Serum Lactate Dehydrogenase Level Predicts the Prognosis in Bladder Cancer Patients

Abstract

Background

Recently, several studies investigated the association between lactate dehydrogenase (LDH) level and the prognosis of urothelial carcinoma. However, no studies explored the role of serum LDH level in the survival of overall bladder cancer (BC). In this study, we intended to address the association of LDH level with the prognosis of BC.

Methods

206 patients with BC were included in this study. The clinical data and blood samples of patients were collected. The overall survival (OS) and progression-free survival (PFS) were used. Kaplan-Meier method and Log rank test were used to evaluate the effects of LDH level on the survival of BC. Univariate and multivariate Cox regression analyses were utilized to identify prognosis predictors of BC.

Results

Data indicated that serum LDH level in the BC patients was significantly higher than those in controls. In addition, this study suggested that serum LDH level was associated with T stage, N stage, tumor size, M stage, pathological type, treatment methods, and lymphovascular invasion. The Kaplan-Meier analysis found significant differences in the OS and PFS rate between lower and higher serum LDH level groups (LDH ≥ 225 U/L and < 225 U/L). Multivariate Cox regression indicated that pathological type, treatment methods, and higher level of LDH were independently associated with adverse prognosis in BC patients.

Conclusion

The higher serum LDH level (≥ 225 U/L) is associated with poor prognosis in patients with BC. Serum LDH level could be used as a novel predictive biomarker for BC patients.

Introduction

Bladder cancer (BC) is a prevalent malignancy in women and is the fourth most main malignancy in men [1]. More than 430,000 BC patients were diagnosed globally every year, which brings a heavy burden for the society [2]. A host of risk factors including cigarette smoking, alcohol intake, diets low in fruits and vegetables, urban living, occupational exposure to carcinogens, and genetic factors were associated with the development of BC [3]. Among these factors, cigarette smoking is the most primary risk factor for BC. It is estimated that tobacco was responsible for approximately 50% of all BC patients [4]. The most common symptom of BC is haematuria [5]. However, some BC patients with microscopic haematuria were not adequate [6]; and there were no effective screening methods for BC, except for the cystoscope. Thus, detecting novel prognostic and diagnostic biomarkers is urgently required for patients with BC.

Most cancer cells utilize glucose metabolism by glycolysis to produce enough ATP to promote the development of cancer, which is served as the Warburg effect [7]. Among the enzymes participating in glycolysis, lactate dehydrogenase (LDH), an oxidoreductase, could converts pyruvate to lactate at the end of glycolysis; thereby, LDH is regarded as a promising molecular target for the development of new glycolytic inhibitors used in cancer therapy [8]. Serum LDH level may be a biomarker of tumor development due to its association with tumor metabolic changes [9].
Urothelial carcinoma (UC) mainly consists of urothelial carcinoma of the bladder (UCB) and upper tract urothelial carcinoma (UTUC). Recently, a host of studies investigated the link between LDH level and the prognosis of UC.[10–24] Among these studies, no studies investigated the role of serum LDH level in the survival of overall BC; and previous studies mainly focused on UCB patients (the main type of BC). In this study, we intended to address the association of LDH level with clinicopathological markers of BC. In addition, we aimed to evaluate the ability of pretreatment serum LDH level in predicting the prognosis of BC patients.

Methods

Patients

This study was conducted after approval from the Ethics Committee of Huaian No.1 People’s Hospital. 206 consecutive BC patients were recruited between September 2019 and May 2017 from our Hospital. BC patients were divided into two groups, which was according to the pretherapeutic median levels of LDH. The diagnosis of BC patients was according to histological results. All relevant clinical and pathologic information was collected for each patient. Inclusion criteria for BC patients were as follows: one, patients with definite histologic evidence; two, patients older than 18 years; three, patients with complete data. Patients with other cancers, hepatitis, heart disease, and hemopathy were excluded. The informed consent form was provided by all subjects. This study was approved by the Ethics Committee of this hospital. This study was in consistent with the Declaration of Helsinki.

