https://orcid.org/0000-0002-8488-7539
Figures
Abstract
Background
To evaluate the clinical value of Aldo-keto reductase family 1 member B10 (AKR1B10) in the diagnosis and prognosis of hepatocellular carcinoma (HCC).
Methods
A search of the PubMed, China Biology Medicine, Cochrane, and Embase databases was performed to conduct meta-analyses to evaluate the accuracy of AKR1B10 in diagnosing HCC and to assess the impact on prognosis of patients after curative resection of HCC.
Results
A total of 12 different cohorts from 11 studies including 2747 HCC patients and 2053 controls showed that the pooled specificity and the pooled sensitivity of AKR1B10 for the diagnosis of HCC were 0.78 (95% CI: 0.69–0.85) and 0.85 (95% CI: 0.77–0.90), respectively. The pooled sensitivity and specificity of serum AKR1B10 for the diagnosis of HCC were 0.80 (95% CI: 0.70–0.86) and 0.87 (95% CI: 0.77–0.93), respectively. The pooled sensitivity and specificity of AKR1B10 in malignant tumor tissue for the diagnosis of HCC were 0.78 (95% CI: 0.61–0.89) and 0.82 (95% CI: 0.69–0.90), respectively. The pooled sensitivity and specificity of AKR1B10 to distinguish HCC from benign liver disease were 0.71 (95% CI: 0.62–0.78) and 0.84 (95% CI: 0.77–0.89), respectively. The sensitivity and specificity of AKR1B10 combined with alpha fetoprotein (AFP) in the diagnosis of HCC were 0.84 (95% CI: 0.79–0.88) and 0.88 (95% CI: 0.73–0.95), respectively. The pooled sensitivity and specificity of AKR1B10 in malignant tumor tissue for the diagnosis of early-stage HCC were 0.85 (95% CI: 0.62–0.95) and 0.88 (95% CI: 0.81–0.93), respectively. A meta-analysis of five studies including 798 patients demonstrated that high AKR1B10 expression in liver malignant tumor was associated with better overall survival in patients with HCC after hepatectomy (HR = 0.54, 95% CI: 0.41–0.72, p < 0.001).
Citation: Wang Z, Pei Y, Li W, Zhang J, Liu J (2022) Clinical value of AKR1B10 in hepatocellular carcinoma: A systematic review and meta-analysis. PLoS ONE 17(12): e0279591. https://doi.org/10.1371/journal.pone.0279591
Editor: Muhammad Tarek Abdel Ghafar, Tanta University Faculty of Medicine, EGYPT
Received: April 26, 2022; Accepted: December 10, 2022; Published: December 30, 2022
Copyright: © 2022 Wang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Hepatocellular carcinoma (HCC) is one of the most common malignant tumors worldwide [1–3]. Since most tumors are already in the advanced stage at the time of discovery, the 5-year overall survival (OS) rate of HCC is only 5–9% [4, 5]. However, it has been reported that the 5-year survival rate of patients can be improved to more than 50% if HCC can be diagnosed at an early stage [6]. The alpha-fetoprotein (AFP) is a marker for routine screening of HCC, but it is not stable for the detection of HCC. It has been reported that the sensitivity of AFP for diagnosing HCC ranges 63%–68% and the specificity ranges 76%–94% [7, 8]. Furthermore, AFP cannot satisfy the diagnosis of early HCC, Galle et al. found that 31% of early HCC patients had serum AFP levels below 20 ng / ml [9]. AFP expression levels were also found to be upregulated in patients with benign liver diseases, such as cirrhosis and hepatitis [10]. The limited diagnostic value of APF for HCC makes it urgent to find an excellent diagnostic biomarker for HCC. Currently, few studies have been performed on the predictive biomarkers after treatment of patients with HCC [9]. Therefore, many researchers have focused on identifying better diagnostic markers and prognostic markers for HCC.
