Association of Micronuclear Frequency with Dysplasia and Cytogenetic Changes (p53 Mutation and p16 Expression) in Oral Potentially Malignant Disorders.

Aims: The present study evaluated the frequency of micronuclei in oral potentially malignant disorders (OPMDs) and their association with the presence of dysplasia on cytology and biopsy as well as their association with p53 mutation and p16 expression. Cytological findings of dysplastic changes in OPMDs were compared to histological diagnoses. Material and
Methods: This was a cross-sectional, observational, descriptive study. Scrape smears (n = 74) were collected from lesions in patients with OPMDs. Punch biopsy was collected in patients showing dysplastic changes. Tissue microarray for p53 mutation and p16 expression was performed using paraffin embedded blocks. Cases were classified into grades of dysplasia using both scrape smears and biopsy. Micronuclei frequency was calculated per 100 cells using scrape smears. Mann-Whitney U test was used for correlation of cytology and histology for grade of dysplasia as well as micronuclear frequency with p53 mutation and p16 expression.
Results: Micronuclear frequency was found to be increased in patients with dysplasia. A significant association of micronuclear frequency with dysplastic changes was seen on cytology. Sensitivity of cytological evaluation was found to be 64.7%. The association of the micronuclear frequency of samples with p53 mutation and p16 expression was nearly significant (n = 28, P = 0.069 and 0.095, respectively).
Conclusion: Micronuclear frequency can be a reliable marker of mutagenic change in OPMDs. Cytological assessment of micronuclei can serve as useful, non-invasive, and relatively inexpensive tool to predict cancerous changes in OPMDs. Copyright:

INTRODUCTION

Oral cancer is defined as malignant neoplasm which is found on lips, floor of mouth, cheek lining, gingiva, palate, or in the tongue.[1] It is highly prevalent across the globe and is one of the leading causes of mortality in India. It has the worldwide incidence of 2.8% and mortality of 1.8%.[2] As per World Health Organization (WHO), oral potentially malignant disorders (OPMDs) are the lesions with morphologically altered tissue having higher likelihood to progress to oral cancer.[3] The common subtypes of OPMDs include leukoplakia, erythroplakia, oral submucus fibrosis (OSMF), lichen planus, and oral ulcers.

Cytology has low sensitivity in screening for evaluation of malignancy or dysplasia; but has advantages of ease of sampling and cost effectiveness.[45] Micronuclei (Mn), which are seen as round to oval cytoplasmic chromatin mass, next to the nucleus, are seen in increased frequency in oral cancer and dysplasia. Quantitative estimation of micronuclei can also serve as a good indicator of genetic damage.[6] Similarly, the p53 mutation and p16 overexpression are the markers of genotoxicity. These genetic alterations can be studied on microarray of biopsy specimen which is also an expensive method.[7]

Hence the current study was undertaken for the assessment of the frequency of micronuclei in buccal smears of the patients with OPMDs and their association with presence of dysplasia on cytology and biopsy as well as microarray findings of p53 mutation and p16 expression. Oral lesions were evaluated on cytology for detection of dysplastic change and were compared to gold standard technique of biopsy.

MATERIAL AND METHODS

The study was conducted in the Department of Pathology at a tertiary health care center and teaching hospital in central India. Approval was taken from the local institutional ethics committee before the commencement of the study. It was a hospital-based descriptive cross-sectional observational study done on the patients having OPMDs and visiting the hospital for treatment. In the given hospital, the patients with oral lesions are managed in Ear Nose and Throat (ENT) and dentistry OPDs. If the lesions had an outward appearance of leukoplakia, erythroplakia, oral submucous fibrosis, lichen planus, and oral epithelial ulcers they were referred to the pathology department for diagnostic evaluation in form of cytology and biopsy. Adult patients (age >18 years) referred to pathology were included in the study after taking informed consent. Patients with lesions that had the overt clinical appearance of squamous cell carcinoma and those with inflammatory lesions were excluded from the study. Seventy six study subjects were chosen for the study by convenience sampling. They were first examined by scrape cytology. Biopsy was performed on the consenting patients who had a significant dysplastic change on cytology or had an outward appearance of lesion suggestive of possible malignant transformation.

Cytology

For cytological evaluation, buccal smears were taken from the lesions with wooden spatula or toothbrush. The choice of the tool was governed by type of the lesion. Afterwards these scrape smears were fixed immediately using 100% ethanol for staining with Hematoxylin and Eosin and Papanicolaou stains. Dry smears were processed in stepwise fashion to apply May Grunwald Giemsa stain. The micronuclear frequency per 100 cells was decided by the consensus of two independent observers.

