Nutritional status is a significant indicator for physical conditions, especially for cancer patients. Several nutritional assessment indexes including nutritional risk index, prognostic nutritional index and the modified Glasgow prognostic score have been applied to predict outcomes of cancers(Reference Sun, Chen and Xu1–Reference Minami, Ogata and Ihara4). Similarly, the pre-operative Controlling Nutritional Status (CONUT) score is known as a relatively objective, validated and emerging index to assess patients’ nutritional status, which is derived from serum albumin level, total lymphocyte count, as well as total cholesterol concentration(Reference Ignacio de Ulíbarri, González-Madroño and de Villar5). A series of retrospective cohort studies had demonstrated that CONUT score was closely related with the prognosis of some type of cancers, including gastric carcinoma(Reference Kuroda, Sawayama and Kurashige6), hepatocellular carcinoma(Reference Lin, Ruan and Jia7), oesophageal carcinoma(Reference Toyokawa, Kubo and Tamura8), colorectal carcinoma(Reference Hayama, Ozawa and Okada9), breast carcinoma(Reference Li, Li and Wang10), ovarian carcinoma(Reference Li, Zhang and Ji11), lymphoma(Reference Nagata, Kanemasa and Sasaki12) and lung carcinoma(Reference Gul, Metintas and Ak13). In addition, the systematic review and meta-analysis about gastric carcinoma, hepatocellular and colorectal cancer have also been released recent years(Reference Takagi, Buettner and Ijzermans14–Reference Takagi, Domagala and Polak16).
The utility of the CONUT score on survival outcomes of patients accepting nephrectomy due to upper tract urothelial carcinoma (UTUC) or renal cell carcinoma (RCC) was first reported in 2017(Reference Ishihara, Kondo and Yoshida17). Since then, this effect has been further examined by clinicians all over the world. However, the value of the pre-operative CONUT score in urinary system tumours has not been proved with evidenced-based medicine until now.
Consequently, the first systematic review and meta-analysis, which is about the relationship between the CONUT score and prognosis of patients with UTUC or RCC, was performed by our team to provide more valid evidence.
Material and methods
Search strategy
This study was designed and conducted on the basis of the Preferred Reporting Items for Systematic Reviewers and Meta-Analyses (PRISMA) guidelines(Reference Moher, Liberati and Tetzlaff18). Major public medical and scientific database including PubMed, Embase along with Web of science were systematically searched to seek out all original articles, which examined the association between the pre-operative CONUT score and prognosis of patients with UTUC or RCC. Search terms were as follows: kidney neoplasm OR kidney cancer OR kidney cancers OR renal neoplasm OR renal neoplasms OR renal cancer OR renal cancers OR neoplasm, kidney OR neoplasm, renal OR neoplasms, renal OR neoplasms, kidney OR cancer of kidney OR cancer, renal OR cancers, renal OR cancer of the kidney OR cancer, kidney OR cancers, kidney OR upper tract urothelial carcinoma AND the Controlling Nutritional Status score OR CONUT score AND nephrectomy. The systematic search was performed on February 2021.
Study selection
Inclusion criteria were presented as follows: (1) patients undergoing radical or partial nephrectomy for UTUC or RCC; (2) the CONUT score of patients was assessed before surgical operation and (3) prognosis indicators including overall survival (OS), cancer-specific survival (CSS) as well as disease-free survival (DFS) were reported.
Exclusion criteria were shown as follows: (1) human studies on other cancers were excluded; (2) patients treated with non-operation therapy were excluded as well; and (3) reviews, case reports, comments, letters, as well as meeting abstracts were also excluded.
Data extraction
After deleting same articles, the titles, abstracts and full text of the remaining records were independently checked and approved by two authors. If there existed disagreement, the divergence was resolved by the third investigator. Using a unified form, variables including study type, issuing time, publishing country of study, essential information of patients, cut-off value of the CONUT score, as well as long-term survival outcomes were extracted.
