Prognostic Role of Serum Lactate Dehydrogenase in Patients With Urothelial Carcinoma: A Systematic Review and Meta-Analysis

Background: To investigate the potential prognostic role of serum lactate dehydrogenase (LDH) in patients with urothelial carcinoma (UC) using the method of systematic review and meta-analysis. Materials and Methods: We searched PubMed, Embase, Cochrane Library, and Web of Science for eligible studies up to February 2020. Pooled hazard ratios (HRs) and 95% confidence intervals (CIs) were used to estimate the relationship. Results: A total of 14 studies including 4,009 patients with UC were incorporated. The results showed that a high pretreatment serum LDH was associated with an inferior overall survival (OS, HR 1.61, 95% CI 1.39–1.87, p < 0.001), cancer-specific survival (CSS, HR 1.41, 95% CI 1.05–1.90, p = 0.022), and disease-free survival (DFS, HR 1.64, 95% CI 1.04–2.59, p = 0.034) in UC. Subgroup analyses identified that a high pretreatment serum LDH was associated with a poor OS (HR 1.97, 95% CI 1.02–3.81, p = 0.042) and DFS (HR 1.64, 95% CI 1.04–2.59, p = 0.034) in upper tract urothelial carcinoma, a short OS (HR 1.71, 95% CI 1.37–2.15, p < 0.001) in urothelial carcinoma of bladder. Conclusion: Our findings indicated that a high level of pretreatment serum LDH was associated with inferior OS, CSS, and DFS in patients with UC. This biomarker can be an important factor incorporated into the prognostic models for UC.


INTRODUCTION
Urothelial carcinoma (UC), mainly consisting of upper tract urothelial carcinoma (UTUC) and urothelial carcinoma of the bladder (UCB), is, respectively estimated to have 85,000 new incidences and 18,000 related mortality in the United States in 2018. These made UC become the 4th and 12th most common malignance for males and females (1). Multiple lesions, high rate of recurrences, and distant metastasis are typical features of UC. Despite advances in surgical techniques and developments of preoperative and postoperative adjuvant therapy, the long-term survival of cases with UC have not significantly changed over these years (2). Especially, for patients with advanced UC, median overall survival was only 3 to 6 months without therapy, and prolonged to 13-16 months when receiving systematic chemotherapy (3). Therefore, it is important to determine prognostic factors for timely adjustment of treatment.
Previous literatures have found that cancer cells metabolize differently from normal cells, which means more lactate seems to be needed for cancer cells (4). Lactate dehydrogenase (LDH), an enzyme that catalyzes lactic acid into pyruvate, may exert a crucial role in the metabolism of tumor cells. High level of serum LDH has been reported to serve as an unfavorable prognostic factor in kinds of malignances, including prostate cancer, renal cell carcinoma, lymphoma, colorectal cancer, and lung cancer (5). Moreover, many studies have identified the prognostic role of pretreatment serum LDH in cases with UC. Because circulating blood LDH is easy to be measured clinically, it can be used as an indicator of cancer burden and a useful biomarker in clinical management. In the present study, we aimed to systematic review literatures studying the prognostic role of pretreatment LDH in UC and merged the quantitative data.

Study Design
The present study was performed according to the PRISMA statements (Supplementary Material) (6). The protocol has been registered on PROSPERO (No. CRD42019147216).

Selection Criteria
In general, the present study included literatures investigating the prognostic role of pretreatment of serum lactate dehydrogenase (LDH) in patients with urothelial carcinoma, and detailed criteria were shown as follows. Inclusion criteria: (1) studies involving patients confirmed diagnosed with urothelial carcinoma; (2) studies that measured serum lactate dehydrogenase of patients before treatment; (3) studies that reported the results of oncological outcomes including overall survival (OS), cancerspecific survival (CSS), and disease-free survival (DFS); (4) studies that directly provided hazard ratios (HRs) and 95% confidence interval (CI) or required data for calculating them. The methods for calculation were reported by Tierney et al. (7). Exclusion criteria: (1) studies wherein we cannot extract HRs and 95% CI; (2) studies that did not analyze serum lactate dehydrogenase as dichotomous variable or did not clearly report the cut-off value; (3) not original article, such as review, abstract, opinion, letter, and so on; (4) duplicated results from the same cohort. Two researchers independently screened and evaluated the literature; the disputes were resolved by discussion.  The quality of each study was evaluated using the Newcastle-Ottawa scale, which embraced three aspects including patient selection, comparability, assessment of outcome, and follow-ups.

Statistical Analysis
We extracted all HRs and 95% CIs of related endpoints from included studies. Subgroup analyses of overall survival were also conducted. The grouping variables included publication year, region, site of malignance, number of patients, cancer stage, cut-off value, and NOS score. Meta-regression was also performed to identify the resources of interstudy heterogeneity. Subgroup analyses of upper tract urothelial carcinoma (UTUC) and urothelial carcinoma of the bladder (UCB) were also performed. HRs and the corresponding 95% CI were used to assess the significance of the prognostic value of serum lactate dehydrogenase for urothelial carcinoma. A pooled value >1 was regarded as indicating an unfavorable outcome for the subjects having a high level of serum lactate dehydrogenase. Cochran's Q test and Higgins I² statistic were both conducted to examine inter-study heterogeneity. A random effects model was applied for all analyses. Publication bias was assessed by funnel plots, Egger's, and Begg's tests. Sensitivity analysis was performed to check the stability of the results. All statistical analyses were performed with Stata 12.0 (STATA Corporation, College Station, TX, USA).

