Findings

Background

Pancreatic cancer (PaC) accounts for 2.2% of all cancers and is the fourth most common cause of cancer related deaths [1]. Because of the lack of early symptoms, aggressive growth and early dissemination, most patients are diagnosed at late stages with advanced distant metastases, which make their disease surgically inoperable. Moreover, this type of cancer is known to be highly resistant to radiotherapy and chemotherapy and has one of the lowest five-year survival rate among solid cancers [2].

MicroRNAs (miRNAs) are small, non-coding RNAs that post-transcriptionally regulate gene expression. They can function as both, oncogenes or tumor suppressors, and play an important role in the regulation of many biological processes [3]. They can also serve as new diagnostic and prognostic biomarkers as well as promising therapeutic targets [4]. A growing number of studies have proved that miRNAs regulate the variety of processes involved in the development, progression and chemoresistance of PaC. In this study, expression of 4 miRNAs (miR-21, miR-217, miR-198 and miR-34a) that have been previously described to be deregulated in PaC [5-7] was quantified in tumor tissues of patients with pancreatic ductal adenocarcinoma (PDAC), healthy tissues and tissues of chronic pancreatitis (CP) with the aim to validate their diagnostic and prognostic potential.

Materials and methods

Patients and tissue samples

Specimens from 74 patients with PDAC (37 males, 37 females) and 18 patients with CP (12 males, 6 females) who had undergone resection from August 2001 through April 2012 at the Department of Surgery (Faculty Hospital Brno, Czech Republic) were used. Moreover, control pancreatic tissue samples without signs of inflammation or dysplastic changes from 9 patients were included. All subjects were of the same ethnicity (European descent). The ages of patients ranged between 30 and 79 years with a median of 60.5 years. Written informed consent was obtained from all patients and the study has been approved by the local Ethical Board.

Extraction of miRNAs

Isolation of total RNA enriched for small RNAs was performed using formalin-fixed paraffin-embedded samples with more than 90% of cancerous, inflammatory or normal tissue. All samples were deparaffinized, treated with DNAse I, proteinase K and RNA extraction was undertaken using mirVana miRNA Isolation Kit (Ambion Inc, Austin, TX, USA) according to the manufacturer's instructions. Concentration and purity of RNA were determined spectrophotometrically by measuring its optical density (A260/280 > 2.0, A260/230 > 1.8) using Nanodrop ND-1000 (Thermo Fisher Scientific, Waltham, MA, USA).

Real-time quantification of miRNAs

Complementary DNA was synthesized from total RNA according to the TaqMan MicroRNA Assay protocol (Applied Biosystems, Foster City, CA, USA) using T100™ Thermal Cycler (Bio-Rad, Hercules, CA, USA). Real-Time PCR was performed according to the standard protocol using the TaqMan MicroRNA Assay kit and the Applied Biosystems 7500 Sequence Detection System (both Applied Biosystems, Foster City, CA, USA).

Data normalization and statistical analysis

The threshold cycle data were calculated by SDS 2.0.1 software (Applied Biosystems, Foster City, CA, USA). All real-time PCR reactions were run in triplicates. The average expression levels of all measured miRNAs were normalized using miR-1233 (Assay No. 002768; Applied Biosystems, Foster City, CA, USA) and subsequently analyzed by the 2-ΔCt method. Statistical differences between the levels of analyzed miRNAs were evaluated by non-parametric Mann-Whitney U-test and Kruskal-Wallis test. Survival analyses were carried out using the log-rank test and Kaplan-Meier plots approach. All calculations were performed using GraphPad Prism version 5.00 (GraphPad Software, San Diego, CA, USA). P-values of less than 0.05 were considered statistically significant.

Results

MiRNAs have potential to differentiate PDAC from CP and normal pancreas

To evaluate the diagnostic potential of 4 selected miRNAs (miR-21, miR-34a, miR-198 and miR-217), FFPE samples of 74 PDAC patients, 18 CP patients and 9 non-tumoral control pancreas were examined by qRT-PCR. Moreover, RNU48, RNU6B, RNU44, miR-1233 and miR-1260 were analyzed as potential reference genes. Using geNorm [8] and NormFinder [9], miR-1233 was selected to be the most appropriate normalization control. Using Man-Whitney U-test, significantly higher levels of miR-198 (P < 0.0001), miR-21 (P = 0.0018), miR-34a (P = 0.0111) and significantly lower levels of miR-217 (P = 0.0001) were observed in PDAC samples compared to healthy pancreatic tissue. Moreover, all analyzed miRNAs had the potential to differentiate PDAC from CP tissues. Whereas miR-198 (P = 0.0003), miR-21 (P < 0.0001) and miR-34a (P < 0.0001) were significantly up-regulated, miR-217 (P = 0.0307) was significantly down-regulated in PDAC samples compared to CP samples (Figure 1A-D). Subsequently, using ROC analysis, miR-21 was shown to have the highest capacity to distinguish between these two groups of samples with the sensitivity 93%, specificity 72% and AUC = 0.9227. The results of ROC analyses are summarized in Table 1.

Figure 1
figure 1

Normalized expression of analyzed miRNAs in control pancreatic tissue, CP tissue and PDAC tissue. A) MiR-198 is increased in PDAC compared to CP tissue (P = 0.0003; CP) and healthy pancreas (P < 0.0001; HP). B) MiR-217 is down-regulated in PDAC compared to CP tissue (P = 0.0307; CP) and healthy pancreas (P = 0.0001; HP). C) MiR-21 is up-regulated in PDAC compared to CP tissue (P < 0.0001; CP) and healthy pancreas (P = 0.0018; HP). D) MiR-34a is up-regulated in PDAC compared to CP tissue (P < 0.0001; CP) and healthy pancreas (P = 0.0111; HP). *P < 0.05; **P < 0.001; ***P < 0.0001.

