Post-chemotherapy miR-146a expression and its prognostic potential in oral cancer patients

Purpose: To determine miR-146a expression level after chemotherapy in oral cancer patients, and its prognostic value. Methods: Quantitative reverse transcription polymerase chain reaction (qRT-PCR) was used for the determination of miR-146 expression levels. Based on the results, the analysis of the miR-146a expression in oral cancer patients was performed by drawing ROC curve to provide information on the prognostic value of miR-146a. The survival of the patients was monitored over a period of 5 years. The patients were categorized into highand low-expression groups, and multivariate Cox regression analysis method was adopted to provide a more comprehensive analysis of individual risk factors influencing the prognosis of oral cancer. Results: The miR-146a expression level in patients after chemotherapy was lower than that in patients before they received chemotherapy (p < 0.05). The specificity of using miR-146a to predict oral cancer was 76.83 %, the sensitivity 69.44 %, and the area between the curve and x-axis 0.78. In contrast, the survival level was significantly greater in high-expression patients (p < 0.05). Conclusion: The independent risk parameters for buccal carcinoma are drinking, smoking, chronic leukoplakia, and miR-146a.


INTRODUCTION
Oral cancer represents a generic term for malignant tumors taking place in the oral cavity. Its incidence has remained high in recent years, and it has become the sixth most prevalent cancer in the world. Its pathological characteristic is mostly squamous epithelial carcinoma, i.e., mucosal mutation [1]. Oral cancer has been on the rise among young people in recent years. In the year 2021, 300,000 cases of oral cancer have occurred around the world, with survival of 50-60% over a 5-year period, and a possibility of approximately 150,000 deaths. Oral cancer is distributed in about 32% of the oral mucosa, 22% of the tongue, 11% of the lower lip, 11% of the palate, 8% of the vestibule, 5% of the alveoli, 5% of the floor of the mouth, and 3% of the gums [2]. At present, the clinical treatment of oral cancer is the same as that of many other types of cancer, and it involves mainly surgery and chemotherapy. In the study carried out by Ketabat [3], it was reported that although surgical resection could be performed in the treatment of oral cancer, surgical resection caused permanent disfigurement, physical weakness, and severe dysfunction. Chemotherapy and radiotherapy produce obvious toxicity. These two treatment methods seriously impact the health and quality of life of cancer patients. Significant advancements have been achieved in the treatment of oral cancer with chemotherapy and radiotherapy, but the prognosis of the disease has not improved significantly over the past five decades. Therefore, the development of new treatment approach is of great importance both for treatment of oral cancer and for prolonging the lives of the patients. Besides, timely detection of early-stage tumors remains a priority and a major challenge for the medical community.
MicroRNAs (MiRNAs) act as endogenous molecules that adjust and control the expression of target genes. Moreover, they are involved in the regulation of a range of pathophysiological functions [4]. Many miRNAs are abnormally expressed during the progression of cancers [5][6][7]. Previous studies have found that miR-146a is closely related to the progression of hepatocellular carcinoma, non-small cell lung cancer and gastric cancer [8][9][10]. But there is no clear evidence whether it has a connection with laryngeal carcinoma. Therefore, the expression of miR-146a in oral cancer was determined. Its value in predicting and prognosis of oral cancer patients was studied, in order to explore new potential diagnosis and treatment targets for clinic.

Patients
A total of 82 oral cancer patients who were diagnosed by doctors at Urumqi Stomatological Hospital and received routine treatment from March 2015 to February 2018 (a period of three years) were chosen as the research group in this study. In addition, healthy subjects were selected. A control group containing 72 healthy persons was used for comparative study. The study group had a male-to-female ratio of 51:31, with a mean age of 42.4 ± 11.2 years. The control group had a male-to-female ratio of 40:32, with mean age of 41.7 ± 10.9 years. This study was approved by the ethical committee of Urumqi Stomatological Hospital, and followed the guidelines of Declaration of Helsinki [11].

Inclusion criteria
The ages of the patients involved in the study ranged from 30 to 70 years. They were all diagnosed and treated at Urumqi Stomatological Hospital, and they fully met the diagnostic criteria for oral cancer, with complete case data. All patients fully agreed and actively cooperate with all arrangements made by the hospital's medical staff.

