Micronuclei: An essential biomarker in oral exfoliated cells for grading of oral squamous cell carcinoma.

BACKGROUND: Micronuclei in exfoliated oral epithelial cells have been shown in some studies to correlate with severity of this genotoxic damage. This severity can be measured in terms of grading of the lesions. AIM: To correlate frequency of micronuclei (MN) in oral exfoliated cells in clinically diagnosed cases of oral squamous cell carcinoma (OSCC) followed by a histopathological grading.
MATERIALS AND METHODS: The study subjects consisted of clinically diagnosed cases of OSCC. Healthy subjects without any tobacco consumption habits formed the control group. The cytosmears from both groups were stained with rapid Papanicolaou stain. MN were identified according to the criteria given by Countryman and Heddle with some modifications.
RESULTS: The frequency of MN was three to four times higher in patients with OSCC as compared to patients in the control group and the difference was found to be highly significant. In 75% cases, the cytological grade as determined by the frequency of micronuclei correlated with the histopathological grade and this observation was statistically significant.
CONCLUSIONS: MN can be a candidate to serve as a biomarker for prediction of the grade of OSCC.

Introduction

Micronucleus (MN) is a recently upgraded topic, especially in the field of oral cancer. MN takes its origin from chromosome fragments or whole chromosomes, which lag behind at anaphase during nuclear division. It could be argued that MN in exfoliated oral epithelial cells represent a preferred target site for early genotoxic events induced by carcinogenic agents. Various studies have shown the correlation of frequency of MN and severity of this genotoxic damage. This severity can be measured in terms of grading of the lesions. This especially holds true for oral squamous cell carcinoma (OSCC). There is lot of controversy surrounding the criteria to be followed for identification and method of counting of MN. Recent studies are attempting to show the correlation between frequency of MN in oral exfoliated cells and histopathological grading of OSCC.[1] Present study has made an attempt to validate such correlation, if any. We also put forward a new modified criterion in relation to assessment of size of MN and the method of counting of MN.

Aims and Objectives

The aim of the study was the evaluation of correlation between the frequencies of MN in oral exfoliated cells from patient clinically diagnosed of OSCC. This was followed by histopathological grading of OSCC in the same subject.

The objective was to test the hypothesis whether MN in oral exfoliated cells can be a parameter for grading of OSCC.

Materials and Methods

Subjects included in the study were divided into the following groups:

Group I (study group): comprised 16 subjects diagnosed clinically of OSCC.

Group II (control group): comprised 16 subjects without any oral lesions and any history of habits of tobacco consumption. The age and gender of this group was matched with that of Group I.

A written consent was taken from subjects of both the groups.

Three observers participated in the study for analysis as mentioned below:

Observer 1: Counting of MN in Group I

Observer 2: Counting of MN in Group II.

Observer 3: Histopathological scoring and grading of OSCC.

The slides of each group were exchanged among all the observers, providing each observer a participation in each type of observation. They were not provided with information regarding the study subjects to prevent observer bias. The average score of three observations by the three observers was calculated for each group to resolve the interobserver bias.

To obtain the smear of exfoliated cells from the oral cavity (buccal mucosa in control group), a wooden spatula was used. It was slightly moistened before applying on the mucosa. Patients were also asked to rinse the oral cavity before taking the samples. This was to remove food debris and necrotic slough, if any, which could hamper the quality of the cytosmear. In subjects suspected of OSCC, the most representative site was selected for obtaining the smear, like the margins of the lesion. A light gentle pressure was applied while scraping. The smears were fixed and stained with Papanicolaou (PAP) stain.Whole of the smear was screened for counting of MN. Smear was first screened at ×400 for testing the quality of staining, followed by examination at ×1000. MN were identified according to the criteria given by Countryman and Heddle[2] with some modifications as suggested by authors.

Measurement of micronuclei

According to the criteria put forward by Countryman and Heddle[2] size of MN should be at 1/3 of the size of the nucleus.We have modified the criteria regarding the assessment of size of MN. Structures ranging from 1/3 to 2/3 of the size of nucleus were also considered as MN. Screening of slide was done in a zigzag manner starting from one end of the slide and approaching towards the other end. A simultaneous counting of MN was done for each subject in each group. The average frequency of MN was further tabulated based on following formula:

Average frequency of MN=Total number of MN/ Total number of cells with MN.

Based on the average frequency of MN, a cytological grade was determined [Table 1].

Table 1

Cytological grade of OSCC based on the average frequency of micronuclei

Histopathological grading for OSCC for the subjects diagnosed of OSCC (histopathologically) was also done. Malignancy point score was counted based on new malignancy grading system given by Anneroth et al.[3] According to this system, six parameters are used to determine the total score. The six parameters are degree keratinisation, nuclear pleomorphism, number of mitotic figures, pattern of invasion, stage of invasion and plasmalymphocytic infiltration. Each parameter is given score of 1–4. Then, malignancy point score is calculated as follows:

Malignancy point score = Total score/Total number of parameters used.

