Detection of human papillomavirus infection in oral cancers reported at dental facility: assessing the utility of FFPE tissues

Incidence of human papillomavirus (HPV)-associated oral cancers is on the rise. However, epidemiological data of this subset of cancers are limited. Dental hospital poses a unique advantage in detection of HPV-positive oral malignancies. We assessed the utility of formalin-fixed paraffin-embedded (FFPE) tissues, which are readily available, for evaluation of high-risk HPV infection in oral cancer. For protocol standardization, we used 20 prospectively collected paired FFPE and fresh tissues of histopathologically confirmed oral cancer cases reported in Oral Medicine department of a dental hospital for comparative study. Only short PCRs (~ 200 bp) of DNA isolated using a modified xylene-free method displayed a concordant HPV result. For HPV analysis, we used additional 30 retrospectively collected FFPE tissues. DNA isolated from these specimens showed an overall 23.4% (11/47) HPV positivity with detection of HPV18. Comparison of HPV positivity from dental hospital FFPE specimens with overall HPV positivity of freshly collected oral cancer specimens (n = 55) from three cancer care hospitals of the same region showed notable difference (12.7%; 7/55). Further, cancer hospital specimens showed HPV16 positivity and displayed a characteristic difference in reported sub-sites and patient spectrum. Overall, using a xylene-free FFPE DNA isolation method clubbed with short amplicon PCR, we showed detection of HPV-positive oral cancer in dental hospitals.


Introduction
India harbors the world's largest number of oral cancer cases [1,2]. Oral cancers are among the largest group of preventable cancers. Regular screening and prevention of high-risk habits like smoking, tobacco chewing, and consumption of alcohol and areca nut could effectively control them [3]. These cancers are biologically and psychologically the worst when diagnosed at an advanced stage. Oral cancer is currently on the rise, particularly in specific geographic regions like India [2].
Lately, based on the status of Human Papillomavirus (HPV) infection, oral cancer has been classified as molecularly and clinically distinct subset [4]. Infection of highrisk HPVs is becoming a significant contributor, a decisive risk factor, and likely an etiological agent for these subsets of oral cancers [5,6], S.19 [7]. These cancers occur early in age and have no prior history of tobacco and/or high alcohol consumption [4]. Comparatively, HPV-positive oral cancers are at lower risk of tumor progression and related mortality with improved chemotherapy/radiotherapy response [8][9][10][11]. The incidence trend, particularly in the USA, demonstrates an increase in reporting of new HPV-related oral cancer [12,13]. Whereas, HPV-unrelated oral cancer is reported as stable or on decline, which is consistent with gradual reductions in tobacco use [14]. Among different HPV's in the oral cavity, HPV16 and HPV18 are the commonly detected genotypes [15,16]. More than 90% of HPV-positive oral cancer are infected predominantly with the HPV16 genotype [17], whereas the presence of HPV18 has been reported only sporadically in a few specific studies [8,[18][19][20][21][22].
Unlike cervical cancer, an established clinical infrastructure for screening of oral cancer or clear clinical guidelines for testing of HPV infection in oral lesions does not exist. Further, there are no validated screening methods for oral cancer. Thus, the field is still in its relative infancy. However, a clear advantage of oral screening for mucosal abnormalities was demonstrated in a randomized trial conducted in India that showed a reduction in oral cancer mortality rates [23]. Therefore, understanding the status of HPV infection in a suspected lesion at an early stage will be extremely helpful in management and control of oral cancer. Majority of oral lesions with or without dental involvements usually encountered first at oral medicine/oral surgery clinics of the dental hospital. Therefore, dental hospitals can serve as primary opportunistic screening sites [24,25] and may cater to prevention and early management of oral cancer.
Formalin-fixed paraffin-embedded (FFPE) tissues represent a potentially useful and readily available resource for HPV detection in oral cancer [26]. Several HPV studies have used FFPE specimens in the past to correlate HPV genotypes with histological classification in oral cancer [19,27,28] and other cancers [29]. Although HPV detection is feasible in DNA isolated from FFPE, it poses several technical challenges for subsequent PCR-based diagnostics [30]. Moreover, there has been no study which measured the concordance of HPV positivity detected in FFPE tissues with paired fresh oral cancer biopsies. Therefore, the objective of this study was to understand the incidence of HPV in this oral cancer lesions at dental hospitals and tertiary cancer care hospitals. We also tried to determine whether DNA isolated from FFPE tissues can be efficiently used for HPV detection by comparing results of HPV-specific PCRs with DNA isolated from fresh biopsies and their paired FFPE tissues. To achieve better results, we modified the xylene-free DNA extraction protocol. Further, we validated the technique by evaluating HPV positivity in FFPE specimen of retrospective oral cancer cases reported in the dental hospital. This is a prospective study with inclusion of retrospective samples for comparison purposes.