Data Collection and Follow-up

Clinical information was collected from the medical records, including age, sex (male and female), smoking status (yes and no), T stage, N stage (N0 and N1-N3), tumor size (< 3 cm and > 3 cm), M stage (M0 and M1), lymphovascular invasion (LVI) status, perineural invasion (PNI) status, pathological type, treatment methods, and multifocality (Unifocal and Multifocal). Serum LDH levels were collected from all BC patients before treatment. Overall survival (OS) was defined as the survival duration from the date of initial diagnosis to death due to any reasons or last follow-up. Progression-free survival (PFS) referred to the time for BC patients to survive without progression or death after treatment. The follow-up strategies were as follows: every two months for the first 2 years, quarterly for the next 2 years, and then semiannually for the remaining time thereafter. The follow-up time ended in June 2021.

Statistical Analysis

The correlation between serum LDH level and clinicopathological indexes was analyzed. We used the receiver operating characteristic (ROC) curve to assess the diagnostic ability of serum LDH level for detecting BC. The log-rank test was used to compare the OS and PFS between different groups. Univariate and Multivariate Cox regression analyses were conducted to identify prognostic variables associated with the survival of BC. Risk factors with a P value < 0.05 in univariate analysis were choose in further multivariate Cox regression analysis. 95% confidence interval (CI) and Hazard risk (HR) were figured out. Statistically significant Differences were set at P < 0.05. SPSS (version 19.0, Chicago, IL, USA), and Graphpad Prism (version 7.0, La Jolla, CA, USA), and MedCalc were used for statistical analyses.

Results

Clinicopathological variables of BC patients and LDH level

The cut-off value of LDH level was defined as 225 U/L, which was the upper limit of normal range in this hospital. Thus, we defined the high LDH level group was over 225 U/L, and the low LDH level group was lower than 225 U/L. Clinicopathological variables for BC patients with LDH ≥ 225 U/L and < 225 U/L are shown in Table 1. 223 patients with BC were evaluated initially, and 206 cases were included finally (Fig. 1). Among the 206 BC patients, 149 cases were in the low LDH group and 57 cases were in the high LDH group. We found that LDH level was associated with T stage, N stage, tumor size, M stage, pathological type, treatment methods, and LVI status. No significant differences were observed in two groups regarding age, sex, smoking status, PNI status, and multifocality status.

Table 1

Demographics of 206 patients with the bladder cancer.
Variables	Total patients	Groups			P
		High LDH		Low LDH
	(n = 206)	< 225 U/L	≥ 225 U/L
Median age (years)					0.073
< 64	93	73	20
≥ 64	113	76	37
Sex, n%					0.709
Male	159	114	45
Female	47	35	12
Smoking					0.592
Yes	95	67	28
No	111	82	29
Pathological type					< 0.001
Transitional cell carcinoma	188	144	44
Squamous cell carcinoma	12	4	8
Adenocarcinoma	6	1	5
T stage					< 0.001
Ta, Tis, T1	122	101	21
T2-3	84	48	36
N stage					< 0.001
N0	183	144	39
N1-N3	23	5	18
M stage					< 0.001
M0	188	146	42
M1	18	3	15
Tumor size (cm)					0.011
< 3 cm	55	47	8
> 3 cm	151	102	49
LVI					< 0.001
Present	76	43	33
Absent	130	106	24
PNI					0.075
Present	29	17	12
Absent	177	132	45
Multifocality					0.559
Unifocal	161	118	43
Multifocal	45	31	14
Treatment methods					< 0.001
Surgery	194	147	47
Chemotherapy	9	2	7
Immunotherapy	3	0	3
LVI, lymphovascular invasion; PNI, perineural invasion, LDH, Lactate dehydrogenase.
Bold values are statistically significant (P < 0.05).

Diagnostic value of LDH level for detecting BC

230 healthy controls were included in this study. The detailed information for controls is summarized in Supplementary Table 1. We measured the serum level of LDH in BC patients and controls, and found that the LDH level in BC group were significantly higher than those in control group. We used the ROC curve to evaluate the diagnostic function of serum LDH level for BC (Fig. 2). The sensitivity and specificity were 47.57% and 71.74% with a cutoff value of 211.53 (Supplementary Table 2), respectively. The area under the curve was 0.615 (95%CI: 0.567–0.661, P < 0.001), suggesting that the diagnostic ability of serum LDH level for detecting BC was relatively moderate.

