Histomorphological Evidence of Complex Cannibalism—An Uncharted Territory in Oral Squamous Cell Carcinoma

Abstract

Complex cannibalism (CxC) is an emerging new seeming entity in oral squamous cell carcinoma (OSCC) embody as a prime metabolic event in determining the aggressive potential. Owing to paucity in literature regarding it, the aim of the present study to deciphering the occurrence of CxC in OSCC. Further, the expression of cluster of differentiation 68 (CD68) was studied in tumor cells and correlated with CxC to ascertain the biological behaviour of OSCC. 30 Hematoxylin and Eosin stained sections of various grades of OSCC were scanned for CxC and correlated with clinicopathological parameters. Immunohistochemical analysis using CD68 was performed. While observing CxC in OSCC, statistically significant in age (p-0.048) and histological grades (p-0.004). CD68 expression in tumor cells was statistically significant in histopathological grades (p < 0.001) whereas on correlating with CxC (p − 0.171) was non-significant. The analogous rise in CxC and CD68 with increasing histopathological grades could aid in recognising CxC as a precise histopathological parameter to assess the aggressive biological potential in OSCC.

Introduction

Oral squamous cell carcinoma (OSCC), a common head and neck malignancy warrants sixth place in its worldwide incidence [1]. Despite extensive diagnostic and therapeutic advancements, the current incidence has increased by 50% thus maintaining the static nature of the mortality and morbidity rates [2]. Grasping adequate knowledge about predictive factors for the progression of OSCC might shed a light on pertinent therapeutic measures there by improving survival. Owing to its multifactorial genesis, the pathogenesis of OSCC has been scrutinized in recent times with major focus directed in analyzing novel molecular markers which can serve in evaluating the biological behaviour and providing better treatment modalities [3]. In day to day practice, the selection of treatment strategies in oral cancer patients hugely relies on histopathological diagnosis. An exhaustive literature search showed that amidst a multitude of parameters cellular cannibalism (CC) is a modern valuable histological prognosticator which might aid in predicting the grading and biological behavior of OSCC [4, 5].

“Cannibalism” is derived from a Caribbean word “carib” which was first used to describe a native tribe in Brazil that participated in rituals where human meat was served [6]. CC is defined as the ability of a cell to engulf another sibling cell as well as cells from the immune system leading eventually to death of the internalized cell. The conventionally characterised cannibalistic cell has a crescent-shaped nucleus that engulfs another cell with a round to oval fading (smudged) nucleus [6, 7]. Several hypotheses suggests its cause to be either increased nutritional demand within the tumor microenvironment, tumor homeostasis or immune evasion [6–8]. Regardless of CC being an elementary histopathological parameter, the exact aetiology still remains ambiguous and it has never been fully explored in the bounds of oral cancers.

Complex cannibalism (CxC) is a term pioneered by Sarode et al. which represents a type of cellular cannibalism, in which one malignant cell engulfs another malignant cell and the complex which is formed is further devoured by a different cell. It can also present as one malignant cell engulfing two other tumor cells at a time. It was seen specifically in advanced stages and poorly differentiated oral squamous cell carcinoma and concluded that complex cannibalism could be suggestive of highly aggressive behaviour in OSCC [9]. Owing to its selective nature of expression in limited number of cases with increased aggressiveness CxC can be considered as a precise histopathological parameter better than CC indicating poorer prognosis.

The mechanism of CC features a diverse array of molecular markers including cluster of differentiation 68 (CD68), lysozyme, caveolin-1, actin, ezrin, cathepsin B etc. Among them a fairly established one is the genetic phagocytic marker CD68, is a 110-KDa heavily glycosylated type I transmembrane glycoprotein, which is mainly expressed by monocytes and macrophages and classified as member of the scavenger receptor family [8]. The human CD68 protein is encoded by the “CD68” gene which maps to Chromosome 17. Expression of CD68 antibody has been reported as an essential diagnostic and prognostic marker in tumor cells of various malignancies, such as breast carcinoma, prostatic carcinoma, renal cell carcinoma, melanoma, endometriosis, and even different soft tissue sarcomas [10–14]. CD68 depicts the phagocytic behaviour of the tumor cells and correlating it with the histopathological parameter of CxC might help us to attain more knowledge regarding the aggressiveness of OSCC.

