Prognostic role of the controlling nutritional status (CONUT) score in patients with biliary tract cancer: a meta-analysis

Abstract

Background

Previous reports have not reached consistent results regarding the prognostic significance of the controlling nutritional status (CONUT) score in biliary tract cancer (BTC). Therefore, the present meta-analysis was conducted to investigate the precise role of the CONUT score in predicting the prognosis of BTC.

Methods

Electronic platforms including Web of Science, PubMed, Cochrane Library, and Embase were comprehensively searched up to May 2, 2023. We also determined combined hazard ratios (HRs) and 95% confidence intervals (CIs) to estimate the role of the CONUT score in predicting the prognosis of patients with BTC.

Results

Ten studies involving 1,441 patients were included in the present study. Nine studies treated patients with surgical resection, and one study used percutaneous transhepatic biliary stenting (PTBS) plus 125I seed intracavitary irradiation. Based on the combined data, a higher CONUT score significantly predicted dismal overall survival (OS) (HR = 1.94, 95%CI = 1.41–2.66, p < 0.001), inferior recurrence-free survival (RFS) (HR = 1.79, 95%CI = 1.48–2.17, p < 0.001) in BTC, and low differentiation (OR = 1.57, 95%CI = 1.15–2.14, p = 0.004). Nonetheless, the CONUT score was not related to sex, lymph node metastasis, microvascular invasion, perineural invasion, distant metastasis, TNM stage, or tumor number in patients with BTC.

Conclusion

Higher CONUT scores significantly predicted worse OS and RFS in patients with BTC. Moreover, BTC patients with high CONUT scores tended to have poor tumor differentiation. The CONUT score could help clinicians stratify high-risk patients with BTC and devise individualized treatment plans.

KEY MESSAGES

1. As far as we know, this study is the first to analyze whether pretreatment CONUT is significant for predicting the prognosis of BTC.

2. A high CONUT significantly predicted worse OS and RFS in BTC patients.

3. CONUT could help clinicians stratify high-risk BTC patients and devise individualized treatment plans.

Keywords:

• CONUT
• biliary tract cancer
• meta-analysis
• evidence-based medicine
• prognosis

Introduction

Biliary tract cancer (BTC) refers to invasive carcinomas such as gallbladder cancer (GBC), extrahepatic cholangiocarcinoma (ECC), and intrahepatic cholangiocarcinoma (ICC) [1]. BTC accounts for approximately 3% of all gastrointestinal cancers and ranks second among hepatobiliary cancers in terms of morbidity, followed by hepatocellular carcinoma [2,3]. According to the disease stage, BTC has a dismal prognostic outcome, with 5-year survival rates of 2%-15%, 2%-30% and 2%-70% for ICC, ECC, and GBC, respectively [4]. Surgery is a potential radical therapy for BTC. Nonetheless, as many patients with BTC do not have early symptoms, they still have a poor prognosis. Over 65% of patients with BTC are not candidates for surgery at the time of diagnosis, and their 5-year survival and 1-year relapse rates are 5–15% and 67%, respectively [5,6]. Therefore, the identification of new and reliable biomarkers for predicting BTC prognosis is urgently needed.

Increasing evidence has shown that nutritional status is an important factor in cancer patients because it can predict treatment tolerance and tumor development [7]. Many blood test-derived biomarkers, including the prognostic nutritional index (PNI) [8], albumin-to-globulin ratio (AGR) [9], geriatric nutritional risk index (GNRI) [10], and C-reactive protein-to-albumin ratio (CAR) [11] have been identified as effective prognostic markers in various cancers. Notably, the controlling nutritional status (CONUT) score represents a verified nutritional evaluation system, which contributes to the comprehensive evaluation of host immunocompetence and anabolism [12]. The CONUT score is based on serum albumin levels, total cholesterol (TC) content, and lymphocyte count [12] (Table 1). The CONUT has attracted wide interest as a biomarker for predicting cancer prognosis in breast cancer (BC) [13], peripheral T-cell lymphoma [14], esophageal cancer [15], and thyroid cancer [16]. Previous studies have analyzed whether the CONUT score could be adopted to predict the prognosis of patients with BTC, but their findings have been inconsistent [17–26]. For instance, some studies have shown that the CONUT score is important for predicting BTC prognosis [19,20,23,26], whereas others have suggested that CONUT score is not related to BTC prognosis [21,24]. Therefore, the present meta-analysis reviewed the most recent data to identify the significance of the CONUT score in predicting the prognosis of patients with BTC.

