Pancreatobiliary tract cancer is a highly fatal cancer. Detection of pancreatobiliary tract cancer is difficult because it lacks typical clinical symptoms and because of its anatomical location. Biomarker discovery is therefore important to detect pancreatobiliary tract cancer in its early stage. A study demonstrated that expression levels of miR-1246, miR-3976, miR-4306, and miR-4644 in serum exosomes were higher in pancreatic cancer patients than these levels in healthy control participants. Supposing that microRNA (miRNA) expression profiles in saliva are similar to those in serum, four miRNAs (miR-1246, miR-3976, miR-4306, and miR-4644) in salivary exosomes may also be useful for detection of pancreatobiliary tract cancer. In this study, it was examined whether these miRNAs could be used as biomarkers for pancreatobiliary tract cancer. Twelve pancreatobiliary tract cancer patients and 13 healthy control participants were analyzed as a cancer and a control group, respectively. Unstimulated whole saliva was collected, salivary exosomes were isolated, and total RNA was extracted. Using quantitative real-time PCR (RT-qPCR), the relative expression ratios of miR-1246 and miR-4644 were significantly higher in the cancer group than these ratios in the control group. Receiver operating characteristic (ROC) curves were constructed to analyze the discrimination power of these miRNAs. For miR-1246, the results yielded an area under the curve (AUC) of 0.814 (P=0.008). For miR-4644, the results yielded an AUC of 0.763 (P=0.026). For the combination of miR-1246 and miR-4644, the results yielded an increased AUC of 0.833 (P=0.005). This pilot study suggests that miR-1246 and miR-4644 in salivary exosomes could be candidate biomarkers for pancreatobiliary tract cancer.
Pancreatobiliary tract cancer is a term used to describe malignant carcinoma in pancreatic, gallbladder, and extrahepatic bile ducts. Pancreatobiliary tract cancer is a highly fatal cancer (
Recently, microRNAs (miRNAs) have been reported as potential biomarkers for various types of cancers including pancreatobiliary tract cancer (
Recent studies have also focused on salivary miRNAs as biomarkers for various diseases (
To the best of our knowledge, there is no literature regarding the relationship between miRNAs in salivary exosomes and pancreatobiliary tract cancer. On the other hand, one report is available regarding the relationship between miRNAs in blood exosomes and pancreatic cancer (
The study design was a case-control study. It was not possible to estimate the sample size preliminarily, because there was no prior information on which to base a sample size (
Twelve patients (6 males and 6 females) with pancreatobiliary tract cancers were referred to the Department of Preventive Dentistry, Okayama University Hospital for saliva collection in the morning before the onset of cancer therapy from July 2013 to July 2014.
As a control group, 13 healthy participants were recruited at the Department of Preventive Dentistry, Okayama University Hospital from August 2014 to December 2014 for saliva collection in the morning. Inclusion criteria for healthy control participants were >50 years of age and no history of any cancer. To avoid the effect of systemic conditions on circulating miRNA expressions, the exclusion criteria for healthy control participants were as follows: diabetes (
Measurements were performed before the onset of cancer therapy. Medical charts were reviewed to obtain information about cancer and the body mass index (BMI). A personal interview was performed to gather information about smoking habits (pack-years). Serum levels of hemoglobin A1c (HbA1c), C-reactive protein (CRP), albumin, carcinoembryonic antigen (CEA), and CA19-9 were evaluated. Concentrations of HbA1c, CRP, and albumin were measured by the high-performance liquid chromatography method, the latex agglutination method, and the bromocresol green albumin method, respectively. Concentrations of CEA and CA19-9 were measured by electrochemiluminescence immunoassays.
Unstimulated whole saliva was collected as reported previously with minor modification (
Exosomes were isolated from saliva samples (0.5–1.0 ml) using Total Exosome Isolation Reagent (Invitrogen, Carlsbad, CA, USA), in accordance with the manufacturer's instructions (
Total Exosome RNA and Protein Isolation Kits (Invitrogen) were used for extraction of total RNA from exosome samples. Total RNA was extracted in accordance with the manufacturer's instructions. After extraction of total RNA from salivary exosomes, their quality was confirmed using the Agilent 2100 Bioanalyzer and the Agilent RNA 6000 Pico Kit (both from Agilent Technologies, Santa Clara, CA, USA).