Aldo-keto reductase family 1 member B10 (AKR1B10), originally identified as a human small intestinal aldose reductase and aldose reductase-like protein, is a cytosolic NADPH-dependent reductase that catabolizes a variety of endogenous compounds, such as dicarbonyls, aromatic, aliphatic aldehydes, and certain pharmaceutical ketones. The AKR1B10 gene is located on chromosome 7q33, and the AKR1B10 protein consists of 316 amino acids [11]. Recently, the enzyme was found to be highly expressed in several malignant tumors, including liver and lung tumors [12, 13]. As a result, AKR1B10 has attracted extensive attention as a relevant biomarker for the development of these tumors. In recent years, the rapid development of molecular testing technology has made it possible to detect AKR1B10 in peripheral blood samples, which allows analysis and provides clinicians with information on tumor biology. However, the diagnostic and prognostic significance of AKR1B10 in HCC remains unclear.
In the current study, we aimed to evaluate the diagnostic value of AKR1B10 in serum and tumor tissue for HCC, its ability to distinguish HCC from benign liver disease, and its accuracy in diagnosing early-stage HCC. Furthermore, we aimed to explore the impact of AKR1B10 expression on the OS and recurrence-free survival (RFS) of HCC patients after hepatectomy.
Methods
Data sources and searches
Literature search was performed on PubMed, China Biology Medicine, Cochrane, and Embase databases to identify studies reporting the diagnostic and prognostic value of AKR1B10 in patients with HCC. The deadline for retrieval was April 2, 2022. The search terms included in the search strategy were as follows: ((AKR1B10) OR (Aldo-keto reductase1-B10)) AND ((liver cancer) OR (hepatoma) OR (hepatic carcinoma) OR (hepatocellular carcinoma)).
Screening and selection of studies
All studies were individually screened by the title and abstract, followed by a full-text assessment by two reviewers. Divergences were solved by discussion until consensus was reached. The inclusion criteria were as follows: (1) research population included patients with HCC and non-HCC (benign liver disease or normal person); (2) AKR1B10 was analyzed in all patient groups; (3) prognostic studies needed to provide sufficient data to calculate the HR and its 95% CI; (4) diagnostic studies needed to provide sufficient data to calculate sensitivity and specificity; and (5) there were no language restrictions. The exclusion criteria were as follows: (1) insufficient data provided to calculate the HR; (2) insufficient data provided to calculate the sensitivity and specificity; (3) abstracts, reviews, letters, case reports, or publications; or (4) animal or cell experiments.
Data extraction and outcomes
The data were individually extracted by two authors, following a predefined scheme. Divergences were solved by discussion until consensus was reached. Extracted study characteristics included the following: first author, publication year, type of cancer, number of patients, number of controls, type of sampling, aim of the study, and detection method of the AKR1B10 expression level. On the basis of the aim and outcome of the included studies, studies were classified into the diagnostic and prognostic groups. Studies were classified into the diagnostic group when AKR1B10 was used as a biomarker to diagnose HCC or to differentiate HCC from healthy controls. Studies were categorized into the prognostic group when the association between the high or low expression level of AKR1B10 and survival was provided.
Study quality assessment
The quality of included studies was scored separately by two reviewers. Divergences were solved by discussion until consensus was reached. The quality assessment of diagnostic accuracy studies (QUADAS) criteria was used to assess the quality of studies related to the diagnosis of HCC by AKR1B10. The Newcastle Ottawa scale was used to assess the quality of studies related to the prognosis of HCC.
Statistical analysis
All analyses were conducted using Stata 16 and Meta-Disc 1.4 software. For diagnostic studies, the pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and area under the curve (AUC) and their 95% confidence intervals (CIs) were calculated to assess the diagnostic value of AKR1B10 for HCC. Furthermore, the Spearman correlation analysis method was used to examine threshold effect heterogeneity among diagnostic studies. For prognostic studies, the hazard ratio (HR) and its 95% CI were collected to assess the correlation between AKR1B10 expression and prognosis of HCC. If the study only provided Kaplan Meier survival curves, the Engage Digitizer version 2.11 software was used to extract the relevant values from the survival curves and calculate the HR (95% CI) [14, 15]. The test of inconsistency (I2) was used to evaluate the degree of inter-study statistical heterogeneity. Substantial heterogeneity was indicated if I2 values were ≥ 50%, and the random-effects model was adopted [16, 17]. The Deeks’ funnel plot asymmetry test was used to assess the publication bias of the included studies, and P < 0.05 was considered statistically significant [18].