Table 1. The scoring criteria for the CONUT.

	Degree
Parameters	Normal	Light	Moderate	Severe
Serum albumin (g/dL)	3.5–4.5	3.0–3.49	2.5–2.99	<2.5
Score	1	2	4	6
Total lymphocyte (count/mm^3)	≥1600	1200–1599	800–1199	<800
Score	0	1	2	3
Total cholesterol (mg/dL)	>180	140–180	100–139	<100
Score	0	1	2	3
CONUT score (total)	0–1	2–4	5–8	9–12

CONUT: controlling nutritional status.

Materials and methods

Study guideline

The present study was conducted in accordance with the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement (Supplementary File 1) [27].

Ethnics statement

As our analysis did not directly involve participants, and we extracted the necessary data from previous publications, institutional review board and ethics committee approval was not required for this meta-analysis.

Literature search

We comprehensively searched electronic platforms, including Web of Science, PubMed, Cochrane Library, and Embase, using Medical Subject Heading (MeSH) terms combined with text words: (controlling nutritional status or CONUT) and (biliary tract tumor, biliary tract cancer, biliary tract neoplasm, cholangiocarcinoma, bile duct cancer, gallbladder tumor, gallbladder neoplasm, gallbladder cancer, and gallbladder carcinoma). The literature search was conducted from the inception of each database until May 2, 2023. We manually reviewed the reference lists of the included studies to avoid omitting relevant articles.

Study eligibility criteria

Eligible studies were included based on the following criteria: (1) BTC diagnosed through histology, (2) COUNT before treatment was measured, (3) studies that investigated the association between CONUT score and survival outcomes in BTC, (4) hazard ratios (HRs) and associated 95% confidence intervals (CIs) of survival analysis were obtained or calculated based on available data, (5) the threshold CONUT score was determined, and (6) English-language articles. Studies that met the following criteria were excluded: (1) reviews, case reports, meeting abstracts, letters, and comments; (2) studies that did not provide survival data; and (3) animal studies.

Data collection and quality evaluation

Two reviewers (HJ and ZW) reviewed eligible studies and independently collected the required data. Any disagreements were resolved through negotiations until a consensus was reached. The following data were collected from the included articles: first author, country, publication year, sample size, sex, age, study duration, histology, treatment, study design, threshold CONUT score, study center, follow-up, neoadjuvant or adjuvant treatments, survival endpoints, survival analysis type, HRs with corresponding 95%CIs, and clinicopathological characteristics. The Newcastle–Ottawa Scale (NOS) was used to assess the quality of the included studies [28]. The NOS consists of three domains–selection, comparability, and outcome–with a score of 0–9 stars. An article with an NOS score ≥6 was considered high-quality.

Statistical analysis

The pooled HRs and corresponding 95%CIs were calculated to estimate the significance of the CONUT score in predicting the prognosis of patients with BTC. Heterogeneity among studies was assessed using I² statistics [29] and the chi-square Q test [30]. We selected the random-effects model [31] for the calculation in the case of distinct heterogeneity (I²>50% or p < 0.10); otherwise, we selected the fixed-effects model [32]. An appropriate calculation model was selected based on heterogeneity levels. We also performed subgroup analysis to detect potential sources of heterogeneity. Correlations between the CONUT scores and clinicopathological features were assessed by combining odds ratios (ORs) and 95%CIs. Moreover, a sensitivity analysis was carried out to assess whether individual studies affected the overall outcomes. Funnel plots and Begg’s test were used to assess publication bias. Stata software version 12.0 (Stata Corp., College Station, TX, USA) was used for the statistical analysis. Statistical significance was set at p < 0.05 (two-sided).

Results

Literature search process

As shown in Figure 1, 79 articles were initially obtained, and 55 remained after the removal of duplicates. Through title and abstract screening, 44 irrelevant studies or fundamental research in cell biology were removed. Subsequently, we read the full text of 11 articles and discarded one because it did not provide survival information. Finally, the present meta-analysis included ten studies with 1,441 patients [17–26] (Figure 1; Table 2).