Thus we have formulated the current study to assess the phenomenon of CxC and to correlate it with the phagocytic behaviour of the tumor cells by using CD68 antibody and its clinicopathological significance in order to determine the biological behaviour of OSCC.

Materials and Methods

The present observational study was conducted in the department of Oral Pathology and Microbiology, PGIDS, Rohtak post institutional ethical clearance (PGIDS/IEC/2019/21). 343 OSCC cases (2017–2021) were selected from the tissue archives out of which the 30 cases which displayed CxC were considered as study group. This included 15 cases of well differentiated oral squamous cell carcinoma (WDSCC), 8 and 7 cases of moderately differentiated oral squamous cell carcinoma (MDSCC) and poorly differentiated oral squamous cell carcinoma (PDSCC) respectively. Several forms of cannibalism (partial/complete) along with tumor cells engulfing the neutrophils, lymphocytes, RBCs, mitotic figures, plasma cells and keratin were also considered.

Identification of CxC

H&E stained sections were observed under high power (100X, oil immersion) using light microscope for assessing cannibalistic cells. The morphological features used to describe complex cannibalism proposed by Sarode S et al. [8] are:

1. one malignant cell was engulfing the other and this complex was further engulfed by another cell, or.

2. one cell was seen engulfing two cells at a time.

The counting was done in 10 random high power fields (hpfs), showing maximum density of this phenomenon by three observers. The average of ten different fields was calculated and expressed as number of mean CxC /Hpf [8]. Further arbitrary grading of CxC was done as follows:

• Grade 1: <3cells,

• Grade 2: 4–6 cells,

• Grade 3: >6 cells.

Immunohistochemical Analysis

For the immunohistochemical study, 3 μm sections were obtained from the formalin-fixed paraffin-embedded specimen on poly lysine coated slides. Immunohistochemical staining was carried out using the streptavidin-biotin-peroxidase complex method. The slides were incubated in primary antibody CD68 (1mL Conc. Dil,1: 90, cell marque) at room temperature for 1 h. Diaminobenzidine was used as chromogen. Tonsil was used as positive control for CD68.

Evaluation of CD68 Staining

The expression of CD68 was localized to the cytoplasm of malignant cells and cannibalistic cells. A semiquantitative method was used to score the CD68 (cytoplasmic and/or membranous staining) as previously described by Fedchenko N et al. Scoring was based on (a) intensity of immunostaining (0 = absent, 1 = mild, 2 = moderate, 3 = intense) and (b) the percentage of positive malignant cells (0 = no positive cells 1 = < 10% of positive cells 2 = 10–50% positive cells 3 = 51–80% positive cells 4 = > 80% positive cells). The final immunostaining score was determined by the multiplication of (a) and (b). Final scores ranged from 0 to 12 (0–1 = absent, 2–3 = mild, 4–8 = moderate and 9–12 = strongly positive) [15].

Statistical Analysis

All statistical analyses were performed using Microsoft Excel worksheet and SPSS Statistics for Windows Version 25.0 (IBM Corp., Armonk, New York, USA). As the data were qualitative and semi-quantitative, a two-sided p-value 0.05 was considered statistically significant in the non-parametric Chi-square and Mann Whitney test.

Results

In the present study, out of 30 cases, 26 (86.7%) were males while 4 (13.3%) were females, with male: female ratio being 6.5:1. Mean age of incidence was 51.47 ± 9.98 years with an age range of 30-72years. Mean age of male and female patients was 52.58 ± 9.83 and 44.25 ± 8.81 years respectively. Table 1 shows the association of Clinicopathological parameters and mean CxC in different grades of OSCC.

Table 1.