Correlation between LDH level and the prognosis in patients with BC

We divided all BC patients into two groups (LDH ≥ 225 U/L and < 225 U/L) according to the level of LDH. In order to evaluate the prognostic value of LDH in BC, Kaplan-Meier analysis was utilized to assess the association between LDH level and follow-up data, and the log-rank test was mainly used for analyzing statistical difference. The results showed that the high level of LDH was negatively correlated with OS and PFS (Figs. 3 and 4), suggesting that the survival rate of BC patients with lower level of LDH were significantly higher than those with higher level of LDH. In addition, we compared the survival rates between UCB group and non-UCB group (squamous cell carcinoma and adenocarcinoma). Data indicated that the survival rate of BC patients with UCB was significantly higher than that in non-UCB group (Supplementary Figs. 1 and 2).

Identification of Prognostic Factors of OS and PFS

Univariate and multivariate Cox analyses were used to identify prognostic factors of BC patients. Univariate analysis showed that older age, T2-3 stage, N1-3 stage, M1 stage, LVI, PNI, non-UCB, treatment methods (chemotherapy/immunotherapy vs. surgery), and higher level of LDH were significantly associated with reduced OS (Table 2), while smoking, T2-3 stage, N1-3 stage, M1 stage, tumor size ≥ 3cm, LVI, PNI, non-UCB, treatment methods (chemotherapy/immunotherapy vs. surgery), and higher level of LDH negatively correlated with PFS (Table 3). Multivariate analysis indicated that T2-3 stage, non-UCB, treatment methods (chemotherapy/immunotherapy vs. surgery), and higher level of LDH were independent predictive factors for worse OS, while tumor size ≥ 3cm, non-UCB, treatment methods (chemotherapy/immunotherapy vs. surgery), and higher level of LDH correlated with adverse PFS. In addition, we identified prognostic factors for BC patients with UCB, and found that older age, T2-3 stage, treatment methods (chemotherapy/immunotherapy vs. surgery), and higher LDH level were prognostic factors for worse OS (Supplementary Table 3). Furthermore, we observed that T2-3 stage, M1 stage, Tumor size ≥ 3cm were adverse factors for PFS in UCB (Supplementary Table 4).

Table 2

Univariate and multivariate cox regression analysis for overall patient survival.
Factors	Overall survival
	Univariate		Multivariate
	HR (95%CI)	P-value	HR (95%CI)	P-value
Age
≥ 64 vs. < 64 years	3.48(1.69–7.21)	0.001	1.43(0.64–3.19)	0.384
Sex
Male vs. female	0.88(0.46–1.70)	0.713	-	-
Smoking
Yes vs. no	1.43(0.80–2.54)	0.226	-	-
Pathological type
Squamous/Adeno- vs. Transitional cell carcinoma	16.91(8.69–32.90)	< 0.001	3.18(1.28–7.88)	0.013
T stage
T2-T3 vs. Ta, Tis, T1	8.02(3.75–17.18)	< 0.001	3.43(1.44–8.14)	0.005
N stage
N1-N3 vs. N0	13.92(7.65–25.32)	< 0.001	1.95(0.48–7.82)	0.348
M stage
M1 vs. M0	14.77(7.94–27.46)	< 0.001	0.88(0.26–3.03)	0.839
Tumor size
≥ 3 vs. < 3cm	1.64(0.79–3.40)	0.182	-	-
LVI
Present vs. absent	4.74(2.54–8.87)	< 0.001	1.47(0.65–3.36)	0.358
PNI
Present vs. absent	3.04(1.60–5.76)	0.001	1.05(0.43–2.55)	0.918
Multifocality
Multifocal vs. unifocal	1.04(0.53–2.05)	0.900	-	-
Treatment methods
Chemotherapy/Immunotherapy vs. Surgery	18.20(8.78–37.74)	< 0.001	2.76(1.06–7.18)	0.037
LDH
High vs. low	7.23(3.94–13.24)	< 0.001	3.47(1.72–6.96)	< 0.001
LVI, lymphovascular invasion; PNI, perineural invasion, LDH, Lactate dehydrogenase.
Bold values are statistically significant (P < 0.05).