Based on malignancy point score, a histopathological grade was assigned to each case of OSCC [Table 2]. The obtained cytological grade based on average frequency of MN was correlated with histopathological grade based on malignancy point score. The data obtained was statistically analyzed with the help of Student's t-test and Pearson's correlation.

Table 2

Histopathological grade of OSCC based on the malignancy point score

Results

The average frequency of MN in Group I and in Group II is summarized in [Table 3]. These observations give an unambiguous implication that the frequency of MN is three to four times higher in patients with OSCC as compared to the patients in control group. These results are also supported by statistical analysis [Table 4]. A highly significant difference was noted in the average frequency of MN in patients with OSCC as compared to subjects in the control group. PAP stained cytosmears revealed cells with MN ranging from one to five, with dissimilar sizes ranging from 1/3 to 2/3 the size of the nucleus of the cell [Figures 1 and 2].

Table 3

Frequency of micronuceli in oral squamous cell carcinoma OSCC (group I) and control group (group II)

Table 4

Comparison of frequency of micronuclei between oral squamous cell carcinoma (OSCC) (group I) and control (group II). Statistical analysis is student t test.

Figure 1

PAP stained cytosmears revealing cells with MN (×1000): (a) cells with only one micronucleus; (b) cells with two MN; (c) cells with three MN; (d) cells with four MN and (e) cells with five MN

Figure 2

(a and b) PAP stained cytosmears revealing epithelial cells with varying size of the MN (×1000)

When cytological grade (based on average frequency of MN) was compared with histopathological grade (based on malignancy point scoring) [Table 5], it was observed that out of 16, cytological grade of 13 subjects (75%) was correlating with the respective histopathological grade. This observation was statistically supported by Pearson's correlation indicating highly significant correlation between the cytological grade and histopathological grade [Table 6].

Table 5

Comparison of cytological grade (based on average frequency of micronuceli) and histological grade (based on malignancy point score)

Table 6

Statistical analysis for the comparison of cytological grade and histological grade (Spearman's correlation)

Discussion

MN could arise from the two basic phenomena in mitoticcells, which are chromosome breakage and the dysfunction of the mitoticapparatus. MN are formed from eccentric chromosomes, chromatidfragments, whole chromosomes or chromatids that lag behindin anaphase and are left outside the daughter nuclei in telophase. According to Tolbert et al.[4] MN are regarded as markers of abnormal mitoses involving chromosomal breakage and mis-segregated chromatin. Bloc hing et al.[5] were of view that MN assay can be regarded as an important biomarker to predict the relative risk of occurrence of cancer in the upper aero-digestive tract.

The International Collaborative Project on Micronucleus Frequency in Human Populations (HUMN) was organized to collect data on MN frequencies in different human populations and different cell types to determine the extent to which MN frequency is a valid biomarker of ageing and risk for diseases such as cancer.[6]

The hypothesis of a direct association between the frequency of MN in target or surrogate tissues and cancer development is supported by the findings like clear increase in the frequency of MN in target tissues as well as in peripheral lymphocytes in cancer patients.[7]

In the present study, making use of exfoliated buccal cells for assaying MN can be arguably explained on two bases. First, since approximately 90% of human cancers originate from epithelial cells, they represent a preferred target site for early genotoxic events induced by carcinogenic agents entering the body by way of inhalation and ingestion. The second reason is that the collection of buccal cells is certainly the least invasive method available for measuring the DNA damage in humans. This holds especially true in comparison to obtaining blood samples (for lymphocyte and erythrocyte assays) or obtaining tissue biopsies.[7] Exfoliated buccal mucosa cells can be easily collected using a wooden tongue depressor, a metal spatula, or a cytobrush moistened with water or buffer to swab or gently scrape the mucosa of the inner lining of one or both cheeks. Although cytobrushes appear to be most effective for collecting large numbers of cells, the high expense makes them less feasible for routine purposes. So, we preferred the use of a wooden spatula. Casartelli et al.[8] observed that MN frequencies were higher when cells were collected by vigorous, rather than by light, scraping.

Various methodological factors can affect the levels of MN in buccal cells. The main sources of variability may lie in the scoring criteria and staining procedures used. The effects of these factors on MN assay in the buccal cells have not been properly evaluated or quantified. There are various criteria given by various authors.[24910] According to Countryman et al.[2] and Sarto et al.[9], the criteria for identifying a structure as micronucleus is that it should be of the size less than 1/3 of the diameter of the associated nucleus, but still large enough to discern the shape and color. Whereas according to Belien et al.[10] the size should be less than 1/5 of the size of the parent nucleus.These two different views are creating a perplexity. So, in the present study, the criteria regarding assessment of size of MN have been modified.

Sarto et al.[9] described that a distinction between MN arising from chromosome breakage and spindle disturbances should be made by restricting the area of a micronucleus to 1/5 of the parent nucleus. Although a number of authors do not explicitly distinguish between the two types of MN, they in fact do distinguish them by their restriction on the diameter or area.[9]

After looking at this diverse origin for MN, it becomes very hard to believe that the size of MN is a constant value. Instead, it should be measured in terms of a range. Bigger MN result from exclusion of whole chromosome, following damage to the spindle apparatus of the cell (aneugenic effect), whereas smaller MN result from structural aberrations causing chromosomal fragments (clastogenic effect). Considering these facts, in the present study, the size of MN has been considered as ranging from 1/3 to 2/3 of the size of the nucleus. This proposition has also been supported by the observations of cells with varied sizes of MN [Figure 3].