Figure 1. Flow chart demonstrating the study selection process.

Table 2. Baseline characteristics of included studies.

Study	Year	Country	Sample size	Gender (M/F)	Age (years) Median(range)	Study period	Histology	Treatment	Neoadjuvant /adjuvant treatments	Treatment intent	Cut-off value	Follow-up (months) Median(range)	Survival endpoints	Survival analysis	NOS score
Cui, P.	2018	China	73	44/29	64.7	2012–2017	ECC	PTBS+^125I	No	Palliative	≥2	1–30	OS	Multivariate	7
Miyata, T.	2018	Japan	71	45/26	64.8	2002–2016	ICC	Surgical resection	Neoadjuvant chemotherapy: n = 2	Curative	≥2	36.9	OS, RFS	Multivariate	8
Utsumi, M.	2019	Japan	45	25/20	75(38–95)	2008–2017	GBC	Surgical resection	No	Curative	≥3	35(3–102)	OS	Multivariate	7
Wang, A.	2020	China	94	49/45	54.5(31–73)	2010–2019	ECC	Surgical resection	No	Curative	≥3	1–60	OS, RFS	Multivariate	7
Zheng, Y.	2020	China	167	95/72	57.2	2012–2018	ICC	Surgical resection	Adjuvant chemotherapy: n = 64	Curative	≥3	14(4–58)	OS, RFS	Multivariate	7
Sun, L. J.	2021	China	371	199/172	61.5(23–88)	2002–2017	BTC	Surgical resection	No	Curative	≥2	16	OS	Multivariate	6
Terasaki, F.	2021	Japan	149	115/34	71(39–85)	2002–2016	ECC	Surgical resection	Adjuvant therapy: n = 15	Curative	≥3	53.5	OS, RFS	Multivariate	8
Yugawa, K.	2021	Japan	78	55/23	66(39–87)	1998–2018	ICC	Surgical resection	No	Curative	≥2	1–60	OS, RFS	Univariate	7
Asakura, R.	2022	Japan	169	112/57	70(63–75)	2000–2019	ECC	Surgical resection	Adjuvant chemotherapy: n = 82	Curative	≥3	1–60	OS	Univariate	9
Mito, M.	2022	Japan	224	152/72	72(31–90)	2006–2020	BTC	Surgical resection	Adjuvant chemotherapy: n = 76	Curative	≥4	83.0(6.6–187.6)	OS, RFS	Multivariate	7

ICC: intrahepatic cholangiocarcinoma; ECC: extrahepatic cholangiocarcinoma; BTC: biliary tract cancer; GBC: gallbladder cancer; OS: overall survival; RFS: recurrence-free survival; PTBS: percutaneous transhepatic biliary stenting; NOS: Newcastle-Ottawa Scale; M: male; F: female.

Features of included articles

Table 2 displays fundamental features of the included articles [17–26]. Six studies were performed in Japan [18,19,23–26] and four in China [17,20–22]. The sample size of these studies ranged from 45 to 371, with the median value of 121.5. All studies used a retrospective design. Four articles included ECC cases [17,20,23,25]; three, ICC cases [18,21,24]; two, BTC cases [22,26]; and one, GBC cases [19]. Nine studies treated patients with surgical resection [18–26], and one study adopted percutaneous transhepatic biliary stenting (PTBS) plus 125I seed intracavitary irradiation [17]. Five studies used ≥3 as the cut-off value for CONUT [19–21,23,25], four studies selected ≥2 [17,18,22,24], and one study adopted ≥4 [26]. All 10 studies mentioned the role of CONUT in predicting overall survival (OS) in BTC [17–26]. Six studies provided data on the association between CONUT and recurrence-free survival (RFS) in patients with BTC [18,20,21,23,24,26]. Eight studies calculated HRs together with 95%CIs by multivariable regression [17–23,26], and two studies used univariate analysis [24,25]. For the included articles, NOS scores were 6–9 (median, 7), suggesting the articles were high quality (Table 2). The details of NOS scores are shown in Supplementary File 2.