Full size image
Table 1 Results of the ROC curve analysis to differentiate between pancreatic ductal adenocarcinoma and chronic pancreatitis
Full size table

MiR-21 and miR-198 can serve as new prognostic biomarkers for PDAC

To evaluate the prognostic function of analyzed miRNAs, Kaplan-Meier survival curves have been generated and compared by log-rank analysis. We have proven that patients with high levels of miR-21 and miR-198 have shorter both, disease free survival (DFS; P = 0.0011 for miR-21; P = 0.0001 for miR-198; Figure 2A-B) and also overall survival (OS; P = 0.0427 for miR-21; P = 0.0097 for miR-198; Figure 2D-E). Moreover, when the expression levels of these two miRNAs were combined, it has been shown that patients with low levels of both miR-21 and miR-198 have significantly higher DFS (18.2 months vs. 8 months; Figure 2C) and OS (23.7 months vs. 14.9 months; Figure 2F) compared to the patients with high levels of miR-21 and/or miR-198. Expression levels of miR-217 and miR-34a were not correlated with DFS or OS of PDAC patients (Table 2).

Figure 2
figure 2

Kaplan-Meier analysis of DFS and OS based on the expression of miRNAs. A) PDAC patients with high level of miR-198 have shorter DFS compared to patients with low level of this miRNA (P = 0.0001). B) PDAC patients with high level of miR-21 have shorter DFS compared to patients with low level of this miRNA (P = 0.0011). C) Combined analysis of miR-21 and miR-198 – PDAC patients with low levels of both miR-21 and miR-198 have longer DFS compared to patients with high levels of miR-21 and/or miR-198 (P < 0.0001). D) PDAC patients with high level of miR-198 have shorter OS compared to patients with low level of this miRNA (P = 0.0097). E) PDAC patients with high level of miR-21 have shorter OS compared to patients with low level of this miRNA (P = 0.0427). F) Combined analysis of miR-21 and miR-198 – PDAC patients with low levels of both miR-21 and miR-198 have longer OS compared to patients with high levels of miR-21 and/or miR-198 (P = 0.0015).

Full size image
Table 2 The detail results of Kaplan-Meier survival analysis
Full size table

Discussion

Successful management and treatment of PDAC patients remains one of the key challenges in clinical oncology. Although the early stages of the disease can be treated surgically, most patients are diagnosed at advanced stages, when surgical resection is not possible. Moreover, differential diagnosis of pancreatic lesions is challenging [10]. Therefore, there is an urgent need for novel diagnostic biomarkers that would enable precise differential diagnosis of pancreatic lesions. In addition, molecular biomarkers that could serve as prognostic factors would be very valuable.

MiRNAs have been described to be deregulated in a variety of solid cancers, including PDAC [11-13]. In this study, the utility of miR-21, miR-34a, miR-198 and miR-217 as novel diagnostic and prognostic biomarkers of PDAC was evaluated. Consistently with the previous data, significantly increased levels of miR-21 and miR-198 [7,14,15] and decreased levels of miR-217 [7,15,16] have been observed in PDAC tissue. Despite the fact that miR-34a is generally described as an important tumor suppressor [12,17], the expression of this miRNA has been significantly higher in our PDAC samples compared to healthy tissue. Therefore, it seems that this miRNA may have dual functioning as both oncogene and tumor suppressor, depending on the cellular and tumor microenvironment [18].

Importantly, all analyzed miRNAs had a high potential to differentiate CP from PDAC tissue, therefore, they might be involved in early events of pancreatic carcinogenesis. Habbe et al. [19] showed that miR-21 is highly expressed in early non-invasive intraductal papillary mucinous neoplasms. Further, using in situ hybridization increased miR-21 expression was found in 79% of PaCs; however, only 8% of benign pancreas and 27% of CP expressed this miRNA suggesting its important role in the development of PaC [20]. MiR-217 was described to play a crucial role in regulation of acinar-to-ductal metaplasia [21], in addition, this miRNA is deregulated not only in PDAC but also in its precursor lesions, compared to non-neoplastic pancreatic tissues [15]. The function of miR-34a and miR-198 in early development of PaC has not been described till now, nevertheless, it seems that miR-21, miR-34a, miR-198 and miR-217 could be used as tumor markers to distinguish PDAC and its precursors from a benign lesions.

Given the dismal prognosis of PaC, second aim of this study was to identify miRNAs with the potential to differentiate between patients with good (DFS ≥ 12 months, OS ≥ 18 months) and poor (DFS < 12 months, OS < 18 months) prognosis. We proved that high levels of miR-21 and/or miR-198 significantly correlate with poor prognosis. Concerning miR-21, several studies have been previously published demonstrating prognostic function of this miRNA in PaC [12,20]. Moreover, high levels of miR-21 were associated with a poor response to gemcitabine and its levels were increased after the exposure to this drug [22,23]. Concerning miR-198, there are two contradictory reports analyzing the prognostic function of this miRNA in PDAC. Whereas Marin-Müller et al. [24] described high levels of this miRNA to be associated with good prognosis, Schultz et al. [7] observed correlation between over-expression of miR-198 and poor prognosis. Taken together, our data indicate that miR-21 and miR-198 could be used as potential prognostic biomarkers in PDAC patients. Importantly, the value of clinical utility of these miRNAs could be enhanced by measurement prior to resection in PDAC tissue obtained by endoscopic ultrasound-guided fine needle aspirates [25] with the aim to improve the clinical management of borderline resectable cases and identification the patients who will benefit most from the surgical resection.