Results
Characteristics of the included studies
The flowsheet of the study screening process is shown in Fig 1. Database searches yielded 731 studies. A total of 187 studies were found after removing duplicates. Upon reading the abstracts, 123 studies unrelated to the research question were excluded. The remaining 64 articles were fully reviewed, and 48 articles were removed (20 articles did not provide sufficient data, eight were reviews, six were case reports, two were meta-analyses, and 12 were animal or cell experiments). Overall, 16 studies, including 11 diagnostic [19–29] and five prognostic [30–34] studies, were included in our meta-analysis.
Analysis of threshold effect.
To verify whether there was heterogeneity caused by the AKR1B10 threshold among the included diagnostic studies, we used the Meta-Disc 1.4 software to perform the Spearman test to detect the threshold effect, and the correlation coefficient and P value between the sensitivity logit and the 1—specificity logit were calculated. The results showed that the Spearman correlation coefficient and P value were -0.413 and 0.183 (> 0.05), respectively, indicating that the heterogeneity caused by different AKR1B10 thresholds did not exist in the 11 diagnostic studies.
Accuracy of AKR1B10 for diagnosing HCC
A total of 11 studies [19–29] including 12 different cohorts reported data evaluating the diagnostic accuracy of AKR1B10 for HCC, and they were published between 2014 and 2020. Data from 2747 HCC patients and 2053 controls were collected. All diagnoses of HCC patients were verified by histopathology. Histopathology, ultrasound (US), computed tomography (CT), magnetic resonance imaging (MRI) or hematology were used to diagnose all patients with benign liver disease. The Barcelona Clinic Liver Cancer (BCLC) and tumor node metastasis (TNM) classifications were used to define HCC stages in the included studies. Enzyme linked immunosorbent assay (ELISA), Time-resolved fluoroimmunoassay (TRFIA), Reverse transcription polymerase chain reaction (PT-PCR) and Immunohistochemistry was used to determine the expression level of AKR1B10 in serum and tumor tissue. Detailed characteristics of the included studies are shown in Table 1. Fig 2 shows the details of the quality assessment form.
The pooled sensitivity and specificity of AKR1B10 in discriminating HCC from non-HCC were 0.78 (95% CI: 0.69–0.85) and 0.85 (95% CI: 0.77–0.90), respectively. The PLR, NLR, DOR, and AUC with their corresponding 95% CIs were 5.3 (95% CI: 3.3–8.5), 0.26 (95% CI: 0.17–0.38), 21.00 (95% CI: 9.00–46.00), and 0.89 (95% CI: 0.86–0.91), respectively (Fig 3A–3C). Our meta-analysis suggested that AKR1B10 has good diagnostic efficacy for HCC. However, we found that each article had a different source of AKR1B10. In addition, benign liver disease was used as a control group in five studies [23, 25–28]. Therefore, we performed subgroup analyses to further evaluate the accuracy of AKR1B10 as a diagnostic marker for HCC.
Accuracy of AKR1B10 derived from different samples and detection methods for the diagnosis of HCC.
The tested samples of AKR1B10 included serum and malignant tumor tissue. To verify whether different sources of AKR1B10 have different diagnostic accuracies for HCC, we analyzed the diagnostic accuracy of AKR1B10 levels for HCC in two samples. A total of five articles [19–23] including 6 different cohorts reported the diagnostic accuracy of AKR1B10 in serum for HCC. As shown in Table 2, the pooled sensitivity and specificity of serum AKR1B10 for the diagnosis of HCC were 0.80 (95% CI: 0.70–0.86) and 0.87 (95% CI: 0.77–0.93), respectively. The PLR, NLR, DOR, and AUC with their corresponding 95% CIs were 6.1 (95% CI: 3.5–10.7), 0.24 (95% CI: 0.16–0.34), 26.00 (95% CI: 14.00–49.00), and 0.89 (95% CI: 0.86–0.92), respectively. In addition, a total of six studies [24–29] reported the diagnostic accuracy of AKR1B10 in the malignant tumor tissue for HCC. The pooled sensitivity and specificity of AKR1B10 in malignant tumor tissues for the diagnosis of HCC were 0.78 (95% CI: 0.61–0.89) and 0.82 (95% CI: 0.69–0.90), respectively. The PLR, NLR, DOR, and AUC with their corresponding 95% CIs were 4.3 (95% CI: 2.1–8.9), 0.27 (95% CI: 0.13–0.57), 16.00 (95% CI: 4.00–69.00), and 0.87 (95% CI: 0.84–0.90), respectively (Table 2). Pooled analysis showed that AKR1B10 in the malignant tumor tissue had lower pooled sensitivity, specificity, and AUC for diagnosing HCC compared with AKR1B10 in serum.