Figure 3

PAP stained oral exfoliated cells showing metaphors of MN; (a) structures resembling MN; (b) but not in the same focal plane; (c) presence of vacuoles in the cell; (d) overlap of the nuclei of two adjacent cells; (e) presence of hyperchromatic structures in cell cytoplasm, resembling MN

Sarto et al.[9] have considered the cells from normal mucosa,in order to standardize the size of MN. But this criteria cannot be applied for precancerous and cancerous lesions, as it is very well known that in a cell undergoing atypia secondary to genetic damage, there is always an increase in the size of the nucleus and so also could be the size of micronucleus.[7]

The method of counting of MN also needs to be modified. In many previous studies a method of random screening of 1,000–10,000 cells has been proposed, irrespective of whether they contain MN or not.[157] According to these authors, an average frequency of MN was given as:

Average frequency of MN=Total number of MN/1000–10,000 cells.

All the cells in the denominator of above formula (1000-10,000) do not always necessarily contain MN. So, this denominator value does not hold to be significantly convincing and needs to be replaced by the total number of cells which in point of fact do contain MN. Consequently, we intend to put forward a modified formula as follows:

Average frequency of MN=Total number of MN/Total number of cells with MN.

Belien et al.[10] have concluded that at least 10,000 cells should be screened to monitor a significant reduction of 50% in the number of MN.The present study focused at screening of the whole smear for cells with MN in order to obtain a large sample size which directs us to a more reliable method of counting of MN.

The MN assay has been reported to correlate well with the histological grading of OSCC and leukoplakia. Incidence of MN has been analyzed by various studies in normal patients, oral premalignancies and OSCCs.[1] We detected that the average frequency of MN in patients with OSCC was ranging from 1.05 to 3.1 MN per cell. Although the method of counting of MN was modified as compared to previous studies, these results are correlating with the results of the studies conducted by Palve and Tupkari,[1] where the overall level of MN in the OSCC group was observed to be in the range of 1.1–3.0.

There was a significant correlation between the cytological grade (based on average frequency of MN) and the histological grade (based on malignancy point score) with coefficient of correlation (r value) being 0.654 and level of significance of 0.006 (P<0.05). This observation was similar to those reported by Kumar et al.[11] and Palve and Tupkari,[1] where the frequency of MN increased significantly from grade I to grade II to grade III, respectively, in squamous cell carcinoma group. Comparable results were obtained for increasing grades of OSCC in the present study.

Palve and Tupkari[1] have raised a question regarding subjectivity of histopathological grading system and its dependence on individual experience and assessment of microscopic observation. The reliability of correlation of frequency of MN and histopathological grading decreases if grading itself is not correct. So, they have suggested the new malignancy grading system given by Anneroth et al.[3] instead of other grading systems. As each observer was given the opportunity for counting (Group I and Group II), scoring and grading (in subjects diagnosed of OSCC) of all slides, without any exchange of information regarding case details, we strived to prevent the interobserver bias.

One more aspect regarding MN is the accurate identification of MN because many similes would create confusion and misguide the observer leading to defective count. Figure 3 reveals the various structures similar to MN. An easy way to exclude these is by alteration of the fine adjustment in microscope whereby these structures are lost. This indicates that these structures are not in one plane, and hence not to be considered as MN [Figures 3a–b]. Figure 3c further reveals vacuoles present in the cell, which can also create confusion with MN. Figure 3d exhibits an overlap of nuclei of two neighboring cells, resembling much like MN and has to be kept in mind while observation. Figure 3e displays the presence of structure similar to MN (arrow), but the hyperchromaticity is indicative of some inflammatory cell, probably a neutrophil. Many of such organizations should be differentiated from MN in order to attain a more valid count.

To summarize the results, MN can be seen in normal mucosal cells. Frequency of MN in oral mucosal cells of patients with OSCC was threefold to fourfold higher as compared with the control group. A 75% correlation was found between frequency of micronuclei MN and histopathological grade. Hence, it can be put forward that the frequency of MN in oral exfoliated cells of clinically suspected OSCC can be a candidate for histopathological grading of OSCC in the same subject.

Conclusions

There is a significant increase in the frequency of MN as compared to normal counterparts, indicating strong cytogenetic damage secondary to genotoxic and carcinogenic agents released by tobacco and areca products. The present study has revealed that there is a correlation of frequency of MN and histopathological grading in OSCC. These results should always be compared with a control. Thus, MN in oral exfoliated cells can be a candidate for grading of OSCC.

Some precautions and recommendations put forward are the following. The method of obtaining the sample should be standardized and repeatable. Complete smear needs to be screened for counting the frequency of MN for more valid results. The clarification on the size of the MN, as to whether to consider a constant value or a range, demands further studies.