This was a prospective cross-sectional study. The study was independently reviewed and approved by Ethical Committee of Government Dental College and Hospital, Nagpur, Maharashtra, India (MUHS/PG/E-2/2240/2014). The study was conducted at the Outpatient Department of College, from January 2015 to October 2016 which included 150 patients divided into three groups: Group I: 50 apparently healthy controls not habituated to tobacco chewing and smoking, Group II: 50 clinically diagnosed cases of oral submucous fibrosis, they were not histopathologically confirmed because it is not possible to do the biopsy for each and every OSMF patients unless indicated. All the patients of OSMF were having habit of chewing areca nut and/or tobacco chewing habit in various forms (kharra, pan masala, gutkha etc.,) for over 5 years. The patients were subgrouped according to IIO based clinical grading as per Khanna and Andrade classification [11]. Grade I which is >35 mm IIO consisted of 3 patients, Grade II (26-35 mm IIO) of 17 patients, Grade III (16-25 mm IIO) of 20 patients and Grade IV A (<15 mm IIO) of 10 patients. Group III consisted of 50 clinically diagnosed and histopathologically confirmed cases of OSCC of which 37 had well differentiated, 10 moderately differentiated and 3 poorly differentiated OSCC. These patients were selected irrespective of tumour size, site and grade and all were regular tobacco chewers in various forms for over 10 years.

All individuals included in the study were aged from 18-60 years. The control group includedage and gender matched 50 healthy individuals from a similar socioeconomic background as that of the patient group. They did not have any tobacco habits, systemic or local illness and had visited the hospital for routine check-ups or other minor problems. Patients with HIV, chronic alcoholics, with underlying systemic diseases, cancer other than oral cancer were excluded from the study. Written informed consent was obtained from the patients, after the planned study was explained in detail. A detailed history with thorough clinical examination was performed, and the findings were recorded. The clinical characteristics of controls, OSMF and OSCC patients are shown in [Table/Fig-1].

Five mL of Intravenous blood sample was collected from all 150 patients for estimation of MDA. It was collected using a disposable syringe with 23-gauge needle and was transferred into plain vacutainer. The blood was allowed to clot in an upright position for at least 30 minutes but not longer than 1 hour before centrifugation. These samples were centrifuged in Remi bench top centrifuge for 15 minutes at 3000 RPM within one hour of collection. Serum was seen separated as the top transparent layer. Serum was then transferred to a new test tube and stored in a deep freezer at -20°C for quantitative estimation of serum MDA. MDA forms a 1:2 adduct with thiobarbituric acid and produce coloured complex which can be measured by fluorometry or spectrophotometry. Mean±SD or SE must be established by each laboratory. Variations may occur in individual laboratories due to pipetting, geographic temperature change, incubator temperature etc. As oxidativestress increases Thiobarbituric Acid Reactive Substances (TBARS) which is present in biological specimens also increases. Depending upon the presence of anti-oxidants TBARS return to normal levels over time. TBARS was expressed in terms of MDA equivalents. In the assay a standard curve is constructed using MDA standard against unknown samples. The tests was performed by using the Oxitek TBARS Assay kit and is specially designed for researchers studying oxidative stress and anti-oxidant activity. Every reagent was prepared fresh for each analysis. Absorbance of supernatants was read at 532 nm with the help of BIO RAD SmartSpec Plus spectrometer.

The data was collected, tabulated and analysed by SPSS 16© (Statistical package for Social Sciences) software. Appropriate test of significance was applied. For inter group comparison One-way ANOVA and Post-hoc test were applied and for comparison between two groups independent t-test was applied p-value <0.05 was considered as statistically significant and <0.001 as highly significant.

Comparative statistics for mean serum MDA levels between the healthy control group, OSMF and OSCC groups was done; results are shown in [Table/Fig-2]. The present study has revealed an intriguing aspect of tumour biochemistry. This study showed highest values of MDA in OSCC (57 nmol/mL), in OSMF (25.87 nmol/mL) and lowest (10.50 nmol/mL) in normal healthy control. Thus, a significant increase in serum MDA levels was found from normal healthy individuals to OSMF to OSCC. One-way ANOVA test was applied and p-value was <0.001 i.e., statistically highly significant. When post-hoc test was applied for further inter group comparison; the mean serum MDA level was significantly high in OSCC patients (Group III) as compared to OSMF patients (Group II) and control group (Group I). Estimation of mean serum MDA level of Group II patients with subgroups according to IIO was done, calculated and compared as shown in [Table/Fig-3].