To verify whether AKR1B10 levels detected by different detection method have different diagnostic accuracy for HCC, a subgroup analysis was performed. ELISA was used to measure AKR1B10 expression levels in serum in four studies [19, 20, 22, 23], and the pooled sensitivity and specificity for diagnosing HCC using AKR1B10 expression levels measured by ELISA were 0.80 (95% CI: 0.69–0.90) and 0.85 (95% CI: 0.73–0.94), respectively (Fig 4). Three studies [24, 26, 28] used Immunohistochemistry as detection method of AKR1B10 to diagnose HCC, and the pooled sensitivity and specificity were 0.81 (95% CI: 0.54–0.98) and 0.89 (95% CI: 0.72–0.99), respectively (Fig 4). RT-PCR was used to measure AKR1B10 expression levels to diagnose HCC in three studies [25, 27, 29], and the pooled sensitivity and specificity were 0.68 (95% CI: 0.54–0.81) and 0.76 (95% CI: 0.67–0.83), respectively (Fig 4).
Accuracy of AKR1B10 in distinguishing HCC from benign liver diseases
To assess the accuracy of AKR1B10 in distinguishing HCC from benign liver diseases, such as chronic hepatitis and cirrhosis, we collected and analyzed the relevant data from five studies [23, 25–28] including six different cohorts. The pooled sensitivity and specificity of AKR1B10 in distinguishing HCC from benign liver disease were 0.71 (95% CI: 0.62–0.78) and 0.84 (95% CI: 0.77–0.89), respectively. The PLR, NLR, DOR, and AUC with their corresponding 95% CIs were 4.3 (95% CI: 2.9–5.6), 0.35 (95% CI: 0.25–0.47), 13.00 (95% CI: 6.00–25.00), and 0.85 (95% CI: 0.82–0.88), respectively (Table 2). Therefore, we conclude that AKR1B10 has good ability to distinguish HCC from benign liver disease.
Accuracy of AKR1B10 combined with AFP for the diagnosis of HCC
AFP is a traditional diagnostic marker for HCC. To evaluate the diagnostic accuracy of AKR1B10 combined with AFP for HCC, a total of five articles [19–23] including six different cohorts were collected and analyzed. Patients could be diagnosed with HCC if they were either AFP positive or AKR1B10 positive. The pooled sensitivity and specificity of AKR1B10 combined with AFP in discriminating HCC from non-HCC were 0.84 (95% CI: 0.79–0.88) and 0.88 (95% CI: 0.73–0.95), respectively. The PLR, NLR, DOR, and AUC with their corresponding 95% CIs were 7.1 (95% CI: 3.0–16.8), 0.18 (95% CI: 0.14–0.22), 40.00 (95% CI: 18.00–86.00), and 0.90 (95% CI: 0.87–0.92), respectively (Table 2).
Accuracy of AKR1B10 in the diagnosis of early HCC
A total of three studies [22–24] including four different cohorts provided relevant data for the diagnosis of early HCC by AKR1B10. HCC at stages 0 and A was defined as early HCC when the BCLC classification was used [22, 23]. When the TNM classification was used [24], HCC at stages I-II was defined as early HCC. The sensitivity and specificity of AKR1B10 for diagnosing early-stage HCC were 0.85 (95% CI: 0.62–0.95) and 0.88 (95% CI: 0.81–0.93), respectively. The PLR, NLR, DOR, and AUC with their corresponding 95% CIs were 7.3 (95% CI: 4.3–12.4), 0.17 (95% CI: 0.06–0.49), 42.00 (95% CI: 11.00–160.00), and 0.92 (95% CI: 0.90–0.94), respectively (Table 2). Our meta-analysis suggests that AKR1B10 has an excellent diagnostic efficacy for early HCC.
