This study suggested that serum LDH level was associated with T stage, N stage, tumor size, M stage, pathological type, and LVI status. Serum LDH level had a moderate diagnostic value in detecting BC patients. Multivariate Cox regression indicated that non-UCB, T2–3, 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 [11,12,13]. We assumed the following factors may provide evidence for these conflicting 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 contradictory 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 results.

Regarding UTUC, other studies [15,16,17] 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. 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 [20,21,22]. 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 be investigated in this study, thereby exerting effects on final results. Third, the findings observed by this 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. Last, regarding for the ROC curve (Fig. 2), the AUC value was 0.615, indicating that there was no clear cut differentiation. High LDH level might just be another marker for poor performance.