Serum circulating miRNA‐342‐3p as a potential diagnostic biomarker in parathyroid carcinomas: A pilot study

Abstract Objective To compare the serum miRNA expression profiles between patients with benign and malignant parathyroid tumours. Background Despite recent advances in molecular biology, a histological tissue biopsy is still the method of choice used to diagnose most cancers. The preoperative cytology is not an applicable method for diagnosis of parathyroid cancer (PC); therefore, huge interest exists in terms of finding alternative methodologies to seek specific cancer biomarkers. Design A retrospective cross‐sectional study. Patients and Methods Serum samples of patients with PC (n = 13) and parathyroid adenoma (PA) (n = 11), age (p = .999) and sex (p = .999) were matched and examined via the simultaneous comparative expression analysis of 754 microRNAs (miRNAs). The «TaqMan OpenArray Human MicroRNA Panel» (Applied Biosystems) was used to conduct real‐time PCRs using the «QuantStudio 12К Flex» station (Life Technologies). Results According to the results of a pilot study, significant changes in expression levels between the PC group and the PA group (control) (p < .05) were observed for 17 miRNAs. Among them, the downregulation of miRNA‐342‐3p met the Benjamini‐Hochberg adjustment criteria for multiple comparisons (p = .02). Conclusions Serum miRNA‐342‐3p could be a promising biomarker for PC to improve diagnosis and prognosis.

Russian Federation indicate an increasing incidence of PC, 2,7,8 which may be related to improvement in diagnostics or to an actual increase in incidence. The PC pathogenesis is currently unclear, but it is known that the cell division cycle 73 (CDC73) germline inactivation mutations can play important roles. CDC73, which is comprised of 17 exons and is located on chromosome 1q31.2, encodes the protein, parafibromin, which is associated in the polymerase-associated factor (Paf1) complex. Functions attributed to parafibromin include the downregulation of cyclin D1 expression and direct interaction with β-catenin resulting in the activation of transcription of target genes. More than 50% of CDC73 germline mutations associated with hyperparathyroidism-jaw tumour syndrome (HPT-JT), as well as about 15% of CDC73 germline mutations, were reported in patients with suspected sporadic PC. 9 The prognosis of PC is greatly influenced by surgeon performance, which emphasizes the importance of preoperative diagnosis. 10 In contrast to benign parathyroid tumours, which are effectively treated by selective parathyroidectomy, PCs require en bloc resection implying that this avoids capsule rupture. An insufficient volume and experience of the surgeon increase the risk of distant metastasis, the treatment of which has extremely limited efficiency. 11 Achieving microscopic, cancer-free margins improves disease-free survival. 12 Nevertheless, only 12.5% of patients with PC undergo radical surgery. 7 Despite recent advantages in molecular biology, a histological tissue biopsy is still the method of choice to diagnose most cancers, with the exception of PC. Cytology biopsies of the parathyroid gland were reported to be uninformative, even dangerous due to the possible diffusion of malignant cell seeding along the needle track. 13 Therefore, the majority of PC cases are diagnosed postoperatively by histological examination, 11 presenting an important opportunity to identify new preoperative PC markers to aid in the success of initial operations, thus improving the disease-free survival of patients.
Increasing evidence suggests that microRNAs (miRNAs) may be a new group of biomarkers for various cancers. MiRNAs are small, endogenous, noncoding RNAs about 23 nucleotides long. To date, around 2500 miRNA sequences have been identified in humans (miRBase database 20.0). 14 In theory, one miRNA can target hundreds of genes and one gene can be targeted by multiple miRNAs.
They are involved in almost all biological processes, such as cell proliferation, apoptosis and tumorigenesis, and play a crucial biological role in tumorigenesis and progression, as they affect cell proliferation, differentiation, adhesion, migration, invasion and apoptosis. 15 Moreover, alterations in miRNA expression have been observed in many diseases, including various types of cancer, indicating that miRNAs might be useful for cancer management. Unlike long molecules of RNA (eg mRNA), circulating miRNAs are highly stable in the majority of fluids. In recent years, much attention has been paid to the detection of such biomarkers in body fluids. 16 An important feature of miRNAs is their ability to circulate within the blood, which makes it possible to use them as high-quality biomarkers of human cancer. 17,18 Several studies analysed miRNA expression in parathyroid tissue, including PC, 19 infections, exacerbation of chronic disease during the last month, pregnancy and lactation, and intake of calcium-phosphorus-related medications. Biochemical parameters of fasting blood (serum total calcium (reference interval (RI) 2.15-2.55 mmol/L), ionized calcium (RI 1.03-1.29 mmol/L), and creatinine (RI 63-110 mcmol/L)) were determined using the automatic biochemical analyser ARCHITECT c8000 (Abbott). Intact parathyroid hormone (iPTH, RI 15-65 pg/mL) was evaluated using the second-generation electrochemiluminescent analyser Cobas 6000 (Roche).