Association of clinicopathological parameters and mean CxC in different grades of OSCC

Characteristics	n (%)	Mean CxC (Hpf ± SD)	p Value
Total	30
Gender
Male	26	1.88 ± 1.33	0.531
Female	4	2.50 ± 2.38
Age (mean ± SD)
< 60 years	25	1.76 ± 1.13	0.048
> 60 years	5	3.00 ± 2.55
Habit
Smoking	18(60.0%)	2.00 ± 1.71	0.506
Smoking with alcohol	6(20.0%)	2.33 ± 1.36
Smoking and tobacco	5(16.7%)	1.60 ± 0.55
Tobacco chewing	1 (3.3%)	1.00 ± 0.00
Site
Buccal mucosa	6 (20%)	2.50 ± 2.07	0.697
Mandibular alveolus	9 (30%)	2.22 ± 1.86
Maxillary alveolus	3 (10%)	1.67 ± 0.58
Retromolar region	4 (13.3%)	2.25 ± 1.26
Floor of mouth	3 (10%)	1.33 ± 0.58
Lateral border of tongue	3 (10%)	1.00 ± 0.00
Palate	2 (6.7%)	1.50 ± 0.71
Grade
WDSCC	15 (50.00%)	1.40 ± 0.51	0.004**
MDSCC	8 (26.7%)	1.50 ± 0.53
PDSCC	7 (23.3%)	3.71 ± 2.21

** Tables denotes a highly significant nature of association that the p-value is less than 0.00

Association of Mean CxC with Age and Gender in Study Groups

Upon comparison of mean CxC and age was found to interfered a statistically significant difference (p value-0.048). No significant differences for frequencies were obtained when gender was correlated with CxC (p value − 0.531).

Association of Site of the Lesion and Mean CxC in Study Groups

Chi-square analysis of the site-wise distribution when comparing it with the mean CxC in different grades of OSCC was found to be statistically non-significant (p value − 0.697).

Association of Habits and Mean CxC in Study Groups

The association of these habits and mean CxC using chi-square test was found to yield a statistically non-significant difference (p value – 0.506).

Intergroup Comparison between Mean CxC and Different Grades of CxC

On comparison of the grades of OSCC with frequency of CxC, a statistically highly significant difference (p value − 0.004) was obtained (Table 2; Fig. 1). On applying Mann Whitney test, there was a statistically significant difference between well versus poor (p = 0.001) and moderate versus poor (p = 0.017) whereas there was no statistically significant difference between well versus moderate (p = 1.0) (Table 2; Fig. 1).

Table 2.

Intergroup comparison of frequency of CxC

Type of OSCC		Grade 1	Grade 2	Grade3	Intergroup p value	Overall p value
WDSCC	Count	15	0	0	WDSCC versus MDSCC- 1.000	0.004**
	%within total	57.7%	0.0%	0.0%
MDSCC	Count	8	0	0	MDSCC versus PDSCC-0.017*
	%within total	30.8%	0.0%	0.0%
PDSCC	Count	3	3	1	WDSCC versus PDSCC-0.002**
	%within total	11.5%	100.0%	100.0%

p value: > 0.05- Non-significant, < 0.05*-Significant, < 0.005**-Very Significant

Fig. 1.

Photomicrographs illustrating CxC in WDSCC (A), MDSCC (B), PDSCC (C), engulfing neutrophil (D), lymphocyte (E), RBCs (F), one malignant cell was engulfing the other and this complex was further engulfed by another cell (G), partial CxC (H), one cell was seen engulfing two cells at a time (I) (arrow indicates CxC, H&E;40X)

Association of CD68 Expression in Histological Grades of OSCC

For CD68 expression 33.33% cases were negative and 66.67% were positive in OSCC. There was a statistically significant difference seen for the frequencies among the groups (p value = 0.001) (Table 3; Fig. 2).

Table 3.

Association of CD68 expression in different grades of OSCC and CxC

Parameter	CD68
Grades		n	Mild	Moderate	Strong	P Value
OSCC	WDSCC	15	6 (75.0%)	2(25.0%)	0 (0.0%)	< 0.001**
	MDSCC	8	2 (25.0%)	5 (62.5%)	0 (0.0%)
	PDSCC	7	0 (0.0%)	1(12.5%)	4(100.0%)
CxC	Grade 1	26	8 (100.0%)	7 (87.5%)	2 (50.0%)	0.171
	Grade 2	3	0	1 (12.5%)	1 (25.0%)
	Grade 3	1	0	0	1 (25.0%)

** Tables denotes a highly significant nature of association that the p-value is less than 0.00

Fig. 2.