Clinical value of AKR1B10 for the prognosis of HCC
A total of five studies [30–34] including 798 patients reported the data on the ability of the AKR1B10 expression level to predict the OS of patients after hepatectomy. All HCC patients included in the study had undergone hepatectomy, and AKR1B10 expression levels in tissues of liver malignancy were measured by Immunohistochemistry or RT-PCR. HCC patients with high expression of AKR1B10 were divided into the HCC group and HCC patients with low expression were divided into the control group. We extracted the hazard ratio (HR) and standard error for OS from each study. The HR in Schmitz’s study [31] was derived from the K-M curve, other HRs were obtained directly from the corresponding original articles. If HR was < 1, it indicated that the OS of patients after hepatectomy with AKR1B10 high expression was better than that after hepatectomy with AKR1B10 low expression. As shown in Fig 5, our study suggests that high AKR1B10 expression in malignant tumor tissue was associated with better OS in patients with HCC after hepatectomy (HR = 0.54, 95% CI: 0.41–0.72, p < 0.001) In addition, Sonohara [30] and Ha [33] showed that high AKR1B10 levels are associated with better RFS in HCC patients after hepatic resection. These results suggest that high AKR1B10 expression may be a marker of good prognosis in HCC patients after hepatic resection.
Publication bias
To assess the publication bias of included studies, the linear regression test of Deeks’ funnel plot asymmetry was conducted. The pooled Deeks’ test result of the diagnostic studies was p = 0.55 (Fig 3D), which indicated that there was no obvious publication bias for our diagnostic analysis.
Discussion
In HCC cells, inhibition of AKR1B10 expression resulted in cell cycle arrest and impaired cell proliferation, suggesting that AKR1B10 may exert tumorigenic effects through enhanced cell growth [35]. Upregulated AKR1B10 expression levels have been proven to be a risk factor for HCC [25, 27]. Furthermore, experimental results in vitro and in vivo demonstrated that AKR1B10 is involved in cancer cell proliferation [36, 37]. Therefore, we had a reason to believe that AKR1B10 plays an important role in the course of HCC, and we performed meta-analyses to evaluate the clinical value of AKR1B10 in HCC.
In this meta-analysis, after excluding heterogeneity due to threshold effects, we showed that the overall sensitivity and specificity of AKR1B10 for the diagnosis of HCC were 78% and 85%, respectively. Subsequently, we analyzed the main sources of heterogeneity and divided the included diagnostic studies into six subgroups. Our study showed that the sensitivity and specificity of serum and tissue as test specimens for AKR1B10 in the diagnosis of HCC were similar, which indicates that the different tested specimens were not the main source of heterogeneity. In addition, the control population was not consistent among the included studies, which may have affected the accuracy of AKR1B10 for diagnosing HCC.
With the development of science and technology, simultaneous determination of AKR1B10 and AFP is feasible. Our meta-analysis table indicates a higher sensitivity and specificity of AKR1B10 combined with AFP for the diagnosis of HCC compared with AKR1B10 or AFP alone. In addition, serum AFP levels have been reported to be not only associated with HCC but also to be elevated in patients with benign liver diseases, such as hepatitis and cirrhosis [10, 38]. Our meta-analysis indicated that AKR1B10 as a tumor marker has the ability to distinguish HCC from benign liver disease.
It has been reported that the 5-year survival rate of patients can be improved to more than 50% if HCC can be detected at an early stage [6]. However, Marrero’s study [39] showed that the sensitivity and specificity of AFP for the diagnosis of early HCC were 66% and 82%, respectively. Our meta-analysis indicated that the sensitivity of AKR1B10 for the diagnosis of early-stage HCC is higher than that of AFP. This may be related to the opposite expression pattern of AKR1B10 compared with that of AFP [19]. It has been reported that the unique expression pattern of AKR1B10 may be because of 14-3-3 ε induced AKR1B10 upregulation at the early stage of HCC, leading to decreased retinoic acid levels and subsequent cell proliferation. In advanced HCC, the relationship between 14-3-3 ε and AKR1B10 is inhibited by unknown factors, leading to decreased AKR1B10 levels [40].
