Serum lactate dehydrogenase predicts prognosis and correlates with systemic inflammatory response in patients with advanced pancreatic cancer after gemcitabine-based chemotherapy.

abstract CC BY

Serum lactate dehydrogenase (LDH) concentrations correlate with tumor progression and poor outcome. We evaluated the predictive value of serum LDH level for overall survival (OS) of patients with advanced pancreatic cancer after gemcitabine-based chemotherapy. We retrospectively enrolled 364 patients with locally advanced or metastatic pancreatic adenocarcinoma who were then allocated to training (n = 139) and validation cohorts (n = 225). We evaluated the association between serum LDH levels and OS as well as with markers of systemic inflammation, including neutrophil/lymphocyte ratio (NLR), platelet/lymphocyte ratio (PLR) and lymphocyte/monocyte ratio (LMR). Kaplan–Meier analyses revealed that low serum LDH levels in the training cohort significantly correlated with longer OS. Multivariate analysis identified the serum LDH levels as an independent prognostic predictor of OS (p = 0.005). Serum LDH levels correlated positively with NLR and PLR and correlated negatively with LMR. Similar results were obtained for the validation cohort, except that multivariate analysis identified the serum LDH level as a significant prognostic predictor and only a statistical trend for OS (p = 0.059). We conclude that serum LDH levels were associated with the systemic inflammatory response and served as a significant prognostic predictor of OS. Serum LDH levels predicted OS in patients with advanced pancreatic cancer after gemcitabine-based palliative chemotherapy.

patients’ characteristics CC BY

We enrolled 364 consecutive patients with the characteristics as follows: 229 (62.9%) were men, 177 (48.6%) had CA19-9 concentrations ≥1000 IU/mL and 92 (25.3%) were diagnosed with stage III pancreatic cancer. The tumour was located in the pancreatic head in 143 patients (39.3%) and in the body/tail in 221 (60.7%). All patients received gemcitabine-based palliative chemotherapy. The median number of times that patients received gemcitabine-based interventional therapy or conventional systemic venous chemotherapy (gemcitabine monotherapy or combinations such as gemcitabine plus cisplatin/oxaliplatin or gemcitabine plus albumin-bound paclitaxel) were 2.04 and 3.26, respectively, in the training cohort. The median number of times that patients in the validation cohort received these respective therapies were 2.43 and 3.91,respectively. The total numbers of treatments were calculated as the interval between the date of the first treatment and death (or the last observation point taken).The first 139 patients were assigned to the training cohort, and the remaining 225 were assigned to the validation cohort. The clinicopathological characteristics of the patients are summarized in Table 1.

value of serum ldh levels for predicting os CC BY

Kaplan–Meier analysis of the OS of patients in the training cohort revealed that the serum level of LDH, tumour stage (III vs. IV), level of CA19-9, NLR and MLR significantly affected prognosis (Fig. 1a,c,f–h). In contrast, there was not a significant correlation between the serum level of LDH and tumour location, age, sex, Karnofsky performance status (KPS), metastasis (intrahepatic or extrahepatic), location of intrahepatic metastasis (right, left or both lobes), albumin or PLR (Fig. 1b,d,e,i–m). Moreover, low serum LDH levels (<185 IU/mL) significantly correlated with longer OS of patients in the training cohort (148 days and 308 days for patients with high or low LDH levels, respectively, 95% CI: 174.7–333.3; log rank, 10.504; p = 0.001) (Fig. 1a). We found that the level of serum LDH, stage (III vs. IV), CA19-9, NLR and MLR affected prognosis (Fig. 2a–e). In contrast, there was not a significant correlation between the serum level of LDH and tumour location, age, sex, KPS, metastasis (intrahepatic or extrahepatic), location of intrahepatic metastasis (right, left, or both lobes), albumin or PLR, similar to the findings for the test cohort. The prognostic value of LDH was confirmed in the independent validation cohort (OS 226 days and 301 days for patients with high (≥185 IU/mL) and low LDH levels, respectively, 95% CI: 283.4.7–342.7; log rank, 5.793; p = 0.016) (Fig. 2a). Further, OS was independent of tumour stage (III vs. IV). Together, the results suggest that serum LDH levels were significantly associated with OS of patients with advanced pancreatic cancer.