Table 3

Univariate and multivariate cox regression analysis for progression-free survival
Factors	Progression-free survival
	Univariate			Multivariate
	HR (95%CI)	P-value		HR (95%CI)	P-value
Age
≥ 64 vs. < 64 years	1.03(0.66–1.60)		0.910	-	-
Sex
Male vs. female	1.52(0.84–2.76)		0.165	-	-
Smoking
Yes vs. no	1.59(1.02–2.48)		0.041	1.49(0.93–2.37)	0.097
Pathological type
Squamous/Adeno- vs. Transitional cell carcinoma	9.12(4.76–17.49)		< 0.001	2.60(1.12–6.03)	0.026
T stage
T2-T3 vs. Ta, Tis, T1	2.69(1.72–4.22)		< 0.001	1.68(0.99–2.85)	0.055
N stage
N1-N3 vs. N0	13.67(7.58–24.65)		< 0.001	3.00(0.80-11.35)	0.105
M stage
M1 vs. M0	18.37(9.42–35.84)		< 0.001	2.05(0.54–7.68)	0.289
Tumor size
≥ 3 vs. < 3cm	3.20(1.60–6.42)		0.001	2.59(1.26–5.35)	0.010
LVI
Present vs. absent	1.82(1.17–2.83)		0.008	0.82(0.45–1.50)	0.523
PNI
Present vs. absent	2.18(1.22–3.91)		0.008	1.18(0.54–2.60)	0.672
Multifocality
Multifocal vs. unifocal	1.01(0.60–1.71)		0.978	-	-
Treatment methods
Chemotherapy/Immunotherapy vs. Surgery	19.06(8.96–40.55)		< 0.001	2.80(1.02–7.70)	0.046
LDH
High vs. low	3.08(1.96–4.83)		< 0.001	1.74(1.03–2.94)	0.037
LVI, lymphovascular invasion; PNI, perineural invasion, LDH, Lactate dehydrogenase.
Bold values are statistically significant (P < 0.05).

Discussion

This study suggested that serum LDH level was associated with T stage, N stage, tumor size, M stage, pathological type, treatment methods, and LVI status. Serum LDH level had a moderate diagnostic value in detecting BC patients. Multivariate Cox regression indicated that non-UCB, treatment methods, and higher level of LDH were independently associated with adverse prognosis in BC patients (including OS and PFS). Therefore, this study provided evidence that higher LDH level was associated with worse prognosis in BC and could be a prognostic indicator of BC patients.

Hannisdal et al. from Denmark first showed that UCB patients with LDH > 400 U/L showed shorter survival comparing to those with LDH < 400 U/L [11]. Two Japanese studies found that higher LDH level was negatively associated with the OS of UCB after radical cystectomy [12] or postcystectomy recurrent UCB [13]. However, a study by Yang et al. from Taiwan indicated that serum LDH level was not related with the disease-specific survival of UCB in their population [10], which was not consistent with previous findings [11–13]. We assumed the following factors may provide evidence for these conflict points. One, the results of Taiwanese study was obtained by the univariate analysis [10], which may underpower the credibility of results. Two, clinical heterogeneity should not be ignored. Two Japanese studies investigated the recurrent UCB [12, 13], while the Taiwanese study explored invasive UCB [10]. Three, their indicators evaluating survival differed, such as OS or disease-free survival (DFS). Four, their varied treatment therapies may also contribute to their conflicting results concerning the survival. It is of note that Yang et al. conducted another study [14], and found that serum LDH level ranging from 200 to 300 U/L was an independent factor associated with UTUC after multivariate analysis, but not for UCB. Obviously, the tumor location may be a crucial reason for explaining their contradictory results by Yang et al. [10, 14] Last but not least, the cut-off values of serum LDH level were diverse, which may exert effects on the final result.

Regarding UTUC, other studies [15–17] also delineated the association between serum LDH level and the prognosis of UTUC. Zhang et al. showed that preoperative serum LDH level was as a negative predictor of OS and DFS [15], while Tan et al. indicated that preoperative LDH was not an independent prognostic indicator for patients with UTUC [15]. However, Tan et al. suggested that elevated LDH level correlated with worse OS in UTUC patients with localized disease [16]. Kluth et al. also indicated that serum LDH level was not associated with the survival of UTUC [18]. As for the Japanese study by Ito et al., they observed that LDH ≥ 210 IU/L were significantly related with extraurothelial recurrence in N₀M₀ patients with renal pelvic cancer [17], but not OS or DFS.