CONUT and OS in BTC

Ten articles, including 1,441 patients, mentioned the relationship between the CONUT score and OS in BTC [17–26]. According to the pooled data (Figure 2 and Table 3), a higher CONUT score significantly predicted a dismal OS in BTC (HR = 1.94, 95%CI = 1.41–2.66, p < 0.001). Owing to obvious heterogeneity, we selected the random-effects model (I²=79.8%, p < 0.001). Subgroup analysis revealed that a higher CONUT score was still a prognostic factor for OS regardless of country, sample size, treatment, threshold, neoadjuvant/adjuvant treatments, treatment intent, and type of survival analysis (all p < 0.05; Table 3). Moreover, the CONUT score significantly predicted worse OS in patients with ECC, BTC, and GBC (Table 3).

Figure 2. Forest plot verify the association between the CONUT and overall survival (OS) in patients with BTC.

Table 3. Subgroup analysis of association between CONUT and OS in patients with BTC.

Variables	No. of studies	No. of patients	Effects model	HR (95%CI)	p	Heterogeneity I^2(%) Ph
Total	10	1,441	Random	1.94(1.41–2.66)	<0.001	79.8	<0.001
Country
China	4	705	Random	1.82(1.06–3.14)	0.030	88.1	<0.001
Japan	6	736	Fixed	1.94(1.55–2.43)	<0.001	1.5	0.407
Sample size
<100	5	361	Random	2.75(1.64–4.59)	<0.001	55.0	0.064
≥100	5	1,080	Random	1.52(1.11–2.09)	0.009	78.4	0.001
Histology
ICC	3	316	Random	1.46(0.79–2.70)	0.224	73.3	0.024
ECC	4	485	Random	2.35(1.50–3.67)	<0.001	59.1	0.062
BTC	2	595	Fixed	1.64(1.29–2.09)	<0.001	4.3	0.307
GBC	1	45	–	4.69(1.48–14.88)	0.009	–	–
Treatment
PTBS+^125I	1	73	–	2.02(1.08–3.80)	0.028	–	–
Surgical resection	9	1,368	Random	1.93(1.37–2.72)	<0.001	81.4	<0.001
Cut-off value
≥2	4	593	Fixed	1.65(1.28–2.11)	<0.001	13.1	0.327
≥3	5	624	Random	2.21(1.20–4.07)	0.011	87.8	<0.001
≥4	1	224	–	1.91(1.31–2.76)	0.001	–	–
Survival analysis
Multivariate	8	1,194	Random	2.09(1.42–3.09)	<0.001	84.0	<0.001
Univariate	2	247	Fixed	1.57(1.08–2.28)	0.018	0	0.608
Neoadjuvant /adjuvant treatments
Yes	5	780	Random	2.29(1.35–3.87)	0.011	82.0	<0.001
No	5	661	Random	1.71(1.13–2.59)	0.002	70.4	0.009
Treatment
Palliative	1	73	–	2.02(1.08–3.80)	0.028	–	–
Curative	9	1,368	Random	1.93(1.37–2.72)	<0.001	81.4	<0.001

ICC: intrahepatic cholangiocarcinoma; ECC: extrahepatic cholangiocarcinoma; BTC: biliary tract cancer; GBC: gallbladder cancer; OS: overall survival; PTBS: percutaneous transhepatic biliary stenting; CONUT: controlling nutritional status.

CONUT and RFS in BTC

Six studies, including 783 patients [18,20,21,23,24,26] investigated whether the CONUT score was significant in predicting RFS in patients with BTC. According to the combined results (Figure 3 and Table 4), a higher CONUT score was significantly associated with shortened RFS in patients with BTC (HR = 1.79, 95%CI = 1.48–2.17, p < 0.001). In the subgroup analysis, a higher CONUT score still significantly predicted RFS, regardless of country, sample size, histology, or cutoff value (all p < 0.05; Table 4).

Figure 3. Forest plot verify the association between the CONUT and recurrence-free survival (RFS) in patients with BTC.

Table 4. Subgroup analysis of association between CONUT and RFS in patients with BTC.

