological findings.
Identifying independent predictors of recurrence in CRC patients undergoing laparoscopic surgery. We evaluated which of the factors available in a clinical setting could predict recurrence in patients undergoing laparoscopic surgery for CRC. DFS curves for patients classified according to CEA, CA19-9, mGPS, NLR, and BMI values were generated (Figure 3). Among these factors, BMI <20 kg/m2 was the only factor associated with significantly poorer DFS (log-rank test; p=0.0264; Figure 3E). Multivariate analysis using a Cox proportional hazards model showed that both poorly-differentiated adenocarcinoma and a BMI <20 kg/m2 were independent predictive markers for early recurrence in patients treated by laparoscopic surgery for CRC (p=0.0448 and p=0.0254, respectively; Table III).
A BMI of less than 20 kg/m2 is an indicator of poor prognosis in patients without lymph node metastases from CRC. We also analyzed DFS and OS of patients without lymph node metastases of CRC (stage I-II), who generally do not receive adjuvant chemotherapy after surgery. Our cohort revealed that DFS (log-rank test; p=0.0002; Figure 4A) and OS (log-rank test; p<0.0001; Figure 4B) of patients with BMI <20 kg/m2 was significantly poorer than patients with BMI ≥20 kg/m2.
Discussion
In the present study, we demonstrated for the first time that patients with postoperative recurrence had a significantly lower preoperative BMI. In addition, Kaplan–Meier curves for OS and DFS showed that patients with a BMI <20 kg/m2 had a significantly inferior prognosis compared to those with a greater BMI. Furthermore, multivariate analyses of OS and DFS showed that BMI <20 kg/m2 was an independent predictor of poor prognosis and early recurrence in patients treated with laparoscopic surgery for CRC with curative intent. Interestingly, BMI was significantly correlated with preoperative NLR, which is a surrogate marker for the immune response to malignancy. A lower BMI prior to surgery may be a promising predictor of early recurrence and poor prognosis in patients treated by laparoscopic surgery for CRC.
Many clinical trials have demonstrated that excess body weight (expressed as a BMI of 25 kg/m2 or greater) is associated with a risk of adult malignancies, including esophageal, thyroid, gallbladder, endometrial, kidney, and colon cancer (16). In addition, Wu et al. showed an association between obesity (BMI ≥30 kg/m2), but not overweight (BMI=25-29.9 kg/m2), with increased mortality in patients with CRC based on their overall meta-analysis, although considerable heterogeneity was observed across studies (17). These results were consistent in most but not all of the subsequent sub-group analyses, suggesting that a severely rather than mildly-to-moderately high BMI may have an adverse effect on survival in CRC. Similarly to this finding, a recent meta-analysis showed that the survival rate of obese but not overweight adults with leukemia was lower than that for normal-weight patients with leukemia (18). These results support the observation that cancer-related mortality is increased in obese not overweight people, which was reported by a population-based study (19).
Univariate and multivariate analyses of overall survival in patients with colorectal cancer treated with curative laparoscopic surgery.
Univariate and multivariate analyses of disease-free survival in colorectal cancer patients treated with curative laparoscopic surgery.
Survival curves for patients without lymph node metastasis (stage I-II) of colorectal cancer classified on the basis of body mass index (BMI) prior to laparoscopic surgery: A: disease-free survival (DFS), and B: overall survival (OS).
Epidemiologically, the relationship between the risk of mortality and BMI is known to be U-shaped, with increased risk not only being related to cachexia, associated with a very low BMI, but also to obesity, associated with a very high BMI, whereas emerging data indicate that obesity is associated paradoxically with better prognosis in patients with cancer (20). A recent meta-analysis demonstrated that class 1 obese (BMI=25-35 kg/m2) patients had a 5% lower risk of mortality than those of a healthy weight (21). Moreover, some reports showed that higher BMI protected individuals against poorer outcomes in some malignancies, which is called the ‘obesity paradox’ (22). In cancer cachexia, which is found in patients with advanced cancer with systemic metastasis, systemic inflammation is induced and persists because of hyper-inflammatory cytokine status. This results in decreased protein anabolism and caloric intake, while promoting an increase in protein catabolism, insulin resistance, lipolysis, and resting energy expenditure. Eventually, loss of muscle mass and strength, loss of whole body fat, ineffective antitumor response and impaired immunity occur, which lead to physical disability, diminished quality of life, and reduced survival (23). In contrast, Sinicrope et al. demonstrated from the Adjuvant Colon Cancer Endpoints (ACCENT) database that for patients with CRC (stage II-III) undergoing curative surgery, being underweight (BMI <20 kg/m2) and not having “cancer cachexia” was significantly associated with increased mortality (24). Our data are consistent with this result as our underweight patients (BMI <20 kg/m2) treated by curative laparoscopic surgery for stage I-III disease had a phenotype of early recurrence and poor prognosis. Indeed, the patients in our cohort were not afflicted with cancer cachexia because the majority of our cohort had earlier tumor stages (over 70% of patients in this cohort were node metastasis-negative), and there was no correlation between BMI and serum CRP or albumin levels (data not shown).
Although the mechanism involved in the higher risk of recurrence or poor prognosis in patients with underweight status is not fully understood, by analyzing global gene expression changes in patients with renal cell carcinoma, Hakimi et al. found that the survival advantage in overweight patients might be related to down-regulation of the fatty acid synthase (FASN) gene, and expression levels of the immediately upstream gene for the enzyme acetyl-CoA carboxylase (ACACA), and its encoded protein (22), which are associated with fatty acid metabolism and tumor progression (25). Kuchiba et al. also demonstrated that FASN expression levels were lower among obese patients with CRC in the Nurse's Health Study, and suggested that FASN might confer a selective growth advantage to cells upon nutritional depletion (26).
Another potential mechanism we identified in our study was the significant, inverse correlation between BMI and NLR. Recently, elevated preoperative NLR was found to be associated with poor prognosis and tumor recurrence in various malignancies, including CRC (11, 27). One hypothesis regarding the association between NLR and poor outcome is that relative lymphocytopenia following incremental increase in neutrophils induces a decrease in T4 helper lymphocytes and T8 suppressor lymphocytes, resulting in depletion of innate cellular immunity and increased tumor proliferation and metastasis (28, 29). This evidence suggests that impairment of the anti-tumor immune response might occur in patients who are underweight preoperatively. However, we should further investigate the underlying mechanism of these poorer oncological outcomes to draw definitive conclusions.
In conclusion, although our retrospective study has limitations, including that obese patients as defined in the western world (BMI >30 kg/m2) accounted for fewer than 5% of our cohort, our findings demonstrate that among patients who are underweight before being treated with laparoscopic surgery for CRC tumor phenotypes might be more aggressive than in patients of normal weight or even those who are obese. Although the underlying mechanisms were not fully investigated, impairment of antitumor immunity might occur in underweight patients and is correlated with higher NLR. Therefore, the preoperative BMI might be a significant predictor of recurrence and prognosis in patients treated by laparoscopic surgery with curative intent for CRC.
Footnotes
Conflicts of Interest
The Authors have no conflicts of interest to disclose in regard to this study.
- Received May 22, 2015.
- Revision received June 20, 2015.
- Accepted June 24, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