The main difficulty in treating HCC is the high frequency of tumor recurrence even after liver transplantation and radical resection [41], which cannot be avoided even in well-differentiated tumors [42]. In the present meta-analysis, summary results without significant heterogeneity showed that high expression of AKR1B10 in malignant tumor tissue was significantly correlated with better OS and RFS in HCC patients after hepatectomy. Liu’s study [43] suggested that this may be because of the complex regulation of expression at the time of AKR1B10 transcription. In the AKR1B10 promoter, multiple putative binding sites for oncogenic and tumor suppressor proteins are recognized, including p53, cets-1, and C/EBP. In more malignant HCC cells, the deregulation or nonspecific switching on/off of certain transcriptional machinery may downregulate AKR1B10 expression.
To the best of our knowledge, no meta-analysis has yet been reported on the association between AKR1B10 expression levels and the diagnosis and prognosis of HCC. However, our study has some limitations. First, patients from different included studies had different characteristics, which are common in the evaluation of the accuracy of diagnostic tests. It may have affected the accuracy of diagnostic analysis. Second, the cut-off for AKR1B10 positivity was not uniform in the literature included in this study, which may have affected the accuracy of AKR1B10 for HCC diagnosis. Finally, most of the studies on AKR1B10 for diagnosing HCC were conducted in Asian populations, which may have resulted in a population selection bias.
Conclusions
In conclusion, our meta-analysis suggests that AKR1B10 has a high diagnostic value for HCC and can be used as a novel screening method. Notably, AKR1B10 has excellent diagnostic efficacy for early-stage HCC and shows good performance when combined with AFP to diagnose HCC. In addition, AKR1B10 can predict OS and RFS of HCC patients after hepatectomy. However, more large cohort studies or large randomized controlled multicenter trials are needed to further explore the clinical value of AKR1B10 for patients with HCC.
Acknowledgments
We thank LetPub (www.letpub.com) for its linguistic assistance during the preparation of this manuscript.
References
- 1. Yang S, Liu L, Luo Y, Chen M, Li N, Xu L, et al. (2018). Lower alpha fetoprotein and higher risk of hepatocellular carcinoma, study from the type 2 diabetes mellitus patients. Diabetes Res Clin Pract 143:239–44. pmid:30009939
- 2. Wang ZF, Hu R, Pang JM, Zhang GZ, Yan W, Li ZN. (2018). Serum long noncoding RNA LRB1 as a potential biomarker for predicting the diagnosis and prognosis of human hepatocellular carcinoma. Oncol Lett 16:1593–601. pmid:30008842
- 3. Uthamalingam P, Das A, Behra A, Kalra N, Chawla Y. (2018). Diagnostic Value of Glypican3, Heat Shock Protein 70 and Glutamine Synthetase in Hepatocellular Carcinoma Arising in Cirrhotic and Non-Cirrhotic Livers. J Clin Exp Hepatol 8:173–80. pmid:29892181
- 4. Siegel R, Naishadham D, Jemal A. (2012). Cancer statistics, 2012. CA Cancer J Clin 62:10–29. pmid:22237781
- 5. Schwartz M, Roayaie S, Konstadoulakis M. (2007). Strategies for the management of hepatocellular carcinoma. Nat Clin Pract Oncol 4:424–32. pmid:17597707
- 6. Forner A, Llovet JM, Bruix J. (2012). Hepatocellular carcinoma. Lancet 379:1245–55. pmid:22353262
- 7. He S, Hu XW, Wang D, Han LF, Zhang DC, Wei C. (2016). Accuracy of microRNAs for the diagnosis of hepatocellular carcinoma: A systematic review and meta-analysis. Clin Res Hepatol Gastroenterol 40:405–17. pmid:27016891
- 8. Jin X, Cai C, Qiu Y. (2019). Diagnostic Value of Circulating microRNAs in Hepatitis B Virus-Related Hepatocellular Carcinoma: A Systematic Review and Meta-Analysis. J Cancer 10:4754–64. pmid:31598147