Photomicrographs illustrating CD68 immunoexpression in WDSCC (A), MDSCC (B), PDSCC (C). (IHC;40X)

Association between CxC and CD68 Expression in Different Grades of OSCC

Further the association between immunoexpression of CD68 and frequency Of CxC was assessed and found to yield a statistically non-significant difference (p value − 0.171) (Table 3).

Discussion

First introduced by Leydon in 1904, CC is an altered phagocytic event displayed by tumor cells all throughout the body [6]. Cannibalistic cells can be easily identified on routine (H &E) stained sections and no special advanced diagnostic techniques are required. A current concept given by Sarode et al. aims at shedding light on a refined pattern of CC known as Complex CxC [9]. Since CC has been previously explored in OSCC and CxC is an uncharted territory, our sole focus is to emphasize and understand the occurrence of the same in OSCC cases. Sarode S et al. studied the expression of CD68 in OSCC and concluded it as one of the possible marker for CC [8]. Furthermore, they reported the CD68 expression in OSCC in non-cannibalistic tumor cells which are rendered genotypically altered but not expressed phenotypically [16]. As reported by Sarode S et al. these genotypically altered tumor cells without morphological changes might represent the potential cells, which in future could convert into CC [8, 16]. Thus, CD68 might serve as a potential biomarker for early detection of potential CC and its phagocytic behaviour could point as a sign of aggressiveness in tumor mass.

In the present study, the mean age of incidence in OSCC exhibiting CxC was 51.47 ± 9.98 years. The trend of increasing mean CxC (3.00 ± 2.55) was found more in patients above 60years of age when compared to patients below 60years (1.76 ± 1.13) which was statistically significant. In previous studies by Siddiqui S et al. (2019) and Sarode S et al. (2016) on evaluating CC, the mean CC was found more in patients > 60years of age but without any statistically significant difference. The outcome of the present study is thus evocative of a more aggressive behaviour of OSCC in older age group than younger patients and serves as a better surveyor than CC. Since the present study is a pioneer attempt considering CxC in OSCC, we are unable to corroborate our results with any other elaborate study.

In the present study featuring CxC in OSCC, a male predominance was depicted which cements the paramount habit history of males in our region. While considering the mean CxC, it was found to be much higher in females than that of the corresponding males (F > M; 2.50 ± 2.38 > 1.88 ± 1.33) indicating an elevated aggressiveness of OSCC in females than males. Studies conducted by Siddiqui S et al. and Sarode S et al. in CC showed similar male preponderance and higher mean CC in females than males [8, 17]. However more studies with larger sample size that embrace CxC as an inclusion criteria are recommended to validate the present study findings.

In the present study, the most common site of involvement was noted in mandibular alveolus (9/30; 30%) followed by buccal mucosa (6/30, 22%). Placement of betel quid in the buccal mucosa in close proximity to the alveolar ridge causing sustained irritation could be associated with higher incidence of OSCC [18]. The mean CxC cases in buccal mucosa surpassed the cases in mandibular alveolus in our study. Association of anatomical site and frequency of CxC showed statistically non-significant results (p value = 0.697). Thus, the OSCC cases emerging from buccal mucosa are considered to be showing a higher aggressive potential. While observing CC, Sarode S et al. (2016) found maximum distribution in gingiva-buccal complex (26.78%) [8]. Owing to the limited number of cases for each site elaborated studies with higher sample size and uniform site distribution will help to validate the obtained results.

Predisposing risk factors were also studied with smoking being the most prevalent in 18/30 (60.0%) cases. Association of risk factors and mean CxC yielded statistically non-significant results (p value − 0.506). Although smoking has a higher prevalence in OSCC, yet increased frequency of mean CxC was found in patients with both smoking & alcohol consumption habit. While studying CC, Jose D et al. (2014) and Sarode S et al. (2016) observed 85% and 53% of cases with a positive history of tobacco respectively [6, 8]. So, we can imply that present study results might be due to the synergistic carcinogenic potential of tobacco and alcohol which could further correlate aggressive potential of OSCC.