Clinical specimens and the study design
The comparative study was performed on 146 oral or oropharyngeal cancer cases from a dental and 3 tertiary cancer care hospitals (TCCH) ( All samples were collected prior to any chemo/radiotherapy. Center-wise sample distribution and their subsequent processing for HPV detection are described in Fig. 1. Prior written informed consent was obtained from all the subjects included in the study according to the principle laid down by the declaration of Helsinki, and epidemiological details were taken from their clinical records. For retrospective analysis, archival tissue blocks of histopathologically confirmed oral cancer cases reported at Subharti Dental Hospital were used. The study protocol was approved by the Institutional Ethics Committee of all collaborating hospitals and from the National Institute of Cancer Prevention and Research (NICPR: earlier named as Institute of Cytology & Preventive Oncology), Noida, India, where the laboratory work was carried out. A portion of each biopsy collected in cold 1X phosphate-buffered saline (PBS) was immediately as negative control. International standards for HPV16/18 procured from NIBSC (London) under WHO HPV Lab-Net program were used as reference DNA and positive controls for HPV detection protocols. All other reagents used in the study were of analytical and molecular biology grade and were procured from Sigma-Aldrich (USA) unless specified. Custom-synthesized HPLC-purified primers were procured from M/s Microsynth (Germany) or M/s Eurogentec (Belgium). List of primer used in the present study is described in Table 2.

Genomic DNA isolation from oral cancer fresh biopsies and FFPE samples
Genomic DNA was isolated from freshly collected oral cancer biopsies by the proteinase K digestion followed by phenol-chloroform extraction procedure described earlier [36,37] and referred as protocol P1 (Table 3). PCR amplification was performed following the reference protocol P1. Genomic DNA extraction from FFPE scrape specimens (5 Table 2 List of primers used, their amplicon size, and the annealing temperatures 100 55 [36,37] sections of 5 μm thickness) was performed by two different protocols for comparative analysis. First deparaffinization was done by xylene-based method as described earlier [38] and referred as protocol P2 (Table 3). Briefly, sections were left in xylene for 5 min. or repeated if any paraffin remains and rehydrated using graded alcohol series. Alternatively, paraffin was removed by xylene-free method with slight modification and referred as protocol P3 (Table 3). Variation in present method with respect to the earlier published procedures is outlined in Fig. 2. Briefly, 480 μl of 1X PBS and 20 μl of 0.5% Tween-20 was added to the tissue section in 1.5 ml tube and incubated for 10 min. at 90 °C in pre-equilibrated heating block. After incubation, the tissue was centrifuged at 10,000 g for 15 min. at room temperature and later kept on ice for 2 min. The resulting tissue pellet was obtained by removing the wax disc and the supernatant. The tissue pellet was then dissolved in 500 μl of 1% Tissue Lysis Buffer (0.5 M EDTA, 20 mg/ml proteinase K, 100 mM Tris-Cl, pH 8, 10% SDS) and further incubated overnight at 60 °C in pre-calibrated water bath for digestion. The digested suspension was used for DNA extraction by routine phenol-chloroform method. The DNA pellet was collected by centrifugation for 5 min. at 4 °C, washed with 100% ethanol, air-dried at room temperature or in speed vac and re-suspended in 50 μl of double distilled water, and kept at -20 °C until use. The quality and quantity of isolated genomic DNA from oral cancer biopsies and FFPE tissue sections were assessed by electrophoresis on 1% agarose gel with ethidium bromide in 1X TAE buffer.