Besides UCB and UTUC, some researchers shed light on the investigation of serum LDH level and UC survival. Sengelov et al. first showed that LDH level were related significantly to the survival of UC by univariate analyses [19], similar to the findings by Japanese studies via multivariate analyses [20–22]. In addition, two studies from Japan [23] and Spain [24] indicated that serum LDH level was not a prognostic factor in UC patients. We hypothesized that the negative results in the Spanish study [24] may due to its limited sample size (only 56 UC cases). Additionally, they utilized univariate analyses to obtain these results [24], which underpowered the reliability of their results.

Due to these conflicting findings concerning UCB, UTUC or UC, Wu et al. conducted a meta-analysis to address this issue [25]. They suggested that a high pretreatment serum LDH level was linked with an inferior OS, cancer-specific survival, and DFS in UC patients [25]. Subgroup analyses revealed that high serum LDH level was associated with a poor OS and DFS in UTUC, and a short OS in UCB [25]. It is noticeable that the meta-analysis by Wu et al. did not include a study [18], which was in line with inclusion criteria of this meta-analysis. Another meta-analysis by Zhang et al. also indicated that a high LDH level was associated with an adverse prognosis in many solid tumors [26].

In this study, we included 206 BC patients, and found that a higher pretreatment serum LDH level was associated with an unfavorable prognosis in BC patients. Abovementioned studies primarily investigated the survival of UCB patients (the main type of BC), while this study explored the prognosis of overall BC with UCB, squamous cell carcinoma, and adenocarcinoma. We compared the survival rates between UCB group and non-UCB group, and found that the survival rate of BC patients with UCB was significantly higher than that in non-UCB group, which was not investigated in other studies. Furthermore, multivariate analysis indicated that pathological type (non-UCB vs UCB) was an independent predictive factor for worse OS and PFS.

This study had several limitations. First, the sample size of this study was not large enough. Second, some confounding factors affecting the survival of BC may not investigated in this study, thereby exerting effects on our results. Third, the findings observed by this cohort study were only yielded in one single center; thus, multi-center studies are urgently needed to verify these findings. Fourth, UTUC patients were not investigated in this study. Fifth, diverse treatment strategies may affect the survival analysis, thereby interfering the effect of serum LDH level on the prognosis of BC patients.

Conclusions

Totally, this study observes that a higher pretreatment serum LDH level is associated with adverse prognosis in BC patients. In addition, serum LDH level may predict the occurrence of BC patients. Thus, LDH level could serve as a prognostic biomarker for BC patients.

Declarations

Funding

This study did not receive any funding in any form.

Author contributions

Shuo Gu and Chao Yang: study design. Chao Yang: data collection. Shuo Gu and Chao Yang: manuscript writing. Shuo Gu and Chao Yang: critical revision of the manuscript. Shuo Gu supervision. All authors reviewed the manuscript.

Acknowledgements

Not applicable.

Ethics and consent to participate

This study was approved by the Ethics Committee of The Affiliated Huaian No.1 People's Hospital of Nanjing Medical University. This study was in consistent with the Declaration of Helsinki. Informed consent was obtained from all subjects and/or their legal guardian.

Consent for publication

Not Applicable.

Availability of data and materials

All data generated or analyzed during this study are included in this published article [and its supplementary information files].

Competing interests

We have read and understood the BMC policy on declaration of interests and declare that there are no conflicts of interest.