Group III patients were categorised based on site of lesion in oral cavity and estimation of serum MDA value was done based on site distribution and the results were tabulated as shown in [Table/Fig-4]. It was observed from the results that buccal mucosa was the predominant site for the occurrence of squamous cell carcinoma. Out of 50 patients, 26 patients were having lesion over buccal mucosa but the average serum MDA value was highest in patients involving alveolus and minimum with tongue lesions. When one-way ANOVA test was applied to compare these values, p-value was 0.77 i.e., statistically not significant.

Group III patients when subgrouped according to primary tumour size, for which TNMS classification of clinical grading was used [12]. The results are shown in [Table/Fig-5]. It was observed that there was a significant increase in serum MDA value from T1 to T3 group. One-way ANOVA test was applied and p-value was <0.05 (i.e., statistically significant). This implies that, there was a significant difference in level of mean serum MDA in patients of OSCC with different tumour size.

Group III patients were also grouped according to histopathological grading for which Bryne’s histopathological grading (1989, 1992) (ITF) which is based on InvasiveTumour Front Grading System was used [13]. Bryne M et al., suggested a hypothesis that at the invasive front area of several squamous cell carcinomas, the morphological characteristics reflects tumour prognosis far better than any other parts of the tumour [13]. Estimation of serum MDA value was done according to histopathological differentiation as shown in the following [Table/Fig-6]. Serum MDA level was evaluated according to histological grading but the measured parameter did not show significant changes. In Oral squamous cell carcinoma the level of lipid peroxidation was inversely proportional to the degree of differentiation. MDA levels in well-differentiated squamous cell carcinoma were greater as compared to moderately differentiated and poorly differentiated squamous cell carcinoma, and this difference was statistically not significant.

Areca nut is a known etiological factor for oral submucous fibrosis which contains arecoline, arecaidine, guvacolin and guacine which produces numerous free radicals and ROS. Free radicals can produce lipid peroxidation in membrane, Oxidative modification of proteins, and lesions in DNA which can directly and indirectly stimulate the carcinogenic effect in cells. Intracellular and extracellular antioxidant system neutralises the deleterious effects of ROS. Hence, estimation of the level of ROS in Oral submucuous Fibrosis patients may serve as important biomarker to analyse the progression and malignant transformation of the disease [14]. Nicotine which is present in tobacco causes pH changes during chewing which causes partial reduction of oxygen and produces highly reactive free radicals like hydrogen peroxide (H₂O₂), hydroxyl radical (OH*) and superoxide anion (O2-) in the body fluid, such as blood and enhance lipid peroxidation levels of biological molecules, hence serum MDA levels thus thereby increased, as seen in OSMF and OSCC patients [15].

Naturally occurring lipid peroxidationend product, MDA is highly mutagenic andtumourogenic. The MDA reacts with both the Deoxyadenosine and Deoxyguanosine in the DNA and produce DNA-MDA adduct [14]. Early detection of OSCC and OSMF transforming into malignancy can drastically improve the treatment outcomes and prognosis. In the search for possible causes of malignancies on one hand, and the need for the modality affording early diagnosis, there are various biochemical findings which can be useful in early diagnosis at an early stage of the disease [16]. The MDA value in blood is a measure of the ability of the body to handle the oxidative stress [17]. So, in the present study serum MDA level in OSMF and OSCC was estimated to assess the degree of oxidative damage of the disease so that it can be arrested in early stages to avoid the possible consequences of OSMF turning into malignancy.

There was a statistically significant increase in serum MDA level from group I (10.50 nmol/ml) to group II (25.87 nmol/ml) to group III (57.00 nmol/ml) with p-value < 0.001. The present study findings were in accordance with the study conducted by Chole RH et al., Korde SD et al., D’souza D et al., and Ganesan A et al., [17-20]. ROS-induced lipid peroxidation has been indicative in malignant transformation. The prime targets of peroxidation by ROS are the polyunsaturated fatty acids in membrane lipids. The decomposition of these peroxidised lipids yields a variety of end products which also includes MDA. Thus, the level of MDA indicates the extent of lipid peroxidation and serves as a marker of cellular damage due to free radicals. The increase in MDA may be attributed to excessive formation of free radical due to various tissue abuse habits and due to decomposition of polyunsaturated fatty acids present in membranes. It may also occur due to inadequate clearance of free radicals by poor cellular antioxidant system [21].