- 9. Galle PR, Foerster F, Kudo M, Chan SL, Llovet JM, Qin S, et al. (2019). Biology and significance of alpha-fetoprotein in hepatocellular carcinoma. Liver Int 39:2214–29. pmid:31436873
- 10. Collier J, Sherman M. (1998). Screening for hepatocellular carcinoma. Hepatology 27:273–8. pmid:9425947
- 11. Penning TM. (2015). The aldo-keto reductases (AKRs): Overview. Chem Biol Interact 234:236–46. pmid:25304492
- 12. Bruix J, Sherman M, American Association for the Study of Liver D. (2011). Management of hepatocellular carcinoma: an update. Hepatology 53:1020–2. pmid:21374666
- 13. Matsunaga T, Wada Y, Endo S, Soda M, El-Kabbani O, Hara A. (2012). Aldo-Keto Reductase 1B10 and Its Role in Proliferation Capacity of Drug-Resistant Cancers. Front Pharmacol 3:5. pmid:22319498
- 14. Tierney JF, Stewart LA, Ghersi D, Burdett S, Sydes MR. (2007). Practical methods for incorporating summary time-to-event data into meta-analysis. Trials 8:16. pmid:17555582
- 15. Parmar MK, Torri V, Stewart L. (1998). Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints. Stat Med 17:2815–34. pmid:9921604
- 16. Higgins JP, Thompson SG, Deeks JJ, Altman DG. (2003). Measuring inconsistency in meta-analyses. BMJ 327:557–60. pmid:12958120
- 17. Jackson D, White IR, Thompson SG. (2010). Extending DerSimonian and Laird’s methodology to perform multivariate random effects meta-analyses. Stat Med 29:1282–97. pmid:19408255
- 18. Deeks JJ, Macaskill P, Irwig L. (2005). The performance of tests of publication bias and other sample size effects in systematic reviews of diagnostic test accuracy was assessed. J Clin Epidemiol 58:882–93. pmid:16085191
- 19. Han C, Gao L, Bai H, Dou X. (2018). Identification of a role for serum aldo-keto reductase family 1 member B10 in early detection of hepatocellular carcinoma. Oncol Lett 16:7123–30. pmid:30546447
- 20. Du Z, Liu X, Wei X, Luo H, Li P, Shi M, et al. (2020). Quantitative proteomics identifies a plasma multi-protein model for detection of hepatocellular carcinoma. Sci Rep 10:15552. pmid:32968147
- 21. Zhu R, Xiao J, Luo D, Dong M, Sun T, Jin J. (2019). Serum AKR1B10 predicts the risk of hepatocellular carcinoma—A retrospective single-center study. Gastroenterol Hepatol 42:614–21. pmid:31495535
- 22. Wang Y, Xu W, Chai X. (2020). Application of combined serum AFP and AKR1B10 in diagnosis of hepatocellular carcinoma. Chin J Integr Tradit West Med Liver Dis 30:541–3.
- 23. Ye X, Li C, Zu X, Lin M, Liu Q, Liu J, et al. (2019). A Large-Scale Multicenter Study Validates Aldo-Keto Reductase Family 1 Member B10 as a Prevalent Serum Marker for Detection of Hepatocellular Carcinoma. Hepatology 69:2489–501. pmid:30672601
- 24. Matkowskyj KA, Bai H, Liao J, Zhang W, Li H, Rao S, et al. (2014). Aldoketoreductase family 1B10 (AKR1B10) as a biomarker to distinguish hepatocellular carcinoma from benign liver lesions. Hum Pathol 45:834–43. pmid:24656094
- 25. Mori M, Genda T, Ichida T, Murata A, Kamei M, Tsuzura H, et al. (2017). Aldo-keto reductase family 1 member B10 is associated with hepatitis B virus-related hepatocellular carcinoma risk. Hepatol Res 47:E85–E93. pmid:27084455
- 26. Han C, Gao L, Zhao L, Sheng Q, Zhang C, An Z, et al. (2018). Immunohistochemistry Detects Increased Expression of Aldo-Keto Reductase Family 1 Member B10 (AKR1B10) in Early-Stage Hepatocellular Carcinoma. Med Sci Monit 24:7414–23. pmid:30328412
- 27. Sato S, Genda T, Ichida T, Murata A, Tsuzura H, Narita Y, et al. (2016). Impact of aldo-keto reductase family 1 member B10 on the risk of hepatitis C virus-related hepatocellular carcinoma. J Gastroenterol Hepatol 31:1315–22. pmid:26758591
- 28. Zhao J, Zhang W, Liu A. (2014). Expression and its clinical significance of AKR1B10 protein in hepatocellular carcinoma. Chin Mod Doc 52:49–51.