Dysplasia was defined by presence of following features on cytological evaluation[8]

Hypercellularity

Nuclear hyperchromasia

Nucleocytoplasmic ratio (>0.5/1)

Koilocytes

Nuclear membrane irregularity, nuclear pleomorphism

Bi/multinucleation, prominent nucleoli

Mitosis.

All the scrape smears were examined under microscope for presence of micronuclei per 100 cells by two independent observers. Micronuclei were identified by using Tolbert's criteria as follows: [Illustrated in Figure 1] [9]

Figure 1

Smear showing 6 microneuclei in a cell (PAP, 40×)

Rounded, smooth perimeter suggestive of a membrane

Less than a third the diameter of the associated nucleus but large enough to discern shape and color.

Feulgen positive (i.e., pink in bright field illumination)

Staining intensity is similar to that of the nucleus

Texture similar to that of nucleus

Absence of overlap or bridge to nucleus and in same focal plane as that of nucleus.

The micronuclear frequency per 100 cells was manually calculated and finalised by the consensus opinion of two independent observers.

Biopsy

For the patients who showed dysplastic changes, punch biopsy was collected from the area of mucosal change showing nodularity, verrucous change, ulceration, or change in texture. Punch biopsy was performed from the most representative area using a 6 mm punch. The biopsy specimen was collected in 10% formalin in a sterile container and was followed by routine tissue processing and staining with Hematoxylin and Eosin. On the histological evaluation of the biopsy specimen, the cases were classified into mild, moderate, and severe dysplasia as per standard guidelines.[10]

Tissue microarray

Tissue microarray for p53 mutation and p16 expression was performed using paraffin-embedded blocks for 30 biopsies. Two biopsies were lost to methodological errors during the process of microarray. Hence data regarding p53 mutation and p16 expression of 28 patients was obtained. p53 mutation on the microarray was identified if more than 5% of nuclei showed positive nuclear stain. p16 expression was identified by strong nuclear and cytoplasmic expression in a continuous segment of at least 10–20 cells.[10]

Statistical analysis

The data collected was initially entered in Microsoft Excel and data cleaning was done. The descriptive and inferential statistics were later performed in Epi info 2007 software. The qualitative data has been presented as percent proportion of total and the quantitative variables have been expressed as mean with standard deviation. The distribution of quantitative variables for normality was assessed by the Kolmogorov–Smirnov test and non-parametric test (Mann–Whitney U test) was used as the data were not normally distributed. The confidence interval was set at 95%.

RESULTS

Sociodemographic characteristics of the study population

The mean age and gender distribution of study population has been illustrated in Table 1. The mean micronuclear frequency was not associated with age and sex of the study sample as shown in the table. The frequency of various oral premalignant lesions has been represented in Table 2.

Table 1

Association of age and sex with micronuclear frequency (N=76)

Correlation of Age (in years) with Micronuclear frequency per 100 cells (Spearman correlation test)
Age (Mean±SD)	40.53 (± 9.51)
Micronuclear frequency (Mean±SD)	2 (± 1.9)
Spearman correlation coeff.	0.102
Sig. (2 tailed)	0.382
Association of sex with Micronuclear frequency per 100 cells (Mann–Whitney U test)
Sex*
Male	57 (75)
Female	19 (25)
U	446.0
Sig. (2 tailed)	0.242

* Sex, expressed as Total number (percentage)

Table 2

Frequency of substance used and oral lesions

	Frequency (Percentage)
Substance used
None	15 (19.7)
Oral tobacco	53 (69.7)
Smoking	4 (5.2)
Alcohol	1 (1.3)
Combined use	3 (3.9)
Gross appearance of oral lesions
Leukoplakia	23 (30.3)
Erythroplakia	3 (3.9)
OSMF*	1 (1.3)
Ulcer	49 (64.5)

*OSMF: Oral Submucus Fibrosis

Cytology, histology, and microarray findings of oral premalignant lesions

Cytological evaluation of oral lesions was done for all 76 study participants. Dysplasia was found in more than half of the cases (n = 39, 51.3%) [Figure 2] whereas 37 (48.7%) cases did not show dysplastic changes. Analysis of dysplastic changes in individual lesion subtypes revealed that dysplasia was present in 34.8% of cases with leukoplakia, 66.7% erythroplakia, and 59.2% ulcer patients. Cytology in the single case of OSMF did not show any dysplastic change. Histopathological evaluation of biopsy specimens was done for 36 consenting cases. This evaluation confirmed dysplastic changes in most of the patients (n = 34, 94.4%) whereas no dysplastic change was seen in 2 (5.6%) cases. The microarray of the biopsy specimens of 28 subjects revealed the presence of p53 mutation and p16 expression in 22 (78.2%) and 9 (32.1%) study subjects, respectively [Table 3].