univariate and multivariate analysis of prognostic factors in the training and v CC BY

In the training cohort, 93/139 patients (67%) died during the observation period, and 46 (33%) were alive after a median follow-up of 78 months. The median OS of the training population was 254 days (95% CI: 174.7–333.3). Univariate analysis was performed to determine the prognostic value of serum LDH level and other clinical variables for OS of the training cohort (Table 2). Poor OS was predicted by tumour stage (III vs. IV), CA 19-9 (<1000 vs. ≥1000 IU/mL), serum LDH levels (≥185 IU/mL vs. <185 IU/mL), NLR (<3.42 vs. ≥3.42) and LMR(<3.19 vs. ≥3.19) (p = 0.004, p <0.001, p = 0.002, p < 0.001 and p = 0.001, respectively). These five variables were included in the multivariate analysis using Cox proportional hazards regression. Tumour location, age, sex, KPS (<90 vs. ≥90), albumin (<38.6 g/L vs. ≥38.6 g/L), metastasis (intrahepatic or extrahepatic), location of intrahepatic metastasis (right, left or both lobes) and PLR (<154 vs. ≥154) were not significantly associated with OS (all p > 0.05). Multivariate analysis identified serum LDH levels as an independent prognostic predictor of OS (HR = 1,981; 95% CI: 1.226–3.200; p = 0.005) (Table 2). In the validation cohort, 190 of the 225 patients (84%) died during the observation period, and 35 (16%) were alive after a median follow-up of 78 months. The median OS of the training population was 260 days (95% CI: 225.8–294.2). Univariate analysis revealed that age (<60 years vs. ≥60 years), tumour stage (III vs. IV), CA 19-9 (<1000 vs. ≥1000 IU/mL), serum LDH levels (≥185 IU/mL vs. <185 IU/mL), NLR (<3.42 vs. ≥3.42) and LMR (<3.19 vs. ≥3.19) predicted poor OS (p = 0.056, p = 0.013, p <0.001, p = 0.016, p = 0.004 and p = 0.001, respectively). These six variables were included in the multivariate analysis using Cox proportional hazards regression. Similarly, tumor location, sex, KPS (<90 vs. ≥90), albumin (<38.6 g/L vs. ≥38.6 g/L), metastasis (intrahepatic or extrahepatic), location of intrahepatic metastasis (right, left or both lobes) and PLR (<154 vs. ≥154) were not significantly associated with OS (all p > 0.05). Multivariate analysis (Table 3) confirmed that serum LDH levels have prognostic value for OS (HR = 0.747; 95% CI: 0.552–1.012; p = 0.059).

association of serum ldh levels with markers of systemic inflammation CC BY

Systemic inflammatory response markers such as NLR, PLR and MLR serve as prognostic predictors of OS of patients with certain cancers, including pancreatic cancer. Further, hypoxia and systemic inflammation are associated with the advanced stages of pancreatic cancer, and serum LDH levels serve as an indirect marker of tumour hypoxia. Therefore, we evaluated the relationship between serum LDH levels and markers of systemic inflammation in both cohorts (Table 1). Tests for normality (Kolmogorov-Smirnov) indicated that the data represented a continuous parameter that did not meet the assumptions of the test for normality (p < 0.001), except for LMR in the training cohort (p = 0.2). Therefore, we used Spearman’s non-parametric test. In the training cohort, serum LDH levels positively correlated with NLR (r = 0.218, P = 0.01) and PLR (r = 0.155, P = 0.034) but negatively correlated with LMR (r = −0.157, P = 0.033). These results were further confirmed in the validation cohort of 225 cases. Thus, serum LDH levels positively correlated with NLR (r = 0.151, P = 0.0024) and PLR (r = 0.120, P = 0.036) but negatively correlated with LMR (r = −0.130, P = 0.026) (Fig. 3a,b). We concluded therefore that serum LDH levels were associated with the systemic inflammatory response in advanced pancreatic cancer after gemcitabine-based palliative chemotherapy.