To compare miRNA expression between the control and cancer group, TaqMan RT-qPCR assays were performed. TaqMan MicroRNA Assays (life Technologies, Carlsbad, CA, USA) were used for the RT-qPCR analyses performed on the Mx3000P Real-Time QPCR System (Agilent Technologies) according to the manufacturer's instructions (
To evaluate the discrimination power of candidate miRNA biomarkers for pancreatobiliary tract cancer, receiver operating characteristic (ROC) curves (
Characteristics of the control and cancer group are represented as continuous variables [age, BMI, smoking habit (pack-years), HbA1c, CRP, albumin, CEA, and CA19-9] and categorical variables (gender, cancer site, and cancer stage). The Mann-Whitney U test, the Chi-square test, and Fisher's exact test were used to assess significant differences in clinical variables between the two groups, as appropriate. The Mann-Whitney U test was used to compare the relative expression ratios of each miRNA of the control and cancer group. To calculate the P-value for the AUC, a non-parametric test for AUC=0.5 was performed. Two-sided P<0.05 values were considered to represent significant differences. To assess the correlations of variables in the cancer group, Spearman's rank correlation coefficient and its P-value were calculated for age (years), HbA1c (NGSP) (%), CRP (mg/dl), albumin (g/dl), CEA (ng/ml), CA19-9 (U/ml), relative expression ratios of the miRNAs, and smoking (pack-years) as continuous variables, and cancer stage (I–III, IVa, and IVb) as a categorical variable. These statistical analyses were performed using SPSS software version 20 (SPSS, Inc., Chicago, Il, USA).
After the RT-qPCR assays, miR-3976 was excluded from the following analyses because 24 of 25 (96.0%) samples had high (>40) raw Ct-values. The number of samples having high (>40) raw Ct-values of expression levels of miR-1246, miR-4306, miR-4644, and U6 snRNA were 0, 2, 1 and 0, respectively. Therefore, miR-1246, miR-4306, and miR-4644 were considered as candidate miRNAs. For the control group, the median (25th and 75th percentile) values of log2-transformed relative expression of miR-1246, miR-4306, and miR-4644 were 11.6 (10.9, 12.6), −5.4 (−6.8, −3.9), and −6.5 (−7.6, −4.9), respectively. For the cancer group, those of miR-1246, miR-4306, and miR-4644 were 14.7 (12.6, 16.2), −5.0 (−8.1, −2.9), and −4.1 (−5.7, −2.2), respectively. The Mann-Whitney U test showed that the relative expression ratios of miR-1246 and miR-4644 were significantly higher in the cancer than these ratios in the control group (
To evaluate the discrimination power of miR-1246 and miR-4644 for pancreatobiliary tract cancer, ROC curves were constructed (
In the cancer group, there were significant correlations between CRP and cancer stage (r=0.822), between albumin and HbA1c (r=0.582), between albumin and CRP (r=−0.711), between CA19-9 and miR-1246 (r=0.818), between miR-1246 and miR-4644 (r=0.671), between smoking and HbA1c (r=0.579), and between smoking and albumin (r=0.647) (
This study was performed to clarify whether miR-1246, miR-3976, miR-4306, and miR-4644 levels in salivary exosomes were useful as potential biomarkers for pancreatobiliary tract cancer. Among these four miRNAs, two miRNAs (miR-1246 and miR-4644) in salivary exosomes showed significantly higher expression in pancreatobiliary tract cancer patients than that noted in healthy control participants. In addition, in the ROC curve analysis, the AUCs of both miR-1246 and miR-4644 were >0.7, indicating fair discriminatory power (
Compared to other body fluids including blood, saliva may be immediately exposed to the outside environment and confounded by a wide variety of environmental factors (
In the present findings, the combination of miR-1246 and miR-4644 in salivary exosomes increased the sensitivity for pancreatobiliary tract cancer. This is consistent with previous results, indicating that concomitant pancreatic cancer-initiating cell and miRNA marker expression strengthened the sensitivity for pancreatic cancer (
Serum CA19-9 is one of the most studied and validated serum biomarkers for pancreatobiliary tract cancer (
miR-1246 has been found to be aberrantly expressed in pancreatic cancer tissue (
Studies have investigated the relationship between salivary miRNA profiles and pancreatobiliary tract cancer. For instance, a clinical study showed that miR-21, miR-23a, miR-23b, and miR-29c were significantly upregulated in saliva of pancreatic cancer patients compared to controls (
The limitations of this study are as follows. First, external validity was limited because all participants were recruited at the Okayama University Hospital. In addition, the cancer patients were mainly diagnosed as having advanced-stage disease when they participated in this study. Further studies including patients with chronic pancreatitis and early stage of pancreatobiliary tract cancer are necessary to improve the validity of miR-1246 and miR-4644 in salivary exosomes as screening markers of pancreatobiliary tract cancer. Furthermore, overall screening of miRNAs in salivary exosomes using a microarray will be necessary to discover a new biomarker in pancreatobiliary tract cancer patients.