References

Lenis AT, Lec PM, Chamie K, Mshs MD: Bladder Cancer: A Review. JAMA 2020, 324(19):1980-91.
Bray F, Ferlay J, Soerjomataram I et al: Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018, 68(6):394-424.
Cohen PA, Jhingran A, Oaknin A, Denny L: Cervical cancer. Lancet 2019, 393(10167):169-82.
Freedman ND, Silverman DT, Hollenbeck AR, Schatzkin A, Abnet CC: Association between smoking and risk of bladder cancer among men and women. JAMA 2011, 306(7):737-45.
Ramirez D, Gupta A, Canter D et al: Microscopic haematuria at time of diagnosis is associated with lower disease stage in patients with newly diagnosed bladder cancer. BJU Int 2016, 117(5):783-6.
Elias K, Svatek RS, Gupta S, Ho R, Lotan Y: High-risk patients with hematuria are not evaluated according to guideline recommendations. Cancer 2010, 116(12):2954-9.
Siska PJ, Singer K, Evert K, Renner K, Kreutz M: The immunological Warburg effect: Can a metabolic-tumor-stroma score (MeTS) guide cancer immunotherapy? Immunol Rev 2020, 295(1):187-202.
Gallo M, Sapio L, Spina A et al: Lactic dehydrogenase and cancer: an overview. Front Biosci (Landmark Ed) 2015, 20:1234-49.
Serganova I, Rizwan A, Ni X et al: Metabolic imaging: a link between lactate dehydrogenase A, lactate, and tumor phenotype. Clin Cancer Res 2011, 17(19):6250-61.
Yang MH, Yen CC, Chen PM et al: Prognostic-factors-based risk-stratification model for invasive urothelial carcinoma of the urinary bladder in Taiwan. Urology 2002, 59(2):232-8; discussion 38-9.
Hannisdal E, Fossa SD, Host H: Blood tests and prognosis in bladder carcinomas treated with definitive radiotherapy. Radiother Oncol 1993, 27(2):117-22.
Nakagawa T, Hara T, Kawahara T et al: Prognostic risk stratification of patients with urothelial carcinoma of the bladder with recurrence after radical cystectomy. J Urol 2013, 189(4):1275-81.
Nakagawa T, Taguchi S, Uemura Y et al: Nomogram for predicting survival of postcystectomy recurrent urothelial carcinoma of the bladder. Urol Oncol 2017, 35(7):457 e15-57 e21.
Yang MH, Chen KK, Yen CC et al: Unusually high incidence of upper urinary tract urothelial carcinoma in Taiwan. Urology 2002, 59(5):681-7.
Zhang XK, Zhang ZL, Lu X et al: Prognostic Significance of Preoperative Serum Lactate Dehydrogenase in Upper Urinary Tract Urothelial Carcinoma. Clin Genitourin Cancer 2016, 14(4):341-45 e3.
Tan P, Chen J, Xie N et al: Is preoperative serum lactate dehydrogenase useful in predicting the outcomes of patients with upper tract urothelial carcinoma? Cancer Med 2018, 7(10):5096-106.
Ito K, Asakuma J, Kuroda K et al: Preoperative risk factors for extraurothelial recurrence in N0M0 patients with renal pelvic cancer treated by radical nephroureterectomy. Mol Clin Oncol 2016, 4(4):530-36.
Kluth LA, Xylinas E, Kent M et al: Predictors of survival in patients with disease recurrence after radical nephroureterectomy. BJU Int 2014, 113(6):911-7.
Sengelov L, Kamby C, Schou G, von der Maase H: Prognostic factors and significance of chemotherapy in patients with recurrent or metastatic transitional cell cancer of the urinary tract. Cancer 1994, 74(1):123-33.
Fukushima H, Yokoyama M, Nakanishi Y, Tobisu K, Koga F: Sarcopenia as a prognostic biomarker of advanced urothelial carcinoma. PLoS One 2015, 10(1):e0115895.
Takemura K, Fukushima H, Ito M et al: Prognostic significance of serum gamma-glutamyltransferase in patients with advanced urothelial carcinoma. Urol Oncol 2019, 37(2):108-15.
Suzuki H, Ito M, Takemura K et al: Prognostic significance of the controlling nutritional status (CONUT) score in advanced urothelial carcinoma patients. Urol Oncol 2020, 38(3):76 e11-76 e17.
Abe T, Ishizaki J, Kikuchi H et al: Outcome of metastatic urothelial carcinoma treated by systemic chemotherapy: Prognostic factors based on real-world clinical practice in Japan. Urol Oncol 2017, 35(2):38 e1-38 e8.
Bellmunt J, Albanell J, Paz-Ares L et al: Pretreatment prognostic factors for survival in patients with advanced urothelial tumors treated in a phase I/II trial with paclitaxel, cisplatin, and gemcitabine. Cancer 2002, 95(4):751-7.
Wu M, Lin P, Xu L et al: Prognostic Role of Serum Lactate Dehydrogenase in Patients With Urothelial Carcinoma: A Systematic Review and Meta-Analysis. Front Oncol 2020, 10:677.
Zhang J, Yao YH, Li BG et al: Prognostic value of pretreatment serum lactate dehydrogenase level in patients with solid tumors: a systematic review and meta-analysis. Sci Rep 2015, 5:9800.