Figure 2

A case of dysplasia on scrape smears with cells showing high N/C ratio, nuclear hyperchromasia and nuclear polymorphism (H/E, 40×)

Table 3

Findings of cytology, histology, and microarray of oral premalignant lesions

Gross appearance of OPMD →	Leukoplakia	Erythroplakia	OSMF	Ulcer	Total
Cytology findings (n=76)
Dysplasia present	8 (34.8)	2 (66.7)	0 (0.0)	29 (59.2)	39 (51.3)
No dysplasia	15 (65.2)	1 (33.3)	1 (100.0)	20 (40.8)	37 (48.7)
Histology findings (n=36)
Dysplasia present	10 (83.3)	1 (100)	0 (0.0)	23 (100.0)	34 (94.4)
No dysplasia	2 (16.7)	0 (0.0)	0 (0.0)	0 (0.0)	2 (5.6)
Microarray findings (n=28)
P53 mutation present	5 (62.5)	1 (100.0)	0 (0.0)	16 (84.2)	22 (78.6)
P53 mutation absent	3 (37.5)	0 (0.0)	0 (0.0)	3 (15.8)	6 (21.4)
P16 expression present	4 (50.0)	0 (0.0)	0 (0.0)	5 (26.3)	9 (32.1)
P16 expression absent	4 (50.0)	1 (100.0)	0 (0.0)	14 (73.7)	19 (67.8)

Frequency expressed as Total number (percentage)

Association of micronuclear frequency per 100 cells with dysplasia, p53 mutation, and p16 expression

There was a significant association of Mn frequency with the dysplastic changes on cytology as revealed by the Mann–Whitney U test. (P < 0.001) The number of micronuclei in the samples with dysplasia was significantly higher (mean 2.82 ± 1.93) than those without dysplasia (mean 1.14 ± 1.27). Although the Mn frequency was higher in the dysplastic group, no significant difference between the Mn frequency was found between the dysplastic and non-dysplastic samples of histopathological evaluation (P = 0.436). The mean difference between the Mn frequency of the samples with presence and absence of p53 mutation and p16 expression was close to being significant (n = 28, P = 0.069 and 0.095, respectively) [Table 4].

Table 4

Association of micronuclei frequency per 100 cells with dysplastic changes seen on cytology and histology as well as with p53 mutation and p16 expression

Pathological findings subgroups	Mean frequency (±S.D.) of micronuclei per 100 cells
Cytology findings (n=76)
Dysplasia present (n=39)	2.82 (±1.93)
No dysplasia (n=37)	1.14 (±1.27)
Sig. (2 tailed)	0.000*
Histology findings (n=36)
Dysplasia present (n=34)	2.12 (±2.08)
No dysplasia (n=2)	1 (constant)
Sig. (2 tailed)	0.436
Microarray findings (n=28)
p53 mutation present (n=22)	2.59 (±2.3)
p53 mutation absent (n=6)	1.00 (±0.63)
Sig. (2 tailed)	0.069
p16 expression present (n=9)	1.44 (±1.59)
p16 expression absent (n=19)	2.63 (±2.31)
Sig. (2 tailed)	0.095

(Statistical significance calculated using Mann–Whitney U test)

Sensitivity of cytology for diagnosis of epithelial dysplasia in the oral premalignant lesions

Considering histopathology as a gold standard to diagnose epithelial dysplasia the sensitivity of cytology in the identification of dysplasia in OPMDs was found to be 64.7% [Table 5].

Table 5

Comparison of findings on cytology and histology. (n=36)

	Histology dysplasia		Total
	Absent	Present
Cytology dysplasia
Absent	1	12	13
Present	1	22	23
Total	2	34	36

Sensitivity of the cytology to measure dysplasia: 22/34 * 100=64.7%

Prediction of dysplasia on the basis of Mn frequency per 100 cells using cytological diagnostic techniques

A Receiver-Operating Characteristic (ROC) curve was plotted to assess the optimal cut off for the frequency of micronuclei per 100 cells for the presence of dysplasia on the cytological diagnosis. The ROC curve analysis showed the optimal cut-off point at the Mn frequency of 1.5. The area under the curve of 0.758, sensitivity of 74.4%, and specificity of 70.3% was noted for the Mn frequency of 1.5. [Graph 1] Hence it can be interpreted that the Mn frequency of more than or equal to two can predict the presence of dysplasia on cytological techniques.