Prevention of cross-contamination
Strict attention was observed while sectioning and subsequent processing. Paraffin sections were prepared according to the sandwich method: the first and last sections were Table 3 Evaluation of protocols for DNA isolation from FFPE specimen in comparison to standard DNA isolation protocol (Proteinase K-Phenol Chloroform; PK-PC) performed on freshly col- stained by hematoxylin and eosin to check for tumor presence and to guide microdissection. The tumor specimens were divided into two parts. One part was fixed in 10% formalin and examined at Department of Pathology for histopathological evaluation, while the other part was stored at − 80 °C until transferred to the Molecular Oncology Laboratory at NICPR for analysis. From each tissue block, ten 5-μm-thick sections were cut and placed in two 1.5 ml microcentrifuge tubes for DNA extraction and HPV analysis. After each block, the microtome sectioning area was cleaned with acetone and 70% ethanol. To avoid cross-contamination, each tissue block was handled with new gloves and fresh microtome blade was used each time a new block was sectioned, disposable filter tips were used for all pipetting steps, and separate zones were used for pre-and post-PCR handling. Extraction buffer aliquots without tissue samples were used as negative controls during DNA isolations (one buffer control per 10 samples). Water samples (one per 10 tissues) and PCR master mix without template (one per 10 tissues) were included to monitor reagents contamination and potential cross-contamination in PCR experiments.

Internal control PCR amplification for DNA quality assessment
Prior to detection of HPV infection in oral cancer by PCR using consensus degenerate sequence-specific primers PGMY09/PGMY11 and GP5+/GP6+ and its typing the adequacy of DNA was assessed by subjecting the genomic DNA (100 ng/reaction) isolated from above-mentioned methods for β-globin or p53 Exon 5 PCR which were used as internal controls targeted to host genomic DNA. The PCR products were checked using 2% agarose gels stained with Ethidium Bromide and compared with HaeIII-digested Φx174 DNA as molecular weight marker. All negative samples with adequate DNA on agarose were subjected to re-extraction to remove potential PCR inhibitors and rechecked by PCR for internal control.

HPV detection and typing
PCR was carried out to amplify L1 region of about 33 HPV types that included all high-risk HPVs using consensus PGMY09/11 and GP5+/GP6+ primers that target L1 region but of different sizes. After amplification, the products were analyzed on 2% agarose gel electrophoresis. Similarly all samples were further processed for HPV genotyping by HPV type-specific PCR for HPV16 and HPV18 with short amplicon sizes as described previously [36,37]. The PCR procedures were as follows: initial denaturation and hot-start activation at 95 °C for 10 min. followed by 35

Statistical analysis
Statistical analysis for HPV percent positivity with respect to different variables and hospital type was performed by the chi-square test for proportion using Epi Info 6 version 6.04d PLUS statistical software. Chi-square and Fisher's exact test were employed as and when required to find an association between the results obtained from both the centers with respect to staging, grading, site, and sub-site using IBM SPSS statistics software version 2015. For all analyses, p ≤ 0.05 was considered significant. Concordance between the techniques was observed by calculating the agreement and their respective Cohen kappa value (к) by online available Kappa value calculator website EasyCalculation.com.

Isolation of genomic DNA from FFPE specimens from oral cancer and their performance in HPV PCRs
Efficiency of DNA isolated from FFPE tissues and its suitability for subsequent HPV PCRs were tested. Evaluation of DNA isolated from FFPE tissues in comparison to the DNA isolated from freshly collected paired biopsies using reference protocol showed the presence of detectable genomic DNA only in xylene-free method ( Fig. 3 left panels). In contrast, no DNA could be detected in the xylene-based method (data not shown). To ascertain the quality of DNA isolated from FFPE in PCR-based assays, this DNA was subjected to amplification of genomic fragments of β-globin (268 bp in size) and p53 (184 bp in size). PCR product hereby obtained indicated the amplification of the 184 bp fragment, the p53 gene (Fig. 3 middle and right panels). There was no β-globin amplification observed in these samples (Table 3). PCR using p53-specific primers showed amplifications in all the DNA samples of xylene-free method, whereas only 20% (4/20) specimens showed p53 amplification in DNA isolated from xylene-based pre-treatment method. HPV PCRs targeted to L1 region using primers targeted to long (PGMY9/11; 450 bp) and short (GP5+/GP6+; 127 bp) region of HPV genome demonstrated efficient amplification of short fragment PCR in oral cancer DNA isolated from biopsies or xylene-free protocol, but completely failed in xylene-based protocol (Fig. 4) ( Table 3). The HPV detection using GP5+/GP6+ primers in FFPE DNA isolated from xylene-free protocol was concordant with reference results from fresh biopsies. Similarly, HPV18 PCR that amplified 100 bp fragment of HPV18 E6 was positive in 2 cases and corresponded well with results obtained using reference protocol (Fig. 5).