Quality assessment
A quality assessment system which was on the basis of the Newcastle–Ottawa Scale for cohort studies were applied to evaluate methodological quality of all included articles(Reference Stang19). The minimum and maximum score for Newcastle–Ottawa Scale of included studies were 6 and 9, respectively, and the study would be considered as a high-quality study if total score was 6 or higher(Reference Man, Pang and Zhou20). The quality and level of evidence for all included studies were evaluated independently by two authors and the results were presented in Supplementary Table S1.
Statistical analysis
The value of the CONUT score on survival was examined using Review Manager 5·4 (Cochrane Collaboration, 2014). The pooled hazard ratios (HR) with 95 % CI were calculated for dichotomous variables by applying the inverse variance method. To assess heterogeneity among studies, we calculated the I2 values and the χ 2 test. P-value < 0·05 was regarded as statistical significance and I2 values > 50 % indicated that there was heterogeneity among included studies. The random effects meta-analysis was used if the I2 values > 50 %, otherwise, we conducted the fixed effects models.
Results
Study selection was performed according to the procedures of Fig. 1. After deleting the duplications, 95 relevant articles in total were acquired initially using the search strategy above. Then we reviewed the titles and abstracts, and 84 studies were further excluded on account of reporting the CONUT score of non-urinary tumours or for that the content has nothing to do with the CONUT score. The full text of the remaining eleven articles were screened next and five studies of them were excluded for that these articles were examining the prognostic significance of CONUT score on prostate and bladder carcinoma and the intervention of one article was not surgery.
Flow diagram of evidence acquisition.
Finally, six full-text articles were regarded as qualified and included for the pooled analysis, as represented in Table 1 (Reference Bao, Li and Guan21–Reference Ishihara, Kondo and Yoshida26). The survival rates for patients in all included studies were showed in Table 2. Four studies reported that 5-year OS, CSS and DFS in the low CONUT group were better than that in the high CONUT group. These articles came from different countries including China, Japan as well as Germany and were released from 2017 to 2020. All selected articles were single-centre retrospective study and 3529 patients altogether were enrolled with sample sizes ranged from 107 to 1046 cases. All patients included were treated with nephrectomy. Six studies included were considered as high-quality due to the Newcastle–Ottawa Scale score was more than 6.
Literatures about the impact of the CONUT score on patients undergoing nephrectomy for UTUC or RCC
CONUT, Controlling Nutritional Status; UTUC, upper tract urothelial carcinoma; RCC, renal cell carcinoma; OS, overall survival; CSS, cancer-specific survival; DFS, disease-free survival; n.a., not available.
* Score from a maximum of 9 evaluated by the Newcastle–Ottawa quality assessment scale for cohort studies.
Studies investigating the impact of the CONUT score on survival in patients with UTUC or RCC
CONUT, Controlling Nutritional Status; UTUC, upper tract urothelial carcinoma; RCC, renal cell carcinoma; HR, hazard ratio; n.a., not available.
Data are shown for high CONUT group v. low CONUT group unless otherwise indicated. HR is shown with 95 % CI.
* Multivariable analysis.
Impact of the Controlling Nutritional Status score on overall survival
Prognostic value of the CONUT score on OS was examined by all included studies with 3529 patients. Pooled analysis revealed that the CONUT score was related to OS and the OS of high CONUT score group was inferior to that of low-score group (HR 2·32, 95 % CI 1·58, 3·41, P < 0·0001, I2 = 68 %, P = 0·008). Subgroup analysis was further performed in accordance with various primary tumours. In the subgroup analysis, the pre-operative CONUT score was found to be associated with OS of UTUC (HR 1·86, 95 % CI 1·13, 3·08, P < 0·02, I2 = 74 %, P = 0·02) and RCC (HR 3·05, 95 % CI 2·07, 4·49, P < 0·0001, I2 = 0 %, P = 0·93), respectively (Fig. 2(a)).