Urothelial Carcinoma
First, all included studies were analyzed together. For the three endpoints (overall survival, cancer-specific survival, disease-free survival), there were 12 (2, 14-17, 19, 20, 22-26) (Figure 2). Subgroup analyses of overall survival were also conducted, and the results are presented in ( Table 4). The subgroup variables included year of publication, region, site of carcinoma, sample size, survival analysis, and NOS score, which did not obviously change the results. Moreover, except for survival analysis (p = 0.033), other subgroup variables were not the resources of inter-study heterogeneity (P > 0.05 for all).

Publication Bias and Sensitivity Analysis
Since most analyses embraced insufficient literatures, we only checked publication bias and performed sensitivity analysis for overall survival of patients with urothelial carcinoma. The funnel plot showed an approximately asymmetric result, and quantitative tests identified significant differences (Begg's test: p = 0.024; Egger's test: p = 0.005) (Figure 4A). The result of sensitivity analysis showed that excluding any study did not significantly change the merged data ( Figure 4B).

DISCUSSION
Previous literatures have found that cancer cells metabolize differently from normal cells. Even with sufficient oxygen, malignant cells will preferentially metabolize glucose through glycolysis to produce adequate energy for growth, which is known as the Warburg effect and is one of the major metabolic changes in the process of malignant transformation (4). The serum level of LDH, the enzyme involved in the glycolytic pathway, can reflect the metabolic alterations (27). A high level of serum LDH has been reported to serve as an unfavorable prognostic factor in several types of malignances. Several studies have examined the prognostic role of LDH in subjects with UC and have reported inconsistent results. Hence, the present systematic review and meta-analysis about this issue was performed. After systematic literature searching and screening, 14 studies embracing 4,009 patients were included. All studies were of lowto-moderate risk of bias, and the NOS score ranged from 6 to 8. The HRs and 95% CIs extracted from these studies were merged. The results showed that a high pretreatment serum LDH was associated with an inferior overall survival, cancer-specific survival, and disease-free survival in UC. Subgroup analyses of OS showed that grouping variable did not obviously change the results. Moreover, sensitivity analysis further confirmed the stability and reliability of the results. Subgroup analyses of UTUC and UCB were also performed. We identified that a high pretreatment serum LDH was associated with a poor overall survival and disease-free survival in UTUC, a short overall survival in UCB. Since circulating blood LDH is easy to be measured clinically, it can be used as an indicator of cancer burden and a useful biomarker in management of UC.
For the prognostic significance of pretreatment LDH in cases with UC, only one meta-analysis previously was reported. In 2016, Zhang et al. (28) performed a meta-analysis to evaluate the prognostic role of LDH for cases with urological cancer. They included two studies about bladder cancer and one study about UTUC. A high level of serum LDH has been reported to serve as an unfavorable prognostic factor. With more and later studies being included, the present study may provide more comprehensive and reliable results for this issue.
Metabolically, the most prominent feature of cancer cells is an increase in lactic acid production due to their increased glucose uptake rate and reduced oxidative phosphorylation rate, regardless of the availability of oxygen. The phenomenon, called aerobic glycolysis, was first discovered 70 years ago by Otto Warburg (4). As an important substance of the Warburg effect, LDH exists in nearly all kinds of tissues, which has six different isoenzymes. These isoenzymes are assembled from two protein subunits, LDHA and LDHB, into a homo-or hetero-tetramer structure (29). The third protein subunit, called LDHC, forms the testicular specific subtype LDH-6 (30). The main function of lactate dehydrogenase is to catalyze the reversible reaction between pyruvate and lactic acid. NAD + is produced along with this process and is necessary for the continuous production of ATP to maintain glycolysis. Among the LDH isoforms, LDH-5 has the highest catalytic efficiency (31). When cancer tissue is necrotic, high intracellular LDH levels are released into the blood, increasing serum LDH concentrations (32). In addition, when distant metastases occur, tumor cells can damage adjacent organs, such as the lung, liver, and bone. Damage to these organs can also increase serum LDH levels (31,(33)(34)(35). In conclusion, serum LDH seems to be a significant factor in the development of malignance. Its level can reflect tumor burden and serve as cancer biomarker.
There were several limitations for the present study. First, due to limited literatures, we studied all types and stages of UC together. The differences in type, stage, and treatment for UC may be the sources of inter-study heterogeneity. Despite both belonging to UC, UTUC, and UCB may have different biological behaviors and disease characteristics. Hence, subgroup analyses of UTUC and UCB were also performed. Second, serum LDH may be affected by other non-tumor diseases, such as anemia, hepatic disease, muscular dystrophy, and heart failure, which have not been examined in these studies. Moreover, the present study also included results from univariable analysis. These uncontrolled factors may affect the results, and the methods of survival analysis have been found to be the source of inter-study heterogeneity. Third, taken that different cut-off values were used in the included literatures, it is a problem to choose the best one, which may affect the application of this biomarker in clinical practice. Fourth, a significant publication bias was found in the metaanalysis of overall survival, which cannot be overcome by statistical methods. Regardless, the present study represents the most comprehensive and latest systematic review and meta-analysis about the prognostic role of LDH in patients with UC.
In general, the present study proved that a high level of pretreatment serum LDH was associated with inferior OS, CSS, and DFS in patients with UC. Moreover, an elevated level of serum LDH was associated with poor OS and DFS in patients with UTUC, a short OS in patients with UCB. Despite more high-level studies were needed to verify our results, the level of serum LDH can be an important factor incorporated into the prognostic models for UC.

AUTHOR CONTRIBUTIONS
MW and PL: protocol/project development. LX, ZY, and QC: data collection or management. MW, HG, and CL: data analysis. MW and PL: manuscript writing/editing. All authors read and approved the final manuscript.