Variables	No. of studies	No. of patients	Effects model	HR (95%CI)	p	Heterogeneity I^2(%) Ph
Total	6	783	Fixed	1.79(1.48–2.17)	<0.001	49.2	0.080
Country
China	2	261	Random	2.46(1.04–5.81)	0.041	77.0	0.037
Japan	4	522	Fixed	1.69(1.36–2.12)	<0.001	34.1	0.208
Sample size
<100	3	243	Random	1.89(1.03–3.48)	0.041	75.0	0.018
≥100	3	540	Fixed	1.95(1.51–2.52)	<0.001	0	0.620
Histology
ICC	3	316	Fixed	1.46(1.14–1.88)	0.003	0	0.569
ECC	2	243	Random	2.70(1.41–5.18)	0.003	59.5	0.116
BTC	1	224	–	2.25(1.42–3.55)	0.001	–	–
Cut-off value
≥2	2	149	Fixed	1.37(1.00–1.86)	0.048	0	0.435
≥3	3	410	Random	2.20(1.42–3.41)	<0.001	54.2	0.113
≥4	1	224	–	2.25(1.42–3.55)	0.001	–	–
Survival analysis
Multivariate	5	705	Fixed	2.02(1.62–2.52)	<0.001	24.0	0.262
Univariate	1	78	–	1.25(0.85–1.83)	0.252	–	–
Neoadjuvant /adjuvant treatments
Yes	4	611	Fixed	1.88(1.49–2.37)	<0.001	0	0.721
No	2	172	Random	2.15(0.69–2.72)	0.188	87.5	0.005

ICC: intrahepatic cholangiocarcinoma; ECC: extrahepatic cholangiocarcinoma; BTC: biliary tract cancer; GBC: gallbladder cancer; RFS: recurrence-free survival; PTBS: percutaneous transhepatic biliary stenting; CONUT: controlling nutritional status.

CONUT and clinicopathological factors

Six studies with 1,076 patients [18,20–23,26] provided information on the relationship between the CONUT score and the clinicopathological characteristics of BTC. Based on the pooled results, a higher CONUT score significantly predicted dismal differentiation (OR = 1.57, 95%CI = 1.15–2.14, p = 0.004; Table 5). However, as shown in Table 5, there was a non-significant relationship between CONUT score and sex (OR = 0.94, 95%CI = 0.72–1.21, p = 0.617), lymph node metastasis (OR = 1.64, 95%CI = 0.95–2.82, p = 0.074), microvascular invasion (OR = 1.22, 95%CI = 0.92–1.62, p = 0.158), perineural invasion (OR = 1.15, 95%CI = 0.74–1.80, p = 0.530), distant metastasis (OR = 1.15, 95%CI = 0.76–1.76, p = 0.505), TNM stage (OR = 1.23, 95%CI = 0.86–1.76, p = 0.251), and tumor number (OR = 1.02, 95%CI = 0.34–3.03, p = 0.977) of BTC cases.

Table 5. The Correlation between CONUT score and clinicopathological features in patients with BTC.

Factors	No. of studies	No. of patients	Effects model	OR (95%CI)	p	Heterogeneity I^2(%) Ph
Gender (male vs female)	6	1,076	Fixed	0.94(0.72–1.21)	0.617	28.3	0.223
LN metastasis (yes vs no)	6	1,076	Random	1.64(0.95–2.82)	0.074	65.7	0.012
Microvascular invasion (yes vs no)	5	982	Fixed	1.22(0.92–1.62)	0.158	45.7	0.118
Perineural invasion (yes vs no)	5	1,005	Random	1.15(0.74–1.80)	0.530	52.9	0.075
Differentiation (poor vs well/moderate)	4	781	Fixed	1.57(1.15–2.14)	0.004	31.1	0.226
Distant metastasis (yes vs no)	3	744	Fixed	1.15(0.76–1.76)	0.505	0	0.789
TNM stage (III–IV vs I–II)	3	536	Fixed	1.23(0.86–1.76)	0.251	12.9	0.317
Tumor number (multiple vs single)	2	238	Random	1.02(0.34–3.03)	0.977	54.6	0.138

LN: lymph node; BTC: biliary tract cancer; CONUT: controlling nutritional status.

Sensitivity analysis

A sensitivity analysis was performed to examine whether each study affected the pooled outcomes. As shown in Figure 4, after removing the included studies individually, the pooled HRs and 95%CIs for OS and RFS remained almost unchanged, indicating the reliability of our results.