- 29. Wang Y. (2016). Expression and clinical significance of AKR1B10 in hepatocellular carcinoma. Gua Med 2013:2021–4.
- 30. Sonohara F, Inokawa Y, Hishida M, Kanda M, Nishikawa Y, Yamada S, et al. (2016). Prognostic significance of AKR1B10 gene expression in hepatocellular carcinoma and surrounding non-tumorous liver tissue. Oncol Lett 12:4821–8. pmid:28105190
- 31. Schmitz KJ, Sotiropoulos GC, Baba HA, Schmid KW, Müller D, Paul A, et al. (2011). AKR1B10 expression is associated with less aggressive hepatocellular carcinoma: a clinicopathological study of 168 cases. Liver Int 31:810–6. pmid:21645211
- 32. Wang YY, Qi LN, Zhong JH, Qin HG, Ye JZ, Lu SD, et al. (2017). High expression of AKR1B10 predicts low risk of early tumor recurrence in patients with hepatitis B virus-related hepatocellular carcinoma. Sci Rep 7:42199. pmid:28181486
- 33. Ha SY, Song DH, Lee JJ, Lee HW, Cho SY, Park CK. (2014). High expression of aldo-keto reductase 1B10 is an independent predictor of favorable prognosis in patients with hepatocellular carcinoma. Gut liver 8:648–54. pmid:25287169
- 34. Li Z. (2018). Clinical significance of AKR1B10 expression in hepatocellular carcinoma and correlation analysis of clinicopathological features. Chin Med 15:1077.
- 35. Wang J, Zhou Y, Fei X, Chen X, Chen Y. (2018). Biostatistics mining associated method identifies AKR1B10 enhancing hepatocellular carcinoma cell growth and degenerated by miR-383-5p. Sci Rep 8:11094. pmid:30038373
- 36. Yan R, Zu X, Ma J, Liu Z, Adeyanju M, Cao D. (2007). Aldo-keto reductase family 1 B10 gene silencing results in growth inhibition of colorectal cancer cells: Implication for cancer intervention. Int J Cancer 121:2301–6. pmid:17597105
- 37. Satow R, Shitashige M, Kanai Y, Takeshita F, Ojima H, Jigami T, et al. (2010). Combined functional genome survey of therapeutic targets for hepatocellular carcinoma. Clin Cancer Res 16:2518–28. pmid:20388846
- 38. Oka H, Tamori A, Kuroki T, Kobayashi K, Yamamoto S. (1994). Prospective study of alpha-fetoprotein in cirrhotic patients monitored for development of hepatocellular carcinoma. Hepatology 19:61–6. pmid:7506227
- 39. Marrero JA, Feng Z, Wang Y, Nguyen MH, Befeler AS, Roberts LR, et al. (2009). Alpha-fetoprotein, des-gamma carboxyprothrombin, and lectin-bound alpha-fetoprotein in early hepatocellular carcinoma. Gastroenterology 137:110–8. pmid:19362088
- 40. Liu TA, Jan YJ, Ko BS, Wu YJ, Lu YJ, Liang SM, et al. (2015). Regulation of aldo-keto-reductase family 1 B10 by 14-3-3ε and their prognostic impact of hepatocellular carcinoma. Oncotarget 6:38967–82. pmid:26516929
- 41. Wang Z, Zhang G, Wu J, Jia M. (2013). Adjuvant therapy for hepatocellular carcinoma: current situation and prospect. Drug Discov Ther 7:137–43. pmid:24071575
- 42. Takenaka K, Adachi E, Nishizaki T, Hiroshige K, Ikeda T, Tsuneyoshi M, et al. (1994). Possible multicentric occurrence of hepatocellular carcinoma: a clinicopathological study. Hepatology 19:889–94. pmid:8138262
- 43. Liu Z, Zhong L, Krishack PA, Robbins S, Cao JX, Zhao Y, et al. (2009). Structure and promoter characterization of aldo-keto reductase family 1 B10 gene. Gene 437:39–44. pmid:19236911