Generic and type-specific prevalence of HPV infection in oral cancer using FFPE tissues
Overall, HPV positivity in the regional dental hospital was evaluated by incorporating the study of retrospective FFPE samples of 30 additional oral cancer cases reported in the dental hospital from which the DNA was isolated with xylene-free protocol. Among these, 3 samples failed on internal control PCR yielding cumulative sample adequacy of 94% (Table 4). Overall, 11% (5/47) samples were positive for PGMY9/11, whereas 23.4% (11/47) samples were positive for GP5+/GP6+PCR in oral cancer cases reported at the regional dental hospital. Among these HPV-positive samples, 5 samples were found to be HPV18 positive and none showed HPV16 positivity.
To validate this unexpected HPV18 positivity, HPV prevalence and type-specific presence of HPV16 and HPV18 were examined in prospectively collected fresh oral/oropharyngeal biopsies from cases reported at three TCCH of the region which revealed an overall HPV positivity of 10.2% (9/88) and 22% (20/88) by PGMY09/11 and GP5+/ GP6+primers, respectively. All PGMY09/11-positive cases were also positive for GP5+/GP6+. Therefore, for all subsequent analysis, only PCR results from GP5+/GP6+, which also performed efficiently in FFPE tissue, were taken into consideration for further analysis. Among 3 cancer hospitals, the HPV positivity varied considerably from 0% (AIIMS) to 35.7% in (LHMC) when we compared it with the oral cancer cases reported at dental hospital irrespective of the primers used for generic HPV detection. Higher incidence of HPV infection was detected by GP5+/GP6+PCR (23.4% in dental and 22.7% in cancer hospital compared to 11% and 10.2% in PGMY9/11 PCR, respectively, concordance 87.5% and к-value = 0.558). Interestingly, 53% of the oral cancer cases reported as HPV positive were harboring HPV16, whereas no HPV18 infection was detectable in any of the cancer hospital samples.

Comparative analysis of HPV positivity and type-specific prevalence in oral cancer specimen from dental and cancer hospitals
Clinicopathological and demographic evaluation of HPV negative and HPV-positive oral cancer cases revealed a distinct spectrum of patients in dental hospital (Table 5). Lower age group of oral cancer cases reported at the dental hospital Hind III digested λ molecular weight marker; Lane 2-5: Biopsy samples DNA (1B-4B); FFPE samples DNA (1F-4F). Middle and left panels show representative 2% EtBr-stained agarose gel pictures with qualitative assessment of DNA isolated from protocol P1 and P3 for amplification of β-globin and p53 exon 5 by PCR. Lane M: HaeIIIdigested Φx174 DNA molecular weight marker (mean age 46.6 yr vs 53.6 yr (oral cancer cases only) vs 54.6 yr (oral and oropharyngeal cancer cases); p < 0.001). The mean age of the HPV-positive case was relatively lower than the HPV-negative case (39.6 ± 15.0 yr vs 48.8 ± 14.0 yr) p = 0.06. Proportionately more women reported this malignancy at dental hospitals and with a higher percentage of them were found as HPV positive. Evaluation of oral cancer sample distribution with respective oral sub-sites demonstrated cancers of the oropharyngeal region were not reported in dental hospitals, whereas the frequency of tongue cancers reported in dental hospitals was low. Nevertheless, the overall HPV positivity in the dental hospital when compared with all cancers of the oral cavity reported at cancer hospital (that include the case of oropharyngeal cancer also) was very similar in both settings. However, further segregation and comparison with only oral cancer  Table 4 Results of PCRbased detection of HPV infection and its genotypes in oral cancer samples (N = 50) from regional dental hospital using prospective (20) Table 5 Comparative evaluation of HPV positivity and HPV type 16 and HPV18-specific prevalence in oral and oropharyngeal SCC reported in regional dental and tertiary cancer care hospitals Regional dental hospital  2 -cases of dental and cancer hospitals showed a considerably higher frequency of HPV positivity at the dental hospital (23.0% vs. 12.7%; p = 0.01). Overall, reporting of alveolar gingival oral cancer was similar at two centers, but the HPV positivity on this site varied. Alveolar gingival oral cancer of the dental hospital showed higher HPV positivity (26.6% in dental vs. 16.6% in cancer hospital; p = 0.015). It is also to note that none of these tissues were HPV16 positive even in cases reported in cancer hospitals. On the other hand, tongue cancer that was reported at a higher proportion in cancer hospitals showed variable HPV positivity; tongue oral cancer cases were reported at frequencies that could not be analyzed due to the small number of cases (1/4 in dental; 1/14 in cancer hospital) in both hospital settings with similar HPV positivity. Oropharyngeal cancer cases reported at cancer hospitals showed the highest HPV positivity compared to oral cancer cases reported at cancer or dental hospital even if they were clubbed together. Oropharyngeal cancer 39.4% vs oral cancer dental 23.0% vs oral cancer hospitals 12.7% vs (17.8%; p = < 0.05). A comparison of oral cancer presented with different histopathological grades in two setting demonstrated a higher proportion of cases of Well-Differentiated Squamous Cell (WDSCC) type (72.3% in dental vs. 54.0% in cancer hospital). Further, these WDSCC cases reported in dental hospitals displayed a higher HPV positivity (20.6% vs. 6.7% in WDSCC of cancer hospital; p-0.014). Evaluation of clinical stage-wise comparison distribution of cases between dental and cancer hospitals as well as their HPV positivity did not show any statistically significant difference.