Forest plots demonstrating long-term outcomes in terms of low CONUT v. high CONUT score. (a) Overall survival; (b) cancer-specific survival and (c) disease-free survival. UTUC, upper tract urothelial carcinoma; RCC, RCC, renal cell carcinoma; CONUT, Controlling Nutritional Status
Impact of the Controlling Nutritional Status score on cancer-specific survival
Six studies involving 3529 patients demonstrated the connection between the pre-operative CONUT score and CSS. Pooled analysis confirmed the statistically significant predictive role of the CONUT score on CSS, and the CSS was better for the low CONUT score group when compared with the high-score group (HR 2·68, 95 % CI 1·69, 4·26, P < 0·0001, I2 = 68 %, P = 0·009). Subgroup analysis revealed that the COUNT score had a potential value to predict the CSS of UTUC (HR 2·24, 95 % CI 1·17, 4·26, P = 0·01, I2 = 79 %, P = 0·009) and RCC (HR 3·47, 95 % CI 2·12, 5·68, P < 0·00001, I2 = 0 %, P = 0·81) (Fig. 2(b)).
Impact of the Controlling Nutritional Status score on disease-free survival
Data about the effect of pre-operative CONUT score on DFS were synthesised from 5 studies including 2894 patients. Our evidence synthesis revealed that the higher CONUT score was associated with reduced DFS, in other words, the high CONUT score group was more likely to relapse than the low-score group (HR 1·62, 95 % CI 1·37, 1·92, P < 0·00001, I2 = 26 %, P = 0·25). In the subgroup analysis, this result was also applied to UTUC (HR 1·54, 95 % CI 1·28, 1·84, P < 0·00001, I2 = 5 %, P = 0·35) and RCC (HR 2·21, 95 % CI 1·42, 3·45, P = 0·0005, I2 = 11 %, P = 0·29), respectively (Fig. 2(c)).
Discussion
This systematic review and meta-analysis was conducted to reveal the impact of CONUT score on the survival of patients with UTUC or RCC undergoing nephrectomy. The relationship between the CONUT score and OS had been demonstrated previously; however, the value of CONUT score on CSS and DFS remained controversial on account of discrepancy in reported articles. Multivariable analyses were performed by all selected studies, respectively, and the evidence that the high CONUT score was an independent risk factor for OS, CSS and DFS was demonstrated by most of the included studies. However, Bao et al. found that there was no critical association between the CONUT score and CSS(Reference Bao, Li and Guan21). Ishihara et al. reported that the pre-operative CONUT score had little significance in predicting DFS. Therefore, a systematic review and meta-analysis was needed urgently to guide clinical diagnosis and treatment. Our pooled analysis revealed that UTUC and RCC patients with high CONUT score had inferior OS, CSS and DFS when compared with those having low score. The same conclusion was also demonstrated by other tumours such as gastric carcinoma, hepatocellular carcinoma and colorectal cancer(Reference Takagi, Buettner and Ijzermans14–Reference Takagi, Domagala and Polak16).
As we all know, the CONUT score is assessed by the level of serum albumin, peripheral lymphocyte and total blood cholesterol. Each index of pre-operative CONUT score had been reported to be related to the prognosis of different tumours. There exist a lot of reasons to account for the predictive effect of the CONUT score on the prognosis of UTUC and RCC. First of all, albumin, a major element of serum total proteins, plays a great role in reflecting nutrition status and metabolic status(Reference Fruchtenicht, Poziomyck and Kabke3,Reference Lv, An and Sun27) . On the other hand, albumin was strongly related to the extent of the systemic inflammatory response by regulating concentrations of C-reactive protein(Reference Al-Shaiba, McMillan and Angerson28,Reference McMillan, Watson and O’Gorman29) . For patients with renal cancer, serum albumin has been proved as an independent prognostic factor(Reference Stenman, Laurell and Lindskog30). In addition, low albuminemia could lead to poor prognosis of patients with cancer by affecting immune response(Reference Al-Shaiba, McMillan and Angerson28). Secondly, lymphocyte count can reflect the level of immunological and systematic inflammatory reaction(Reference Chen, Li and Huang31). In the tumour microenvironment, the high lymphocyte count indicates the body’s immunoreaction against tumour. The antitumour effect of lymphopenia is achieved by promoting cell apoptosis, restraining the growth and migration of tumour cell, and mediating cytotoxicity reaction(Reference Mantovani, Allavena and Sica32). Lymphopenia was reported to be independently correlated with the inferior survival outcomes of clear cell renal cancer(Reference Saroha, Uzzo and Plimack33). What’ s more, neutrophil to lymphocyte and platelet to lymphocyte ratio, two kinds of indexes based on the lymphocyte