| MiRNA isolation from serum
The miRNAs were isolated from the serum samples using the mir-Vana PARISTM kit (Life Technologies) as per the total RNA isolation procedure for liquid samples according to the manufacturer's recommendations. Briefly, 300 µl of each serum was mixed with 300 µl of 2 × denaturing solution at room temperature, combined with 600 µl of acid-phenol: chloroform and mixed thoroughly with vortex. Then, the sample lysate was centrifuged at 15,000 g for 5 min to separate the mixture into aqueous and organic phases. The aqueous phase was transferred to a fresh tube, mixed with 750 µl of 96% ethanol and applied to a filter cartridge. The filter cartridge was washed once with 700 µl of miRNA wash solution and twice with 500 µl of wash solution 2/3. The RNA was eluted from the filter in 100 µl of preheated (95°C) elution solution and stored at −70°C.

| DNA sequencing
Genomic DNA was extracted from peripheral leukocytes using

| Statistical analysis
Statistical analysis was performed using the software packages STATISTICA 13 (StatSoft) and SPSS (IBM). The quantitative results were reported as the median and the 1st and 3rd quartiles [Q1; Q3]. Spearman's rank test was conducted to assess the correlation between miRNA and laboratory indicators. A ROC curve was constructed, and the AUC was calculated to evaluate the diagnostic value of the selected miRNA and it combination with serum calcium and iPTH levels. The cut-off point was selected so that the sum of the diagnostic sensitivity and diagnostic specificity was the maximum for it. Differences in the expression of miRNA and age between groups were evaluated using the Mann-Whitney test (U test) as well as differences in terms of quantitative characteristics. The frequencies of the signs were compared with each other using the chi-square test (χ 2 ). The Yates amendment was applied when it was necessary.
A p-value less than 0.05 was considered statistically significant. For multiple comparisons, the Bonferroni correction and Benjamini-Hochberg adjustment were applied. The confidence intervals of the frequencies are calculated by the Klopper Pearson method.

| Quantification of miRNA
Raw data files (.eds) were analysed using the ExpressionSuite v1.1 data analysis software (Life Technologies). Ct values were normalized using a global normalization algorithm as an option of the ExpressionSuite v1.1 which uses the mean expression value of all expressed miRNAs in each sample as a normalization factor for miRNA real-time quantitative PCR data. 24 Targets with amplification scores of <1 and with Cq confidence values of <0.8 were rejected during analysis. For comparison of the miRNA-level delta cycle relative threshold (ΔCRT) between the PC and PA samples, the differences were considered to be significant when p-values were less than 0.05 after Benjamini-Hochberg adjustment. 25 Note, that in data modelling with ExpressionSuite, the Benjamini-Hochberg algorithm was used as one of the options and the differences in the fold change of log2 <−1 or >1 were considered relevant.

| Clinical and laboratory characteristics of the study population
Twenty-four patients with PHPT were included in this study ( Table 1.

| Diagnostic value of miRNA-342-3p
Receiver operator characteristic (ROC) curve analysis was used to evaluate the diagnostic accuracy of the serum miRNA-342-3p level for the PC group. The ROC curve was constructed for the expres-

| DISCUSS ION
In previous studies, miRNA signatures were examined in parathyroid tissues. For instance, miRNA was analysed in the PC tissue samples compared with the normal parathyroid glands 19 [19][20][21] The inhibition of the miRNA-296-5p expression was detected in distant metastases, while the miRNAs C19MC and miRNA-372 were activated in the samples. 21  Recently, circular RNA profiles were described in the PC for the first time, suggesting that those also play a role in parathyroid tumorigenesis. 26  However, at present, the role of miRNA-342-3p in the PC is still not well understood. Interestingly, most of the deregulated miRNAs in the PC were repressed, including the revealed miRNA342-3p in this study, suggesting onco-suppressive roles in this disease.
Interestingly, noncoding RNAs (ncRNAs) could be also applied as potential therapeutics in the near future. The replacement of the tumour suppressor miRNA-34a by using MRX34, the synthetic product, exhibited promising initial results in the first phase I study in humans. 32 Since then, the usage of ncRNAs in cancer therapy is permanently improving. The data are available suggesting that ncRNAs could be implicated in predicting the chemo-and radio-resistance, as well as the therapeutic targets. 33 Interactions between the miR-NAs and Toll-like receptors have been reported underlining the implication of ncRNAs in the cancer immunology. 34 miRNA expression patterns were also analysed to better understand and direct the immune checkpoint inhibitor therapy. 35 The purpose of this study is not to compare the value of calcium,

| Benefits and limitations of the study
Our study is hypothesis setting, and thus, it will require further vali- In addition, the differences between the two groups could be related to other, non-cancer-related, differences, for example height of Ca or iPTH levels. Second, some miRNAs may have opposite expression patterns depending on the cancer type. Thus, the miRNA could be an additional tool for the preoperative diagnosis of the PC and to identify the groups with increased cancer risk (high serum Ca and iPTH levels). Despite these limitations, this approach has several benefits. All the included patients were clinically well-characterized.
All the samples were collected at the same time period using the same collection technique and the identical lot number of bloodcollection tubes. The exact reagent set was used for all analysed samples. Moreover, differences in the miRNA-342-3p expression have remained significant after Benjamini-Hochberg adjustments, which is a promising result for further investigation.

| CON CLUS ION
The main objective of this study was to compare the serum miRNA

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon request.