Figure 4. Sensitivity analysis. (A) OS; (B) RFS.

Publication bias

Publication bias in the included studies was assessed using Begg’s test and funnel plots. Symmetric plots were observed in the funnel plots for OS and RFS (Figure 5). Begg’s test results (p = 0.097 and 0.080 for OS and RFS, respectively) indicated an absence of publication bias in the present study.

Figure 5. Publication bias test. (A) Begg’s test for OS, p = 0.097; (B) Begg’s test for RFS, p = 0.080.

Discussion

The prognostic effect of the CONUT score in patients with BTC has not been consistent in prior studies. This study extracted information from ten articles involving 1,441 patients [17–26] and found that a higher CONUT score predicted inferior OS and RFS in BTC patients. Moreover, the CONUT score had a stable prognostic effect among diverse subgroups. We also identified a significant connection between the CONUT score and poor differentiation in BTC, indicating that BTC patients with higher CONUT scores had more aggressive and malignant tumors. Therefore, an elevated CONUT score could be a strong predictor of tumor recurrence and poor long-term prognosis of BTC. To the best of our knowledge, this study is the first to analyze whether the pretreatment CONUT score is significant in predicting the prognosis of patients with BTC.

The CONUT score involves the serum albumin level, TC concentration, and lymphocyte count; therefore, the potential mechanisms of the prognostic effect of the CONUT score for BTC are interpreted from several aspects. First, serum albumin level is a critical predictor of nutritional status as a lower serum albumin level indicates protein loss due to systemic responses [33]. In cancer patients, hypoalbuminemia indicates malnutrition and cachexia. Hypoalbuminemia has been found to impair immune responses because nutrition is a crucial determinant [34]. Lymphocytes play major roles in cell-mediated immunity, cancer immune surveillance, and immune editing. As tumor suppressors, lymphocytes produce cytokines that inhibit the proliferation and metastatic activity of cancer cells and induce cell death [35]. Cholesterol is a major component of mammalian cell membranes [36]. Circulating blood is a source of cholesterol, which is composed of endocytosed low-density lipoprotein receptors. In patients with cancer, low cholesterol levels are associated with poor prognoses [37]. Thus, as a combined index of albumin level, TC content, and lymphocyte count, the CONUT score can be a valuable prognostic marker for BTC.

Significant heterogeneity was observed among the studies in the OS analysis (Table 3). Heterogeneity is derived from many variables, such as study type, sample size, study subjects, selection bias, and cutoff value. We selected a random-effects model based on the significant heterogeneity. The studies included in this meta-analysis remained comparable because they were selected based on uniform inclusion and exclusion criteria. The CONUT cutoff values were not uniform in the included studies because they used different methods to determine the cutoff values. This is a limitation of this meta-analysis, as shown below. Moreover, a subgroup analysis was conducted to evaluate the impact of the cutoff values on our results. The results showed that various cutoff values did not affect the prognostic value of CONUT for OS and RFS in patients with BTC (Tables 3 and 4).

It should also be noted that BTC is a heterogeneous disease that primarily comprises CCA and GBC. GBC is the most common BTC, comprising 80–90% of cases according to autopsy studies [38]. Previous studies have indicated that patients with eCCA, iCCA, and GBC have different prognoses [39]. The histological types used in the meta-analysis are listed in Table 2. We also conducted a subgroup analysis according to the histological subtype. As shown in Tables 3 and 4, CONUT score significantly predicted worse OS in patients with ECC, BTC, and GBC, but not in those with ICC. Moreover, a higher CONUT score significantly predicted RFS regardless of histology. This finding indicates that ICC patients with a high CONUT score may experience tumor recurrence, but the OS duration is not significantly different from those with a low CONUT score.

Notably, this meta-analysis focused on CONUT [12], a nutritional assessment tool based on serum albumin level, TC content, and lymphocyte count (Table 1). The total CONUT score is the sum of the three components. CONUT was established in 2005 [12] and the scoring system has been used until now. Table 1 is consistent with the original CONUT scoring system in the literature [40]. Compared to other nutritional tools and parameters, such as the Glasgow prognostic score (GPS), modified GPS [41,42], PNI [43,44], AGR [45,46], systemic immune inflammation index [47–50], GNRI [51–53], and CAR [40,54,55], CONUT has the following advantages. First, CONUT is a standardized scoring system with clear scoring criteria (Table 1). Therefore, promoting and disseminating CONUT scores is easier. Second, the CONUT score consists of three elements: serum albumin level, TC content, and lymphocyte count. These indicators comprehensively reflect nutritional status. Therefore, CONUT is more reliable and comprehensive. Third, CONUT scores ranged from 0 to 12. The CONUT score is a more refined and easily quantifiable indicator than other nutritional evaluation tools.