Exclusion criteria
Patients who died during treatment, and those who had vital organ failure and damage or cardiovascular and cerebrovascular diseases, as well as patients suffering from other cancers, were excluded. Patients with diabetes, fatty liver, stroke, heart disease, comorbidities with other autoimmune diseases; mental diseases, central nervous system diseases, speech disorders, and visual disorders, were also excluded. Moreover, pregnant patients and those with hearing disorders, taste disorders, and other diseases that may directly or indirectly affect the results in this research were excluded.

Processing of blood samples
Fasting venous blood was collected in the morning, preserved at 4 °C for half an hour, and the serum specimen was obtained after the blood was centrifuged for 10 min at 3000 rpm. The extraction and storage of the supernatant were conducted at -80°C.

Equipment and chemicals
Detailed information on the primer sequences is presented in Table 1.

Determination of mRNA expression
Total RNA was extracted from freshly-separated intravenous serum at a temperature of 8 o C using TRIzol kit strictly in line with the kit protocol. The amount RNA and its level of impurity were measured employing UV spectrophotometry. The RNA absorbance was between 1.8 and 2.1. If the extracted RNA failed to satisfy this criterion, it extraction was done for a second time. After extraction, the integrity of the RNA was examined via electrophoresis on a denaturing agarose gel at a concentration of 1 %. The reaction system for total RNA was reconfigured. The researchers correctly followed the instructions of the miRNA reverse transcription kit in the reverse-transcription of the miRNA under specific conditions, in order to synthesize its corresponding cDNA which was preserved at -20 °C for fresh-keeping. The fluorescent RT-PCR was carried out employing an ABI StepOne Plus fluorescent quantitative PCR instrument. The reaction volume was 12.33 uL, and DEPC water was added to make it up to 20 uL. The PCR conditions were: 95 ℃ for 5 min, 95 ℃ (45 s), 60 ℃ (60 s), and 72 ℃ (45 s). The internal reference employed for the reaction was U6. The experiment was conducted at least thrice, and the results of the relative mRNA expressions were calculated with 2 -△△ct procedure.

Patient follow-up
A three-year detailed follow-up on the health of the two groups was conducted through telephone calls and outpatient medical records, on the 4 th , 8 th and 12 th months of each year.

Determination of main outcome predictors
The miR-146a levels in subjects suffering from oral cancer were monitored, and its predictive potential was determined.

Secondary observation indicators
The subjects were separately assigned to highand low-miR-146a categories, among whom the grouping condition was the miR-146a expression levels (the grouping did not consider age, gender, location and other undifferentiated conditions). After grouping, the patients were closely observed and their 5-year survival and physical health status were accurately recorded. Thereafter, independent factors that affected the patients' prognosis were collated and analyzed using multivariate Cox regression analysis.

Statistical analysis
The data collected by the researchers were statistically analyzed using the SPSS20.0 method. GraphPad 7 was utilized to plot the graphs, while KS test was employed to investigate the distribution of the data. Normallydistributed data are presented as mean ± standard deviation (mean ± SD). Paired t-test was used for comparison within groups. Count data are expressed as number and percentage [n (%)], and they were analyzed with chi-square test. Independent sample t-test was used for between-group comparisons; paired t-test was employed for within-group comparisons, while the Kaplan-Meier approach was utilized to assess the 5-year survival of the compared patients. The predictive potential of miR-146a in buccal carcinoma was determined using ROC curves. Multi-factor Cox regression analysis was performed to identify risk factors impacting prognosis. Differences were considered significant when p < 0.05.

Patient profile
There were no statistical differences between the clinical data of the study and control groups in terms of gender, BMI, age, marital status, nationality, place of birth and residential area. This information was comparable between the two groups (p> 0.05), However, the differences in bad habits such as drinking, smoking, eating of betel nut, and duration of bad habits were all statistically significant (p <0.05) ( Table 2).

Expression levels of miR-146a in the sera of oral cancer patients
The results indicated that miR-146a expression level in serum before chemotherapy was 6.73 ± 1.02, while the serum miR-146a expression level after chemotherapy was 1.29 ± 0.84 (p <0.05) ( Figure 1)

Potential of miR-146a for prediction of oral cancer
Analysis of the ROC curve in Figure 2 led to the conclusion that miR-146a had a specificity of 76.83% and a sensitivity of 69.44% in forecasting oral cancer when the cut-off value was 6.05, and the area of the curve enclosing the x-axis was 0.78. These results are presented in Table 3 as well as Figure 2.

