Abstract
Background/Aim: Nutritional status is strongly associated with cancer prognosis. The aim of this study was to identify the most useful combination of nutrition-related serum markers for predicting prognosis of patients with colorectal cancer (CRC). Patients and Methods: A total of 523 patients who underwent proctocolectomies for CRC at our hospital were enrolled in this study. Serum concentrations of albumin, cholinesterase and total cholesterol, and total peripheral lymphocyte count (TLC) were used as nutrition-related markers. Results: In multivariate analysis of nutrition-related markers, serum albumin and cholinesterase levels were found to be independent prognostic indicators. Cut-off values from receiver operating characteristic analyses were used to sort patients as ChEHigh or ChELow (serum cholinesterase level ≥ or <221.5), and as AlbHigh or AlbLow (serum albumin level ≥ or <3.85). We then sorted them into three groups: ChEHigh/AlbHigh (Group A); ChEHigh/AlbLow or ChELow/AlbHigh (Group B); and ChELow/AlbLow (Group C). Their 5-year overall survival rates differed significantly (Group A: 81.6%, Group B: 62.1%, Group C: 42.7%, p<0.0001); as did their 5-year disease-specific survival rates (Group A: 90.1%, Group B: 73.8%, Group C: 62.2%, p<0.0001). Conclusion: The combination of serum cholinesterase and albumin levels is useful for predicting the prognosis of patients with CRC.
Colorectal cancer (CRC) is one of the most commonly diagnosed cancers. Recent advances in surgical techniques, perioperative management, and chemotherapy, including the use of molecular-targeting drugs, have improved the prognosis of patients with CRC; however, it still ranks fourth with respect to cancer-related deaths worldwide (1).
Patient outcomes can vary widely even for those with the same stage of the same type of cancer, mainly because of differences in patient-related factors, including inflammation, immunocompetence, and nutrition. Nutritional status is an important predictor of postoperative morbidity and mortality (2-4), and various tools for assessing nutritional status have been reported (5-7). Recent studies also demonstrated a close association between nutritional status and prognosis in cancer patients, including patients with CRC (8-10).
Serum markers are frequently used to assess nutritional status in the routine clinical setting because they can be measured quickly, easily, and non-invasively. Commonly measured serum markers that reflect patients' nutritional status include albumin, total cholesterol, cholinesterase, and total peripheral lymphocyte count (TLC). Nutritional-status scoring systems based on these markers include the Controlling Nutritional Status (CONUT) score and prognostic nutritional index (PNI). CONUT scores include serum concentrations of albumin and cholesterol and TLC (11), whereas PNI includes serum concentrations of albumin and TLC (12). The CONUT score and PNI are both closely associated with prognosis of various cancers, including CRC (13-16). Notably, these indicators are more useful in predicting prognosis of CRC patients than any of their component markers used alone (17). However, which combination best predicts the prognosis of CRC patients is unclear. Therefore, the current study aimed to identify the best combination of nutrition-related serum markers to predict prognosis of CRC patients.
Patients and Methods
Patients. We retrospectively enrolled 523 CRC patients who underwent proctocolectomies at our hospital between January 2007 to December 2015 into this study. Clinicopathological findings were determined according to the 8th edition of the Japanese Classification of Colorectal Carcinoma (18). Patients were periodically checked for early recurrence by diagnostic imaging (chest X-ray, colonoscopy, ultrasonography, and computed tomography). Causes of death and patterns of recurrence were determined by reviewing medical records, including laboratory data, ultrasonography, computed tomography, scintigrams, and laparotomies, or by direct inquiry with family members. Serum concentrations of albumin, cholinesterase, total cholesterol, and TLC from patients' preoperative blood test results (taken within 1 month before surgery) were collected. PNI was calculated as follows: 10 × Alb concentration + 0.005 × TLC (12). CONUT score was calculated based on serum albumin and total cholesterol level, and TLC. These factors were scored according to cut-off values, and the sum of the scores was used as the CONUT score (11).