discussion CC BY

Here we investigated the correlation between LDH levels and the prognosis of patients with advanced pancreatic cancer after they were administered gemcitabine-based palliative chemotherapy. The results show that high vs. low serum LDH levels significantly correlated with shorter OS. Further, serum LDH levels correlated positively with NLR and PLR and negatively with LMR. The oxidoreductase LDH, which converts pyruvate to lactate when oxygen is absent or in short supply, plays a crucial role in the metabolism of cancer cells25. Tetrameric human LDH comprises three different monomeric subunits LDH-A, LDH-B and LDH-C. LDH-A is overexpressed in hypoxic carcinomas as well as metastatic cancer cells, and its levels correlate with tumour viability26. In many tumour types, serum LDH levels serve as an indirect marker of tumour hypoxia, neo-angiogenesis and worse prognosis. PDACs are characterized by a dense desmoplastic stroma and a sparse vascularization that limit the availability of oxygen and nutrients. This tissue architecture induces severe hypoxic stress in tumour cells, which then become resistant to chemotherapy and develop increased invasiveness and metastatic potential27. Together, these factors led us to explore the relationship between LDH levels and the prognosis of patients with advanced pancreatic cancer. Increased LDH levels are present in patients with malignancies, including pancreatic cancer, and elevated LDH levels serve as a prognostic marker for numerous human malignancies1617181920212223. Further, LDH levels can guide the treatment of pancreatic cancer. For example, the inhibition by oxamic acid of LDH production by cancer cell lines potentiates the antiproliferative activity of tyrosine kinase inhibitors such as sorafenib28, Moreover, patients with high pretreatment levels of LDH may represent optimal candidates for inclusion in clinical trials that employ a multimodality treatment approach (e.g. transcatheter arterial chemoembolization and inhibitors of vascular endothelial growth factor signaling) to improve time to progression and OS29. Here we found a significant association between serum LDH levels and OS in 346 patients with advanced pancreatic cancer after they were administered gemcitabine-based palliative chemotherapy. Specifically, low serum LDH levels significantly correlated with longer OS. Only tumour stage and CA19-9 are widely used for predicting prognosis of patients with advanced pancreatic cancer30. CA19-9 is the most commonly used marker for pancreatic cancer to evaluate prognosis31. Evidence suggests that other serum tumour markers such as LDH may have prognostic relevance for patients with advanced pancreatic cancer3233. Several preclinical studies provide evidence that the pathogenesis and progression of advanced pancreatic cancer involves a hypoxic tumour microenvironment that is associated with high tumour volume, high tumour LDH expression, and frequent inflammation. Inflammation is the seventh hallmark of cancer that can affect a patient’s response to chemotherapeutic agents and survival. Evidence indicates a correlation between hypoxia and systemic inflammation. We show here that serum LDH levels correlated positively with NLR and PLR and negatively with LMR. The modified Glasgow prognostic score and C-reactive protein (CRP) levels are valuable and commonly used markers of systemic inflammation, which correlate with the prognosis of numerous human cancers, including pancreatic cancer. Further, interleukin-6 (IL-6) is an important predictor of OS. However, we were unable to evaluate the clinical significance of IL-6 levels here, because assays for CRP and IL-6 were not routinely performed in our clinic. Because of this limitation, we will evaluate the prognostic value of other common markers of systemic inflammation, including CRP and IL-6, in patients with pancreatic cancer and their correlation with serum LDH levels in future studies. In conclusion, we found that serum LDH had prognostic value and was associated with systemic inflammation in patients with advanced pancreatic cancer after they were administered gemcitabine-based palliative chemotherapy. To accurately determine the roles and clinical significance of candidate biomarkers for improving the management of patients with pancreatic cancer, large prospective studies are required that employ uniform disease staging and standardized cut-off levels of biomarkers.