Five-year survival level of patients
In this research, the results were categorized into 2 groups based on the median value of the expression level of miR-146a. There were 41   Figure 3).

Univariate Cox analysis results
Based on the median miR-146a levels, the results were categorized into 2 groups: 41 cases of miR-146a high-expression group and 41 cases of low-expression group. The miR-146a value in high-expression patients was ≥ 1.293, while the miR-146a value in low-expression patients was <1.293. The clinical data of the two groups were analyzed based on single factors. The results showed that age, gender, area of residence, and marital status were comparable. However, differences in smoking, drinking, erythema, leukoplakia, betel nut chewing, and miR-146a were all statistically significant (p < 0.05) ( Table  4).

Multivariate Cox analysis results
Results from multivariate Cox regression analysis showed clearly that independent risk factors for patient prognosis did not include differences in age and gender, while chronic diseases such as leukoplakia and erythema, bad habits, drinking and smoking, and miR-146aspecific manifestations were independent risk factors for patient prognosis (Table 5).

DISCUSSION
Smoking and drinking are considered to be two of the chief culprits in oral cancer. Alcohol is an organic solvent which promotes the dissolution and absorption of carcinogens in tobacco. In recent years, head and neck cancer has become the seventh most common malignancy worldwide. Annually, more than 600,000 new cases are diagnosed. At the same time, the incidence of oral cancer is on the rise, and oral cancer accounts for half of the head and neck cancer cases. In addition, according to surveys, there is uneven distribution of head and neck   [12]. In addition, the incidence of head and neck cancer is also associated with gender and age. The incidence in men is higher than that in women, and it increases with age [13]. Due to its high incidence and aggressiveness in recent years, the identification of a key biomarker for oral cancer is particularly important for disease detection and comprehensive treatment.
It is known that miRNA is able to degrade target genes and inhibit their translation, thereby accomplishing post-transcriptional gene silencing. Studies have shown that miRNAs [14] affect and regulate almost 30% of the protein encoded in the body. They inhibit or promote tumor development by adjusting target genes. A study has revealed that miR-146a exerts antitumor effect [15]. However, there is no specific research showing whether miR-146a has antitumor effect in oral cancer.
This research specifically compared the basic clinical information of oral cancer patients and discovered differences in unhealthy habits such as smoking, drinking, and eating betel nut. Gupta et al believed that oral cancer should be a disease of multifactorial origin, i.e., risk factors and effects vary considerably in different populations [16].
Currently known risk factors include various forms of smoking, betel nut chewing, alcohol abuse and underlying chronic diseases such as chronic erythema and leukoplakia. These known risk factors are generally consistent with the results of this research. In this study, it was observed that the serum miR-146a was significantly lower in healthy individuals than in patients with oral cancer. After chemotherapy, a significant reduction in miR-146a could be seen. This suggests that miR-146a may be a potential diagnostic and therapeutic target for laryngeal cancer. The ROC curve and the coordinate xaxis constituted an area of 0.780. This value represents higher specificity and sensitivity, and can be used as a clinical prediction and diagnostic index for oral cancer.
The study has demonstrated that 5-year survival of patients who expressed high miR-146 levels (46.34 %) was significantly better than that of miR-146a low-expression group (65.85 %). These results imply that miR-146a could be employed as a predictor of survival. Based on Cox multivariate analysis, it can be inferred that smoking, drinking, chewing betel lang, chronic leukoplakia, and miR-146a are independent risk factors that affect the prognosis of patients.
In addition, according to the studies conducted by Osazuwa-Peters and other researchers, it is evident that at least 75 % of head and neck cancers are inextricably linked to the consumption of tobacco and alcohol. It is estimated that the frequency of buccal cancer in smokers and alcoholics is 30 -48 % higher than that in the normal population [17].
This preliminary study has demonstrated the clinical potential of miR-146a in prediction of oral cancer. However, there are still certain limitations in the study. This research did not conduct basic cell experiments and rat experiments. Therefore, more in-depth experimental analysis is needed to further confirm the results of this research and provide more valuable clinical information. In addition, quitting unhealthy habits such as smoking, alcoholism, and chewing betel lang can effectively reduce the incidence of oral cancer.

CONCLUSION
This study has shown that miR-146a has a high expression level in oral carcinoma patients. Habits such as drinking, smoking, chronic leukoplakia and erythema are independent risk factors for oral cancer. This finding provides valuable reference information for clinical practice.