CONUT is a promising tool for guiding the management of patients with BTC. When a low pretreatment CONUT score is detected in an individual BTC patient, nutritional support should be provided. Moreover, the risk of tumor recurrence and poor survival and more aggressive therapeutic strategies should be considered. During follow-up, the CONUT score should also be monitored to facilitate the early detection of tumor progression.

Recently, many meta-analyses have reported that the CONUT score is significant for predicting the prognosis of solid tumors [56–61]. According to Peng et al. in a meta-analysis of 5,410 cases, a higher CONUT score indicated worse OS, RFS, disease-free survival (DFS), and cancer-specific survival (CSS) in upper urinary tract urothelial carcinoma and renal cell carcinoma [62]. Zhang et al. demonstrated that a high CONUT score had an unfavorable impact on OS, DFS, CSS, and progression-free survival (PFS) compared to low CONUT scores [57]. According to a recent meta-analysis involving 2,294 patients, a higher CONUT score predicted dismal OS in pancreatic cancer [59]. Takagi et al. conducted a meta-analysis of 2,601 cases and found that a higher CONUT score predicted dismal OS, CSS, and RFS in colorectal cancer patients undergoing surgery [61]. In a meta-analysis by Lu et al. that included 1,811 patients, a high CONUT score predicted worse OS and PFS in patients with hematological cancer [63].

This study had some limitations. First, all included studies were conducted in Asia, which may make them less applicable to other populations when applied on a large scale. Second, all the studies had a retrospective design, which may have reduced the power of our analysis. Third, the threshold for defining the high CONUT score group was not uniform among the studies, which could have introduced a selection bias. Fourth, most included studies (9 out of 10) enrolled patients with BTC who underwent surgical resection. Therefore, our results may be more applicable to BTC patients undergoing surgery. Therefore, prospective studies with larger sample sizes are needed to validate our results.

In summary, a high CONUT score significantly predicted worse OS and RFS in patients with BTC. Additionally, BTC patients with higher CONUT scores tended to have poor tumor differentiation. The CONUT score could help clinicians stratify high-risk BTC patients and devise individualized treatment plans.

Authors’ contributions

HJ and ZW design the project; HZ and ZW searched databases and performed literature screen; HZ and ZW extracted and analyzed the data, analysis; HZ and ZW evaluated the quality of included literature; HZ contributed to writing the manuscript. Final draft was approved by all the authors. All authors contributed to the article and approved the submitted version.

Abbreviations
CONUT	=	controlling nutritional status
BTC	=	biliary tract cancer
HR	=	hazard ratio
CI	=	confidence interval
OS	=	overall survival
RFS	=	recurrence-free survival
GBC	=	gallbladder cancer
ECC	=	extrahepatic cholangiocarcinoma
ICC	=	intrahepatic cholangiocarcinoma
HCC	=	hepatocellular carcinoma
PNI	=	prognostic nutritional index
AGR	=	albumin to globulin ratio
GNRI	=	geriatric nutritional risk index
CAR	=	C-reactive protein to albumin ratio
TC	=	total cholesterol
BC	=	breast cancer
PRISMA	=	Preferred Reporting Items for Systematic Reviews and Meta-Analyses
MeSH	=	Medical Subject Headings
NOS	=	Newcastle-Ottawa Scale
OR	=	odds ratio
PTBS	=	percutaneous transhepatic biliary stenting
OS	=	overall survival
RFS	=	recurrence-free survival
LNM	=	lymph node metastasis
DFS	=	disease-free survival
CSS	=	cancer-specific survival
PFS	=	progression-free survival.

Acknowledgments

We would like to thank Editage (www.editage.com) for English language editing.

Disclosure statement

The authors declare that there is no conflict of interest.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.