patients’ characteristics and clinical features CC BY

All patients, their relatives or both provided written informed consent for their clinical and pathological information to be used for research and stored in the hospital’s database. The Ethics Committee of Fudan University Shanghai Cancer Center (FUSCC) approved the methods and experimental protocols used in the present study, which were performed in accordance with the ethical standards of our institutional research committee and the tenants of the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Methods (Ethics Number: 050432-4-1212B) were performed in accordance with the approved guidelines. We retrospectively recruited 364 patients with histologically or cytologically confirmed locally advanced or metastatic pancreatic adenocarcinoma who were treated at FUSCC between January 2010 and December 2013. The criteria for locally advanced disease included tumour invasion of the celiac trunk, superior mesenteric artery or both, corresponding to stage III pancreatic cancer according to the International Union Against Cancer TNM Classification of Malignant Tumors (Sixth Edition). Dynamic Contrast-enhanced abdominal computed tomography (CT), magnetic resonance imaging (MRI) and MR-cholangiopancreatography were used for determining the TNM stage.

serum ldh assay CC BY

Serum LDH was assayed during routine workups to exclude liver-function disorders before diagnostic interventions or cancer treatment, and these levels were subjected to statistical analyses. LDH was measured in sera prepared from peripheral venous blood collected on the day of hospital admission using a method described by the International Federation of Clinical Chemistry and Laboratory Medicine. The median LDH concentration was 185 IU/mL and was selected as the cut-off between a high and low LDH levels.

laboratory measurements CC BY

Routine laboratory measurements, including counts of white blood cells, neutrophils, lymphocytes, monocytes and platelets as well as albumin concentrations were performed before diagnostic intervention or treatment. The neutrophil/lymphocyte ratio (NLR), platelet/lymphocyte ratio (PLR) and lymphocyte/monocyte ratio (LMR) were calculated. The median NLR, PLR, MLR values and albumin levels of 139 samples were 3.42, 154, 3.19 and 38.6 (g/L), respectively, and were selected as the cut-off values.

statistical analysis CC BY

Patients’ characteristics before diagnostic intervention or cancer treatment are reported using descriptive statistics. The χ2 or Fisher’s exact tests were used to compare quantitative outcomes between groups. Overall survival (OS) was defined as the interval between a definitive diagnosis and death or last follow-up. Survival curves were calculated using the Kaplan–Meier method, and the log-rank test was used to assess the significance of differences in survival. Univariate and multivariate analyses were performed using a Cox proportional hazards regression model as implemented in SPSS version 21.0 (IBM Corp., Armonk, NY, USA). Variables with p < 0.05 determined using univariate analysis were included as candidates in a multivariate Cox regression model (conditional backward selection). Hazard ratios (HRs) estimated from the Cox analysis are reported as relative risk (RR) with corresponding 95% confidence intervals (CIs). The correlation tests used a Spearman’s non-parametric test, and p < 0.05 was applied throughout to indicate a significant difference.

ethics statement CC BY

All patients and/or their relatives provided a written informed consent for their clinical and pathological information to be used for research and stored in the hospital database; this study, methods and experimental protocols were approved by the Ethical Committee of FUSCC (Fudan University Shanghai Cancer Center). All the procedures performed in our study were in accordance with the ethical standards of our institutional research committee and the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Methods(Ethics Number:050432-4-1212B) were carried out in accordance with the approved guidelines.

additional information CC BY

How to cite this article: Yu, S.-L. et al. Serum lactate dehydrogenase predicts prognosis and correlates with systemic inflammatory response in patients with advanced pancreatic cancer after gemcitabine-based chemotherapy. Sci. Rep. 7, 45194; doi: 10.1038/srep45194 (2017). Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.