In group II patientsclinical grade wise analysis showed that mean MDA level gradually increased from grade I to grade IV and the result were statistically significant (p<0.05) and so there exists a direct correlation between clinical grading and MDA value. However, there was slight reduction in mean serum MDA value in grade IVA probably because of unequal distribution of sample size in each grade. It was established that the lipid peroxidation increases with the severity of disease thus indicating the extent of tissue injury [19]. These findings were similar to Metkari SB et al., and Paulose S et al., [22,23]. Contrasting results were reported by Tejasvi MA et al., [24]. Our findings were also not in accordance with the findings by Shakunthala GK et al., in which the difference in serum MDA levels among different clinical staging of OSMF patients were not statistically significant [25]. The present study results showed slight reduction in serum MDA level in grade IVA than in grade III could be due to self–defence mechanism by the proliferative tumour cells to resist deleterious effects of lipid peroxidation on their cell membranes or possible reason could be due to utilisation of MDA in crosslinking of collagen. Chojkier M et al., studied the effects of MDA on collagen production and they found that the addition of MDA (200 μM), to the cultured human fibroblasts incubated in MEM {Eagle’s minimal essential medium} without fetal calf serum (to minimise MDA binding to media proteins), increased 2-fold the production of collagen without affecting non-collagen protein production. The concentration of MDA in the media was 80% of the initial concentration after 3 hour incubation [26]. Thus, it can be concluded that the addition of MDA to cultured fibroblasts increase collagen production by 2-3 times [27].

Group III patients were categorised based on site of occurrence of lesion, serum MDA level in patients involving alveolus was highest. In contrast to the results obtained by Subapriya R et al., who found highest serum MDA level in buccal mucosa [28], the results of the present study revealed that the tumour site and blood comprise two separate metabolic compartments with respect to their susceptibility to lipid peroxidation levels.

Comparison of mean serum MDA level done based on tumour size, a statistically significant difference (P<0.05) was found with mean serum MDA value being highest in T3 patients. Therefore a positive relationship between tumour size and serum MDA was found. Similarly, Manoharan S et al., obtained results in oral cancer patients and found a significant increase in serum MDA level with increase in clinical staging [29]. Our findings were also in accordance with the findings of Srivastava KC et al., [30]. Our results were not in accordance with the study done by Rasheed MH et al., Gupta A et al., who showed no statistical difference in mean serum MDA level with increase in clinical staging [31,32]. Advance stage Head and Neck cancer patients are exposed to higher oxidative stress. Measurement of lipid peroxidation by product in circulation of oral cancer patients may thus be helpful in assessing the clinical stage of oral squamous cell carcinoma. Cell membranes of erythrocytes and other cells are mainly composed of Polyunsaturated Fatty Acid (PUFA) which is considered to be highly susceptible to oxidative attack and also become the major substrate for ROS mediated damage. Due to such damage the fluidity and permeability of the membranes are altered. Thus, large volumes of MDA levels in plasma could be attributed to its increased formation in erythrocytes and its consequent leakage into the plasma or due to inadequate clearance of free radicals by the cellular antioxidants.

A significant decrease in serum MDA level was observed from well differentiated to poorly differentiated squamous cell carcinoma and was not found to be significant statistically (P >0.05). Similar results were obtained by Metgud R et al., Patait MR et al., [15,33]. A complex relationship is anticipated in histopathological differentiation and serum MDA levels which is very poorly explored till date. In our study, there was no significant difference in serum MDA levels of Group III subjects based on histopathological differentiation, probably due to random selection of the patients and thereby unequal distribution of patients in each grade, hence; there is need to conduct further studies in this area, with equal number of patients in each group.

This study is an effort to find out a cause and effect relationship between OSMF, OSCC and serum MDA levels. The patients were selected randomly for this study. In this study an attempt to correlate serum MDA levels with histopathological grading of OSMF patients was not done as it was not possible to obtain biopsy of these patients unless it is advised/indicated. For OSCC patients instead of the entire TNMS clinical staging only tumour size was considered for the study purpose since the patients were selected randomly and enrolled prior to the treatment at the institute. The disease was diagnosed based on their clinical, radiological, and histopathological examinations. Large scale studies with long term follow up and equal distribution of samples among different grades of OSMF and OSCC should be carried out in order to establish MDA as a potential biomarker for oxidative stress.

The estimation of serum MDA levels might be helpful to determine the possibility of malignant transformation of OSMF, assessing the status of OSCC and thereby possibly in clinical intervention of OSCC patients to some extent. The present study suggests a role of MDA as one of the diagnostic biomarkers and innovative tools to monitor oxidative stress and their impact on progression of oral potentially malignant disorders and epithelial malignancy. Serum can be used to determine the impact of redox imbalance on the progression of OSCC but it is difficult to prove cause-effect relationship or to predict its prognostic use with such observational study with small sample size.