Ethical approval. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Statistical analysis. The Chi-square was used to compare distribution of individual variables between patient groups. Receiver operating characteristic (ROC) analysis was used to determine the area under the curve (AUC) for survival analysis and optimal cut-off value of each variable. Survival curves were calculated according to the Kaplan–Meier method. Differences between the curves were examined with the log-rank test. Multivariate analysis of factors considered prognostic of overall survival (OS), with Cox's proportional hazards model and a stepwise procedure were used. p<0.05 was considered significant. GraphPad Prism (GraphPad Software, Inc., La Jolla, CA, USA) and SPSS statistics version 24.0 (SPSS Inc, Chicago, IL, USA) software were used for statistical analyses.
Results
Mean (±standard deviation) serum concentrations were as follows: cholinesterase: 266.2±88.4 U/l, albumin: 3.9±0.6 g/dl, and total cholesterol: 189.1±39.0 mg/dl, and mean TLC was 1494±572.2/μl. In ROC analysis of these markers for OS, AUC values were as follows: cholinesterase: 0.689, albumin: 0.714, cholesterol: 0.610, and TLC: 0.537. In multivariate analysis of these four markers, serum cholinesterase and albumin levels were found to be independent prognostic indicators (Table I). Therefore, serum cholinesterase and albumin levels were used for further analysis. ROC analysis with respect to OS indicated that optimal cut-off values of serum cholinesterase and albumin levels were 221.5 U/l and 3.85 g/dl, respectively. Based on these results, patients were grouped as ChEHigh (serum cholinesterase level ≥221.5; n=370) or ChELow (serum cholinesterase level <221.5; n=153), and as AlbHigh (serum albumin level ≥3.85; n=346) or AlbLow (serum albumin level <3.85; n=177). Five-year OS rates differed significantly between the ChEHigh group (77.8%) and the ChELow group (49.1%; p<0.0001, Figure 1a), and between the AlbHigh group (80.1%) and the AlbLow group (48.1%; p<0.0001; Figure 1b).
To determine whether the combination of serum cholinesterase and albumin levels was more useful than the other individual indicators in predicting CRC prognosis, patients were sorted as ChEHigh/AlbHigh (Group A), either ChEHigh/AlbLow or ChELow/AlbHigh (Group B), and ChELow/AlbLow (Group C). The combination of serum cholinesterase and albumin level was significantly related to age (p<0.0001), tumor size (p<0.0001), tumor location (p<0.0001), depth of invasion (p<0.0001), distant metastasis (p<0.0001), and lymphatic invasion (p<0.0001; Table II).
Five-year OS rates differed significantly in Group A (81.6%), Group B (62.1%) and Group C (42.7%; p<0.0001, Figure 2a); as did 5-year disease-specific survival rates (Group A: 90.1%; Group B: 73.8%; Group C: 62.2%; p<0.0001; Figure 2b). Then the patients in Groups A, B, and C were assigned 0, 1, and 2, respectively, according to our previous study (19). AUC values for OS were as follows: serum albumin level: 0.714, serum cholinesterase level: 0.689, and the combination of serum cholinesterase and albumin level: 0.721, which indicate that the combination of serum cholinesterase and albumin levels is a more accurate predictor of prognosis of patients with CRC than are either serum albumin or cholinesterase level alone. Furthermore, the AUC value of combined serum cholinesterase and albumin levels was higher than that of PNI (0.693) or CONUT score (0.626). Finally, multivariate analyses (Cox's proportional hazards model with a stepwise procedure) of age, gender, tumor size, histology, tumor location, depth of invasion, lymph node metastasis, distant metastasis, lymphatic invasion, vascular invasion, and combined serum cholinesterase and albumin levels as prognostic factors in patients with CRC showed combined serum cholinesterase and albumin level was an independent prognostic indicator, along with age, depth of invasion, lymph node metastasis, and distant metastasis (Table III).
Cancer-related death significantly varied by group (Group A: 7.5%, Group B: 26.1%, and Group C: 31.6%; p<0.001), as did death from cancer-unrelated disease and accidents (Group A: 8.8%, Group B: 14.7%, Group C: 24.6%; p<0.001, Figure 3).