Discussion
To our knowledge, this is the first study that compared HPV prevalence in oral cancer at two different clinical settings viz-a-viz dental and cancer care hospitals. In addition, we explored the option of FFPE-based evaluation of oral lesions for HPV detection. We report efficient isolation of DNA from FFPE specimens using a modified xylene-free approach and found the conditional suitability of the DNA isolated for subsequent PCR-based HPV detection assays. The study also compared the spectrum of oral cancer malignancies and their HPV positivity in these two settings. We found a higher HPV positivity in oral cancer cases reported at the dental hospital as compared to the TCCH of the same region. Genotyping of the infection revealed the presence of HPV type 18 in FFPE samples from dental facility in contrast to detection of HPV16 in fresh biopsies from TCCH.
Detection of HPV infection using DNA isolated from FFPE samples is a challenging approach. DNA from FFPE tissues are usually extracted in fragmented form [18,39]. A series of investigations have been performed using the FFPE specimens with variable HPV positivity ranging up to 74% ( Table 6). Some of the investigators attempted to compare the results obtained from FFPE with freshly collected casematched biopsies [40]. However, they used non-malignant tonsillar tissue where they did not find any HPV positivity. In the present study, we evaluated FFPE tissues and their paired fresh biopsies for analysis of concordance and observed that DNA isolated shows fully concordant results only when the xylene-free method is used in combination with PCRs that amplify short targets. A similar direct approach on oral FFPE tissues was applied in some other studies which claimed over 90% adequacy (Fig. 2, Table 6), but we could not get any amplifiable DNA using this approach (data not shown). Among other few approaches for pre-treatment of FFPE sections, the xylene-based method of wax removal was most preferred (Fig. 2) which was also found associated with compromised sample adequacy for PCR-based detection in our study and by others [41].
On the other hand, choice of assay selection for downstream HPV PCR amplification was entirely based on availability of validated short amplicon consensus and type-specific PCR primers in our laboratory. There are a number of assays which employ shorter fragment lengths for HPV detection, which arguably could be more sensitive than the primers used in the present study (Table 6). Notably, the high concordance rate observed in our study was feasible only in PCRs that targeted shorter fragments of DNA for amplification. These results show that DNA fragment length is an essential factor in the evaluation of DNA isolated from FFPE specimen. While using FFPE samples, best PCR amplification with primers amplifying short bp amplicons (~ 200 bp) has been proposed [26]. Difficulties with PCR amplification > 200-300 bp were reported [49,30]. An antagonistic effect of formaldehyde fixation due to DNA modification and crosslinking has also been implicated in the inhibition of PCR amplification of longer (> 200 bp) amplicons [39], and efficiency of the primer pair has been proposed to correlate inversely with the length of the amplicons [18]. Thus, primer sets, like GP5+/6+, which generate short PCR products, have been able to obtain reliable results for formalin-fixed specimens [50].
The overall analysis of HPV prevalence in oral cancer using FFPE specimens from the dental hospital showed positivity of 23.4% by GP5+/GP6+PCR. Our previous investigations for the detection of HPV positivity in freshly collected biopsies of oral cancer from the same geographical region demonstrated a similar prevalence [51,28]. HPV prevalence reported in oral cancer across different regions of India suggests region-specific variation apart from technical differences in HPV positivity (Fig. 6, Supplementary  ST1). However, there was a sharp contrast in the results when we examined the presence of HPV types 16 and 18. Our earlier reports generated on samples collected from cancer care hospitals, however, always indicated the presence Interestingly, a recent study performed on 23 archival tissues of oral cancers from dental hospital settings from the same region has reported the presence of HPV18 with a comparable frequency [21]. Reasons behind such discrepant observations are not known, but observations with such discordant HPV results have been reported by other investigators [52]. Fixation times, especially longer than 12-24 h, highly acidic fixatives, longer periods of storage, and storage in warm environments all contribute to DNA decay. Over time, these factors lead to the destruction of any remaining DNA in the sample [50]. Thus, study could not rule out the loss of HPV16 and a few other genomes due to formalin fixation that resulted in a reduction in their sensitivity of PCR detection. Nevertheless, finding only HPV18 in oral cancer from the dental hospital was unusual and required further investigation. A pooled analysis of specimens from the three centers revealed HPV positivity that closely resembled to the one observed in FFPE tissues from dental hospital. The reported prevalence of HPV16 in oral cancer ranges from 58 to 66% [64,65], whereas HPV18 is the second most common type of HPV detected in oral cancer. It is found much less frequently in HPV-positive oropharyngeal SCCs (2.9%) compared with HPV-positive oral SCCs (34.5%) or HPV-positive laryngeal SCCs (17.2%) [64,65]. Disproportionately high HPV18 prevalence has been reported from Indian or other geographical regions [19,66,21,22,64,60].
The spectrum of oral cancer cases presented at dental hospital differed significantly with respect to sub-sites. The dental hospital setting completely lacked reporting of oropharyngeal cancers, lesser number of tongue cancers, higher proportion of cancers in women, cases with lower mean age, and over-representation of tumor tissues with WDSCC histopathology. The most apparent reason for the skewed spectrum of oral cancers could be their association with reporting malignancies that likely had dental involvement such as pain, mobility of teeth, ulceration, intraoral swelling or mass, tooth extraction wounds or ill-fitting dentures that do not heal in case of alveolar cancer [67]. Additionally, dentists screening for oral cancer risk often refer high-risk patients to specialists for biopsy-based diagnosis [68], These cancer cases displayed a higher HPV positivity in general when compared to similar cases of oral sub-sites compared from cancer hospitals. HPV18 positivity was found associated with alveolar/gingival, buccal mucosa, and retromolar space. HPV18 has a special tropism for glandular tissue and is the most frequently detected type in adenocarcinomas of the cervix. Adenocarcinomas are rare in the head and neck [64] and occur mainly in salivary gland tumors [69]. Further, a higher HPV positivity was detected in oral cancer cases of the dental hospital that were having well-differentiated histopathology. The reasons for such observations are not known. In general, HPV positivity has been associated with differentiated lesions [70,28],S. [71], but the reason(s) for the observed over-representation only in dental hospital cases are not known.