In conclusion, the present results demonstrated that miR-1246 and miR-4644 in salivary exosomes could be useful biomarkers for identification of patients with pancreatobiliary tract cancer. We hope that clinical use of simple and non-invasive salivary tests will contribute to the screening of pancreatobiliary tract cancer patients and improve their survival.
This study was supported by a Grant-in-Aid for Scientific Research (26670904) from the Ministry of Education, Culture, Sports, Science and Technology, Tokyo, Japan.
area under the curve
body mass index
carbohydrate antigen 19-9
carcinoembryonic antigen
C-reactive protein
hemoglobin A1c
receiver operating characteristic
reverse transcription
microRNAs
threshold cycle
Differential relative expression ratio of miRNAs obtained in the control (n=13) and cancer (n=12) group. U6 snRNA was used as reference. Log2-transformed relative expression ratio of (A) miR-1246, (B) miR-4306 and (C) miR-4644 for each sample is represented as plots. The horizontal lines in the boxes represent the 25th, 50th and 75th percentile values of log2-transformed relative expression ratio of each miRNA, respectively. The P-value was calculated using Mann-Whitney U test. miRNAs, microRNAs.
ROC curve analysis for the discriminatory power of salivary exosomal miRNAs for the control and cancer group. ROC curves for (A) miR-1246, (B) miR-4644 and (C) combination of miR-1246 and miR-4644 are represented. Log2-transformed relative expression ratio using U6 snRNA as reference was used for each miRNA. P-value was calculated using non-parametric test for AUC=0.5. Cut-off, sensitivity, specificity, PPV, and NPV were optimized using the maximum Youden index. Black circle represents the plot with these optimized values. ROC, receiver operating characteristic; miRNAs, microRNAs; AUC, area under the curve; PPV, positive predictive value; NPV, negative predictive value.
Characteristics of the control and cancer group.
Characteristics | Control |
Cancer |
P-value |
---|---|---|---|
Age (years) | |||
Median (range) | 66 (53–83) | 65 (45–84) | 0.728 |
Gender n (%) | |||
Male | 6 (46.2) | 6 (50.0) | 0.848 |
Female | 7 (53.8) | 6 (50.0) | |
Smoking, n (%) | |||
Current | 0 (0.0) | 5 (41.7) | 0.015 |
Cancer site, n (%) | |||
Bile duct cancer | 2 (16.7) | ||
Gallbladder cancer | 1 (8.3) | ||
Pancreatic cancer | 9 (75.0) | ||
T category |
|||
T1 | 0 (0.0) | ||
T2 | 1 (8.3) | ||
T3 | 2 (16.7) | ||
T4 | 7 (58.3) | ||
Unknown | 2 (16.7) | ||
N category |
|||
N0 | 6 (50.0) | ||
N1 | 1 (8.3) | ||
N2 | 1 (8.3) | ||
N3 | 1 (8.3) | ||
Unknown | 3 (25.0) | ||
M category |
|||
M0 | 6 (50.0) | ||
M1 | 5 (41.7) | ||
Unknown | 1 (8.3) | ||
Cancer stage |
|||
I | 0 (0.0) | ||
II | 1 (8.3) | ||
III | 1 (8.3) | ||
IVa | 4 (33.3) | ||
IVb | 6 (50.0) | ||
HbA1c (NGSP) (%) | |||
Median (range) | 5.9 (4.9–8.1) | ||
CRP (mg/dl) | |||
Median (range) | 0.42 (0.06–5.47) | ||
Albumin (g/dl) | |||
Median (range) | 3.7 (2.9–4.5) | ||
CEA (ng/ml) | |||
Median (range) | 12 (1–1,379) | ||
CA19-9 (U/ml) | |||
Median (range) | 414 (1–38,864) |
Mann-Whitney U test, Chi-square test, or Fisher's exact test was used.
General Rules for the Study of Pancreatic Cancer (Japan Pancreas Society) were used. HbA1c, hemoglobin A1c; CRP, C-reactive protein; CEA, carcinoembryonic antigen; CA19-9, carbohydrate antigen 19-9.
Correlation of variables in the cancer group.
Variables | Values | Age |
Cancer stage |
HbA1c |
CRP |
Albumin |
CEA |
CA19-9 |
miR-1246 |
miR-4306 |
miR-4644 |
Smoking |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Age (years) | r |
1.000 | ||||||||||
P-value |
||||||||||||
Cancer stage | ||||||||||||
(I–III, IVa, and IVb) | r | −0.145 | 1.000 | |||||||||
P-value | 0.652 | |||||||||||
HbA1c (NGSP) (%) | r | 0.483 | 0.157 | 1.000 | ||||||||
P-value | 0.111 | 0.626 | ||||||||||
CRP (mg/dl) | r | −0.311 | 0.822 |
−0.209 | 1.000 | |||||||
P-value | 0.326 | 0.001 | 0.514 | |||||||||
Albumin (g/dl) | r | 0.315 | −0.192 | 0.582 |
−0.711 |
1.000 | ||||||
P-value | 0.318 | 0.550 | 0.047 | 0.009 | ||||||||
CEA (ng/ml) | r | −0.196 | 0.527 | 0.179 | 0.214 | 0.302 | 1.000 | |||||
P-value | 0.541 | 0.078 | 0.578 | 0.505 | 0.340 | |||||||
CA19-9 (U/ml) | r | −0.455 | 0.321 | 0.007 | 0.231 | 0.060 | 0.098 | 1.000 | ||||
P-value | 0.137 | 0.310 | 0.983 | 0.470 | 0.854 | 0.762 | ||||||
miR-1246 |
r | −0.291 | 0.122 | −0.112 | 0.116 | 0.004 | −0.091 | 0.818 |
1.000 | |||
P-value | 0.359 | 0.705 | 0.728 | 0.721 | 0.991 | 0.779 | 0.001 | |||||
miR-4306 |
r | −0.123 | 0.168 | −0.098 | 0.049 | 0.134 | 0.406 | 0.336 | 0.510 | 1.000 | ||
P-value | 0.704 | 0.602 | 0.761 | 0.880 | 0.679 | 0.191 | 0.286 | 0.090 | ||||
miR-4644 |
r | −0.077 | −0.252 | −0.249 | −0.315 | 0.250 | 0.063 | 0.406 | 0.671 |
0.790 |
1.000 | |
P-value | 0.812 | 0.430 | 0.435 | 0.318 | 0.434 | 0.846 | 0.191 | 0.017 | 0.002 | |||
Smoking (pack-years) | r | 0.133 | 0.043 | 0.579 |
−0.320 | 0.647 |
0.281 | 0.335 | 0.133 | 0.109 | 0.156 | 1.000 |
P-value | 0.681 | 0.896 | 0.048 | 0.310 | 0.023 | 0.377 | 0.287 | 0.681 | 0.736 | 0.628 |
Log2-transformed relative expression ratio of each microRNA using U6 snRNA as reference was analyzed.
Spearman's rank correlation coefficient and its P-value.
P<0.05,
P<0.01. HbA1c, hemoglobin A1c; CRP, C-reactive protein; CEA, carcinoembryonic antigen; CA19-9, carbohydrate antigen 19-9.