Discussion
In the current study, we first demonstrated that serum albumin and cholinesterase levels were useful prognostic indicators among nutrition-related serum markers. Serum albumin is produced in the liver and is the most abundant blood plasma protein. Serum albumin level is a standard factor in assessing patient nutritional status. Hypoalbuminemia has been associated with poor survival in advanced cancers (20). Two commonly used prognostic indicators for cancers – the Glasgow prognostic score (GPS) and the CRP-to-albumin ratio — are based on serum C-reactive protein (CRP) and albumin levels. The GPS has been shown to be a prognostic factor in advanced gastrointestinal cancers, including esophageal (21) and colorectal cancer (22). The CRP-to-albumin ratio has been linked to prognosis of CRC patients (23). These results indicate the importance of serum albumin levels as a prognostic indicator in CRC patients.
Our study also showed that serum cholinesterase levels were associated with prognosis of CRC patients. The human body makes two types of cholinesterase: acetylcholinesterase, which is present in the brain, nerve, and red blood cells; and butyryl-cholinesterase, a non-specific cholinesterase enzyme that is present in the liver, pancreas, central nervous system and blood (24, 25). In this study, serum concentration of butyryl-cholinesterase was measured as the serum cholinesterase level. Low serum cholinesterase level is associated with liver damage, inflammation, and malnutrition (26), and is reportedly correlated with prognosis in bladder and urothelial cancers (27, 28). Low serum cholinesterase and albumin levels are also reportedly associated with high all-cause mortality in treatment-naïve cancer patients irrespective of tumor entity or stage (29).
To our knowledge, this report is the first to show that serum cholinesterase level is closely related to prognosis in patients with CRC. As combinations of nutrition-related markers found in blood tests, such as PNI and CONUT score, have been shown to help predict CRC prognosis (14, 16), the combination of serum albumin and cholinesterase levels was determined to be more useful than either factor alone. The AUC for OC of combined serum albumin and cholinesterase levels were not only greater than that of serum albumin or cholinesterase level used alone, but was also greater than the AUCs of PNI and CONUT scores. These results indicate that combined serum albumin and cholinesterase levels provide the most accurate prognosis for CRC among nutrition-related indices of similar clinical convenience.
The combination of serum albumin and cholinesterase level was closely related to both cancer-related deaths and deaths due to other causes. Migita et al. previously used PNI to evaluate the preoperative immune-nutritional status of patients and found that low PNI scores were associated with a higher risk of non-cancer deaths (30). A similar study showed that low PNI scores increased the chance of respiratory failure due to pneumonia in older patients with gastric cancer, compared with patients with high PNI scores (31). Overall, these findings show that poor nutritional status increases risk of death from non-cancer-related diseases after surgery, which implies that nutritional markers can help predict patient outcomes.
Our study had some limitations. First it was a retrospective study and was therefore subject to bias. Second, as both serum albumin and cholinesterase are synthesized in the liver, their levels are affected by not only nutrition status but also by liver function. Third, the number of patients included in the current study was small; further large-scale, prospective, randomized, controlled trials are needed to confirm the results.
In conclusion, the combination of serum albumin and cholinesterase levels provide accurate prognostic information for patients with CRC. Because serum markers can be measured quickly, easily, and non-invasively, the combination of serum albumin and cholinesterase level presents a useful clinical index.
Acknowledgements
The Authors thank Marla Brunker, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.
Footnotes
Authors' Contributions
Study conception and design: Saito; Acquisition of data: Yamamoto, Uejima, Tanio, Tada, Matsunaga, Sakamoto, Honjo, and Ashida; Analysis and interpretation of data: Yamamoto and Saito; Drafting of manuscript: Yamamoto and Saito; Critical revision: Fujiwara; Final approval of the article: all authors.
Conflicts of Interest
The Authors declare that they have no conflicts of interest regarding this study.
- Received January 8, 2019.
- Revision received January 21, 2019.
- Accepted January 22, 2019.
- Copyright© 2019, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved