Lack of Association of CYP2C9 genetic polymorphism with oral squamous cell carcinoma

Background: There is increasing evidence for the role of polycyclic aromatic hydrocarbons (PAHs) and heterocyclic aromatic amines (HCAs) in carcinogenesis, including oral squamous cell carcinoma (OSCC). Several of these mutagenic substances are cytochrome (CYP)2C9 enzyme substrates. Methods: In this study, we examined the association of CYP2C9*2 and *3 genetic polymorphisms in 58 OSCC patients and 174 healthy, age, and sex-matched controls. Genotyping was done with allele-specic polymerase chain reaction followed by agarose gel electrophoreses, while selected samples were sequenced for conrmation of genotyping. Results: The wild type genotype (CYP2C9*1*1) was observed at 83%, *1*3 at 8%, *1*2 at 5%, *2*2 at 2% and *2*3 at 2% in combined case and control groups. On further analysis, however, our results did not reveal an association of these variants with OSCC samples (Odds ratio: 0.608, 95% Condence Interval: 0.289 - 1.281, p-value: 0.190). While larger studies are needed to conrm or refute these results, they show a lack of association of CYP2C9*2 and *3 polymorphisms with OSCC in this population. Conclusions: This study demonstrates that the genetic polymorphisms in CYP2C9 genes (CYP2C9*2 and *3) are not causing the risk and are not associated with OSCC. Also, CYP2C9 has no role in the pathogenesis of OSCC in this population of patients. Trial registration: Not applicable.


Background
Head and neck squamous cell cancers (HNSCC) account for about 5% of all cancers worldwide. Among the HNSCC, approximately 90% are squamous cell carcinomas of the oral cavity, pharynx, and larynx (1).
The main risk factors include tobacco and alcohol use (2). The risk is further enhanced by polymorphisms in genes involved in the metabolism of alcohol and tobacco (3). These factors strongly in uence a patient's vulnerability to cancer. Speci cally, polymorphisms in cytochrome P450 (CYP) genes such as CYP2E1 and CYP1A1 are associated with HNSCC (4).
Individuals at higher risk of tumor development may be identi ed by polymorphisms in the genes of drugmetabolizing enzymes. CYP1A1*2A genotype as associated with an enhanced risk of HNSCC in a cohort of Brazilian patients (5). GSTM1 and GSTT1 genotypes exhibited modest association with the risk of HNSCC in a meta-analysis comprised of 8536 controls and 6969 cases (4). A combined effect of GSTM1 homozygous deletion and the CYP1A1*1A/*2A genotype on the risk of developing cancer was also known (6). In a similar study done in the Polish population, the risk of developing HNSCC was associated with CYP1A1*2A genotype and GSTM1 homozygous deletion (7). These studies indicate that polymorphisms in GST and CYP genes confer high risk for developing cancer, especially HNSCC. CYP2C9 is one of the most important drug-metabolizing enzymes in humans and carries out the metabolism of ifosfamide, cyclophosphamide, tamoxifen, warfarin, and etoposide (8). CYP2C9*2 and CYP2C9*3 genotypes result in a 'poor-metabolizer' phenotype and consequently slow down the metabolism of drugs and other substrates metabolized by CYP2C9 (9). In a yeast expression system, CYP2C9*2 and CYP2C9*3 were associated with a signi cant increase in cyclophosphamide clearance (10). Apart from metabolizing various drugs, many mutagens and carcinogens are detoxi ed by CYP2C9 (11,12). The frequencies of CYP2C9*2 and CYP2C9*3 have been reported for multiple populations and display signi cant variations between Caucasian and Oriental populations (13). A study has previously published a higher frequency of CYP2C9*2 polymorphism in lung cancer patients (14)] The risk of distal colorectal adenoma is also reported to be increased with polymorphisms in CYP2C9 (15). However, the association of CYP2C9 genetic polymorphisms with HNSCC patients has not been established yet. The present study was started to nd out if these polymorphisms in CYP2C9 gene are associated with the risk of OSCC in a Pakistani cohort.

Methods
This study was conducted between January 2015-December 2017 at the Dow University hospital, Karachi. The cases group contained 58 OSCC patients, diagnosed and con rmed by a panel of experts.
For control arm, 174 age and sex-matched, unrelated healthy individuals were recruited. Informed written consent was obtained from each participating individual. This research was approved by the Ethics Committee of Shifa International Hospital, Islamabad, and as per the GCP regulations. DNA was isolated from the cancer tissues, embedded in para n and xed in formalin, using QiaAmp FFPE tissue kit (QIAGEN, Germany). The extracted genomic DNA was stored at -20 ο C until PCR and sequencing experimentation.

Genotyping
Polymorphisms in the CYP2C9 gene were determined using allele-speci c PCR. The thermal pro le used was as follows; initial denaturation for 10 minutes at 95 o C, then 37 cycles of denaturation, annealing and extension at 95 o C, 58 o C, and 72 o C, respectively, each for 45 seconds, followed by a nal extension for 7 minutes at 72 o C. Two sets of primers were used for the determination of each genotype. The outer set of primers served as internal control, while the inner set of primers were designed to determine the polymorphism. PCR reactions were carried out in PCR tubes containing 25 µL reaction volume, including 3 µL of DNA template, sets of outer and inner primers, and PCR master mix (Thermo Scienti c, USA). The resulting PCR products were identi ed by resolving them on 5% agarose gel stained with ethidium bromide and visualized under UV light. Results were con rmed by sequencing about half of the samples and can be submitted upon request.

Statistical Analysis
Observed genotypes were compiled and analyzed to nd out allelic and genotype frequencies in case and control groups. Data were recorded and analyzed using SPSS version 23.0. The association between disease status and genotype was investigated using the chi-square and Fisher exact test. A logistic regression model was run to determine the odds ratio and 95% con dence interval for estimating the risk of OSCC. P-value of less than 0.05 was considered to be statistically signi cant.

Results
This study included 58 patients with OSCC and 174 healthy controls. The mean age of the patients was 42.64 years (SD ±12 years) in cases while 40.25 years (SD ± 11.64) in controls. Males were 82% and female 18% in cases while in the control group, 75.9% were male, and 24.1% were female. Of the 58 cases, 24 (41.37%) were localized to buccal cavity, 17 (29.3%) to the tongue, 11 (18.96%) to lips, and 6 (10.34%) to the palate (Figure 1).
To determine whether polymorphisms in the CYP2C9 gene were able to predict the presence of OSCC, genotyping data were compared ( Table 1). The distribution of variant alleles did not show a higher incidence in HNSCC than in control subjects. A statistically insigni cant difference was detected between these groups for CYP2C9*1*1 and CYP2C9*1*2, *1*3, *2*3, *2*2 combined (P =0.190). Similarly, no statistically signi cant association was found for these genotype groups when samples were compared from various anatomic sites (Buccal, p= 0.459, Tongue, p=0.358, Lip, p=0.285, Palate= 0.907) with controls (Table 1). Odds ratio (OR), which quanti es the strength of the association between two events is given for cases and controls along with con dence intervals (CI) (a range of their probabilities).

Discussion
The risk of developing HNSCC is enhanced by association with selected polymorphisms in the genes of certain CYP enzymes, which, apart from metabolizing drugs, are also responsible for the metabolism of several important environmental carcinogens (16)(17)(18). It is suggested that the interaction of these polymorphic CYP enzymes interact with these carcinogens to enhance susceptibility to cancer (19).
Therefore, in the present study, selected polymorphisms in CYP2C9 gene were investigated to nd out if these are associated with HNSCC.
As mentioned earlier, variation in interindividual sensitivity towards carcinogens is a signi cant risk factor for developing various cancers. However, in the present study, CYP2C9 polymorphisms did not reveal any signi cant association with HNSCC patients as compared with the control group. A similar lack of association was observed in another study with 210 HNSCC patients, in which various genetic polymorphisms were studied in important metabolic enzymes, including NAT2, EPHX1, GSTM1, GSTT1, CYP1A1, and CYP2E1 (19). However, another study reported a signi cant association of genetic polymorphisms in CYP1A2 and CYP2E1 genes in HNSCC patients (20).
The tumors arising at four major sites of HNSCC: buccal cavity, lip, tongue, and palate behave differently because of their different properties (21). However, in terms of genetic polymorphisms, minor differences were observed among cancers from these major anatomic sites. Our study revealed that the frequency of CYP2C9 genotypes observed in samples obtained from all the anatomic areas was not statistically different (Table 1). These results indicate that polymorphisms in CYP2C9 gene do not play a signi cant role in oral tumor development in the Pakistani population.
In the majority of squamous cell cancers, CYP1A1, CYP2A6, CYP2E1, and CYP3A genotypes are not frequently expressed in the early stage. Still, in non-small cell lung cancer, they do appear in early-stage and differentiated adenocarcinoma (22). It is speculated that independent of tumor initiation, CYP levels affect various signaling transduction pathways, which alter cell cycle and cause apoptosis or aberrant cell growth and, therefore, might be correlated with carcinogenesis and tumor progression (23,24). However, our study could not nd any association of CYP2C9 genetic polymorphisms with OSCC. There are a large number of studies that have reported a similar lack of association of various CYP genetic variants with OSCC, as reviewed by Vukovic and colleagues (25) and Qiu and colleagues in excellent meta-analyses (26). However, this was the rst study attempting to nd out the association of CYP2C9 genetic polymorphisms with OSCC.

Conclusions
These ndings suggest that the genetic polymorphisms in CYP2C9 genes (CYP2C9*2 and *3) are not predictors of risk and are not associated with OSCC. This indicates a lack of a role for CYP2C9 in the pathogenesis of OSCC in this population of patients.   Relative distribution of different CYP2C9 genotypes in samples obtained from cases and controls.

Figure 3
Overall distribution of CYP2C9 genotypes in cases and controls combined. Relative distribution of different CYP2C9 genotypes in OSCC samples obtained from various anatomic sites in the oral cavity