Graph 1

Receiver operating characteristic curve for association of micronuclear frequency with cytological diagnosis of dysplasia under the non-parametric assumption. Area under curve: 0.763; Asymptotic Sig. (P) < 0.001. Red Circle indicates optimal cut off point as per Youden’s index at the sensitivity value 0.744 and 1-specificity value of 0.296

DISCUSSION

Oral potentially malignant lesions have high rates of malignant transformation.[11] Monitoring of these pathological changes is gaining emphasis for timely diagnosis and management to improve survival in oral cancers.[12] Most oral cancers originate from epithelial cells. Hence assessing micronuclear frequency in the oral epithelial cells by exfoliative cytology is a sensitive tool for screening cellular alteration in OPMDs.[12] Additionally, the Mn frequency has been found to have a significant correlation with histological grading of oral squamous cell carcinoma.[13] Mn in the exfoliated oral cells is considered as a marker of chromosomal damage caused by genotoxic tobacco-related products.[14] Although feulgen and acridine orange reaction and fluorescent in situ hybridization are popularly used staining methods for assessing Mn frequency we used Papaniculalou stain, May Grunwald Giemsa stain, and Hematoxylin and Eosin for the present study as they are simple to use and have better clarity while visualizing epithelial cells to accurately calculate Mn frequency.[1516]

In the present study, the mean number of micronuclei in various OPMDs (leukoplakia, erythroplakia, OSMF, ulcers) was evaluated and correlated with the cytological and histological findings of dysplasia. The mean number of micronuclei in dysplastic lesions which was 2.82 (± 1.93), was found to be significantly higher than the lesions without dysplasia 1.14 (± 1.27). Our findings match with those of Kiran et al.[5] who found the mean Mn frequency in dysplastic (4.1) and normal buccal mucosa (1.46) to be significantly different. This may be because the micronuclei in dysplastic epithelium could be the markers of lagging chromosomes in anaphase, aberrant mitosis, or acentric chromosomes.[10] Since the majority of the patients (69.7%) included in the study used oral tobacco, the increased micronuclear frequency could also mark the genetic damage to the oral epithelium caused by N-nitrosamine which is released from tobacco as observed in the study by Sangle et al.[13]

p53 is a tumor suppressor gene that protects the genome. Mutation of p53 is more frequently found in oral cancer and precancerous lesions compared to normal samples.[17] [Figure 3] As micronuclear frequency also marks genetic damage the correlation between p53 mutation and micronuclear frequency was studied by Salazar et al.[18] Here, a direct correlation was found between Mn frequency and levels of p53 mutation on linear regression analysis. Similarly in our study, the micronuclear frequency difference between p53 positive and p53 negative biopsy specimen of OPMD approximated significant value on the Mann–Whitney U test. (P = 0.069) Similarly, p16, a cyclin-dependent kinase inhibitor that prevents excessive proliferation of tumor cells, is expressed at higher levels in neoplastic as compared to normal cells.[19] [Figure 4] In a study by Gashi et al.[20] which was done on peripheral blood lymphocytes of patients with cervical intraepithelial neoplasia, there was a positive association between Mn frequency and p16INK4a/Ki-67 co-expression. Analogous findings were noted in the current study where p16 expression and micronuclear frequency had association levels that neared significance level (P = 0.095) on the Mann–Whitney U test.

Figure 3

A case of well differentiated squamous cell carcinoma which was p53+ve (40×)

Figure 4

A case of squamous cell carcinoma positive for p16(40×)

Doing the micronuclear assay has certain advantages of being simple to apply, non-invasive, and economical procedure.[21] To add, Mn assay is applicable only on interphase cells making it a good indicator of mitotic interference, chromosomal mutations, and chromosomal breakage. This is the reason for the increase in the use of Mn assessment on exfoliated cells since their introduction in 1970 by Boller.[22] In contrast, the histopathological diagnosis of the biopsy specimens is more tedious and costlier. However, histopathology is the gold standard investigation in the identification of cancerous change. In the study by Babshet et al.,[23] the cytological diagnosis of brush cytology was compared with the histopathologic diagnosis on punch biopsy. The sensitivity of cytology was found to be 77% in the detection of dysplastic change considering biopsy as a gold standard investigation. Somewhat lower figures were obtained in the current study with the sensitivity of oral cytology being 64.7% in the identification of dysplastic change. The difference in the findings could be attributed to differences in sample size and study procedures. On the ROC analysis, it was found that micronuclear frequency of two or more per 100 cells was associated with dysplasia. Our findings echoed the findings of Kiran et al.[5] of mean Mn frequency varied significantly with the presence of epithelial dysplasia. The buccal Mn assay has been previously theorised as a predictor of carcinogenesis.[24] However, the present study is among the few research studies which assessed the optimal Mn frequency cut-off for the purpose.

The study was conducted in the clinical settings on the patients who visited the hospital for treatment. As a result, the people with OPMD in the community not attending the clinic may not be represented thereby hampering the generalizability of findings. There is a huge resource constraint in India for the provision of adequate health care to all. Hence cheaper and easily applicable techniques are needed which can reasonably predict the management required in potentially malignant lesions. More extensive and community-based research may be needed on the validity of cytology and micronuclear assay in the diagnosis of oral precancerous lesions for early identification of malignant change at a lower cost.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form, the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.