count, had been reported to be prognostic factor for UTUC(Reference Itami, Miyake and Tatsumi34,Reference Son, Hwang and Jung35) . Finally, cholesterol, an essential component of cell membrane, plays a crucial role in maintaining the cellular function by affecting the organisation, dimensions and fluidity of plasma membrane(Reference Harikumar, Potter and Patil36). Previous study demonstrated that low cholesterol level was connected with worse outcomes of patients with RCC(Reference Ko, Park and Lee37). A recent study has shown that high serum cholesterol levels can enhance the antitumour effect of natural killer cells in mice(Reference Qin, Yang and Li38). In addition, the higher level of serum cholesterol also indicates the lower overall cancer risk(Reference Wu, Teng and He39). However, the mechanism was unclear. Although in an initial cancer process the expected immunological reaction would be an increase of total lymphocytes, in the case of more serious situation, the nutritional deficit in albumin and cholesterol will prevent the development of the immunological reaction which would be required for the regeneration of the cell membranes. This is also what happens with the energetic and protein needs that are covered by the albumin: In fact, it is the latter one that is in charge of the energetic and protein substrates that are required for cell development as well as lymphocyte proliferation. As a consequence, nutrient deficiency affects all three parameters and is reflected in the controlling nutritional prognostic and clinical risk index. Patients with high CONUT score had low albumin, lymphocyte count and serum cholesterol level. Consequently, it is not difficult for us to understand that the survival rates of high CONUT score group were worse than the low-score group.
What’s more, malnutrition is a common clinical feature of cancer patients and even can evolve into cachexia, especially in advanced cancer patients. Previous evidence supported that malnutrition could influence progression and survival of patients with tumour(Reference Mantzorou, Koutelidakis and Theocharis40). In addition, nutritional intervention of perioperative period has the effect of increasing treatment tolerance and improving prognosis of cancer patients(Reference Paccagnella, Morassutti and Rosti41). However, it has not been confirmed that the perioperative nutritional intervention can affect the long-term outcomes of oncological patients. The present study indicates that the perioperative nutritional support is of importance for survival of patients with RCC and UTUC. The CONUT score of patients is easy to assess and it can help clinicians to identify patients who need nutritional support during the perioperative period.
We should acknowledge that there exist some limitations in the present study. Firstly, although the study included the up-to-date and the most complete articles, there were only six studies involved in the pooled analysis. Secondly, all articles included were retrospective single-centre research, and we can only get the effective data of long-term outcomes without short-term outcomes. In addition, the cut-off values of dividing high CONUT score and low CONUT score were different for included studies. Consequently, further studies are needed to demonstrate the value of pre-operative CONUT score on prognosis of RCC and UTUC. The association between CONUT and prognosis is not necessarily causal. The CONUT score could be a proxy for another aspect of illness. However, CONUT is clinically useful, in that it predicts outcomes.
Conclusions
The first systematic review and meta-analysis that examined the role of pre-operative CONUT score on prognosis of patients accepting nephrectomy for UTUC or RCC was performed by our team. Our pooled analysis revealed that the CONUT score was an effective and convenient index to predict survival of UTUC and RCC patients undergoing surgery, which could provide a reliable reference for clinician.
Acknowledgement
I would like to show my deepest gratitude to all authors who have provided me with valuable help in the process of finishing this study.
This work was supported by a Key Project of National Natural Science Foundation of China (Grant number: 8177060452).
J. C. and D. C. contributed to the study conception and design, and the drafting of the manuscript. Z. P. and P. S. were responsible for data extraction. Z. L. and L. Y. were responsible for data analysis and interpretation. Data integrity and accuracy were confirmed by L. W. and J. Z. Q. W. and Q. D. contributed to the critical revising of the final draft. All authors read and approved the final manuscript.
There are no conflicts of interest.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S0007114521002889