Conclusion
Taken together, our study show the utility of FFPE samples for detection of HPV infection in the malignant oral lesion and can be clubbed with their routine histopathological evaluation. Our study's outcome also suggests potential over-representation of HPV-positive oral cancers with HPV18 positivity in malignancies reported in dental hospital as opposed to the frank locally advanced malignancies reported in TCCH that conventionally report a high incidence of HPV16-positive oral cancer. Moreover, dental hospitals may play a pivotal role in the de-intensification of subsequent treatment of HPV-positive oral lesions right from the beginning if they report the HPV status along with the histopathological findings. Author contribution GV participated in the study design and performed major experimental work and manuscript preparation; NA and SC helped in final preparation of manuscript, communication, and revision; AT assisted in experimental work and data analysis; KV, MJ, and TS helped in experimental work and manuscript preparation; AG, DP, AS, KA, US, and DCD participated in the enrollment of study subjects, their clinical evaluation, provided staging, histopathological grading, tumor diagnosis, and specimen collection; SS participated in the statistical evaluation of data; RM and SMS performed evaluation of clinical and experimental data and assisted in critical review of the manuscript; ACB conceived and designed the study, evaluated data, and critically reviewed, drafted, and communicated the final manuscript. All authors have read and approved the final manuscript. Availability of data and material The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Code availability Not applicable.

Conflict of interest
The authors declare that there are no conflict of interest.