While discussing various demographic details associated with CxC in our study, we compared it with the different studies focusing on CC. Despite elaborate researches conducted by several authors while inspecting cellular cannibalism, our study on CxC yielded better and more significant results stating CxC could be a far superior prognosticator than CC.

On observing CxC in various grades of OSCC, PDSCC revealed the highest value of CxC along with a mean value of 3.71 ± 2.21followed by MDSCC (1.50 ± 0.53) and WDSCC (1.40 ± 0.51) with statistically highly significant difference (p value = 0.004). These findings suggest that quantification of CxC is inversely proportional to degree of differentiation in OSCC.

Sarode S et al. (2012) has previously observed CxC considering only 5 cases (2-PDSCC and 3-MDSCC) and only two cases depicted presence of CxC belonging to PDSCC whereas no CxC was evident in MDSCC and further suggested that PDSCC could show a highly aggressive biological behaviour [9]. Sarode S et al. (2016) furthermore established the phenomenon of complex endocytosis in 9 cases of OSCC [8]. Another study by Jose D et al. stated that they could not find any bizarre cannibalistic cells termed as CxC in OSCC [6]. The contradiction in these above findings might be due to a variable sample size (n = 5,n-20 as compared to n = 30 ours). Jain M et al. focused on studying CxC in relation to metastasis and degree of differentiation in OSCC where, CxC was observed exclusively in metastatic type of MDSCC [7].

In the present study, out of 30 cases of OSCC, 4 displayed endocytosis of neutrophils. Furthermore, lymphocyte and plasma cell engulfed by tumor cells was reported in one case each (Fig. 1). There have been instances of both complete and partial CxC exhibited by the same case. We observed 29 cases depicting complete CxC whereas 6 cases illustrated partial CxC (Fig. 1). However, due to lack of adequate studies on CxC in OSCC, we are unable to compile and corroborate these findings. This can be resolved in the future with extensive studies using adequate sample size.

Given proportional increase of CD68 expression with increasing grades of OSCC and as CxC increases, the expression CD68 also increases in the tumor cells. Thus CD68 and CxC taken together might serve as potential phagocytic biomarkers and hence sign of aggressiveness. However, the results were not statistically significant which might be due to a small sample size. Sarode S et al. were the first one to observe CD68 expression in OSCC where they reported predominantly weak expression in WDSCC and MDSCC while strong positive expression was evident in PDSCC [8]. On the contrary, Sarode S et al. (2014) in OSCC observed mild to moderate positivity of CD68 expression of PDSCC tumor cells only at focal areas [19]. The contradictory findings in the above result might be due to small sample size (n = 7) and their inclusions of only those cases which exhibited neutrophil tumor CC. However, our study being the first to assess CD68 and CxC in conjunction we are burdened with paucity of adequate data to corroborate our findings thus instigating the need for further studies to validate these results.

Conclusion

The present study findings suggest that CD68 could be a useful indicator of aggressive behavior as represented by its increase in rate of immunoexpression during disease progression from WDSCC to PDSCC with associated increase in frequency of CxC. CD68 being a well-established phagocytic marker which helps in exploring the aggressiveness of OSCC, the parallel rise in CxC and CD68 could aid in recognising CxC as a precise histopathological parameter to assess the biological behaviour.

We were unable to obtain the information regarding the follow-up of the patients, metastasis and 5-year survival rate which is a vital confounding factor in the present study and would have clearly enhanced our study results. Forthcoming research with larger sample size with detailed clinicopathological prognosticators correlation is imperative to authenticate the results of present study in order to achieve better therapeutics. Although the present study put forth the correlation between CxC and CD68 in the various histopathological grades of OSCC, there is a distinct lacuna in incorporating other valuable factors like tumor budding, perineural invasion and depth of invasion which can better illuminate the aggressiveness of OSCC. The inclusion of complex cannibalism in routine pathology reports subsequently after validation should be taken into consideration.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declarations

Ethical Approval

The study was approved by the Institute’s scientific & ethical approval committee vide no. PGIDS/IEC/2019/21 dated 30/11/19.

Conflict of Interest

The author declares that they have no conflict of interest to disclose.