Immunoexpression of Ki67, P16 and Beta-catenin in precursor lesions of cutaneous squamous cell carcinoma

Abstract

Cutaneous squamous cell carcinoma (CSCC) is the second most common type of skin cancer, after basal cell carcinoma, representing about 10–20% of all malignant skin tumors. The mortality rates of CSCC approach those of renal and oropharyngeal carcinomas, as well as melanoma, with the increasing of the risk once metastases and perineural invasion occur. Both actinic keratosis (AK) and Bowen’s disease (BD) are direct precursors with the potential for progression to CSCC. In this study, we analyzed the expression of Ki67, P16 and Beta-catenin in the precursor lesions of CSCC in relation to histological prognostic parameters, respectively between them, with the aim of identifying possible correlations with a role in prognosis. Ki67 and P16 presented higher scores in advanced precancerous lesions, such as keratinocyte intraepithelial neoplasia (KIN) III and BD and low scores in seborrheic keratosis (SK). The immunoreactivity to the investigated markers confirms the multistage skin carcinogenesis, and their involvement starting from the initiation phase of the cancer process. The importance of the studied markers in the evolution and prognosis of precancerous lesions of CSCC is also supported by the linear correlations revealed between the immunoexpressions of P16, Ki67 and the membranous immunoexpression of Beta-catenin in AK.

Introduction

Cutaneous squamous cell carcinoma (CSCC) is the second most frequent type of skin cancer worldwide [1] and it usually appears in areas exposed to the sun. The cellular mechanisms underlying the tumor initiation and progression are very important for understanding the evolution and prognosis of this carcinoma. Both actinic keratosis (AK) and Bowen’s disease (BD) are direct precursors with progression potential to CSCC [2, 3, 4]. Seborrheic keratosis (SK), described as a benign skin proliferation caused by the delayed maturation of the keratocytes, also has a chance to evolve into CSCC [5]. Despite the low rate of progression, studies suggest that roughly 60% of CSCC appear in areas with preexisting AK, supporting the idea that these lesions are closely related etiopathogenically. Moreover, AK shows markers which are identical to those seen in CSCC [4]. For BD, considered a carcinoma in situ (CIS), the progression potential to invasive CSCC is estimated to be 3–5%. Although there is no exact epidemiological data about BD, incidence has increased in recent decades, with a higher frequency among the Caucasian people [6, 7].

The high incidence of CSCC and its precursor lesions supports research for the validation of markers involved in biomolecular mechanisms behind tumor progression and aggressiveness of these lesions.

Aim

The aim of our study was to investigate the immunoreactivity of the three precancerous lesions to the P16, Ki67, Beta-catenin markers and the possible existence of correlations between these immunoprofiles.

Materials and Methods

We investigated 46 precursor lesions for CSCC excised in the Dermatology and Plastic Surgery Clinics of the Emergency County Clinical Hospital, Craiova, Romania (40 cases of AK, three cases of BD, three cases of SK). The surgical excision samples were embedded in 10% neutral buffered formalin, included in paraffin, and stained using Hematoxylin–Eosin (HE). The classification of lesions was performed as recommended in the specialized literature according to the degree of lesion in the case of AK [8].

Then, we did serial sections from the corresponding paraffin blocks which went through immunohistochemical (IHC) processing using an amplification polymer-based detection system [polymer–Horseradish peroxidase (HRP) Histofine®, Nichirei, Japan, ready-to-use, code 414151F]. To visualize the reactions, we used 3,3’-Diaminobenzidine (DAB, code 3467, Dako) chromogen and for validation, we used positive and negative external controls (Table 1).

Table 1.

Antibodies used: clone, dilution, retrieval and external positive controls

Antibody	Clone/Manufacturer	Dilution	Antigen retrieval	External control
P16	DO-7/Dako	1:50	Tris-EDTA buffer, pH 9	Tonsil
Ki67	MIB-1/Dako	1:100	Citrate buffer, pH 6	Tonsil
Beta-catenin	Monoclonal β-catenin	1:50	Citrate buffer, pH 6	Liver

EDTA: Ethylenediaminetetraacetic acid

The Ki67 proliferative index (Ki67 PI) represents the average number of labeled tumor cells relative to the total number of cells in 10 microscopic fields (×40 objective), each field containing approximately 1000 cells.

Quantification of reactions for the anti-P16 antibody was done considering the percentage and intensity of reactions at the nuclear and cytoplasmic levels. Depending on the percentage of positive tumor cells, the cases were divided into the following categories: 0 (0–5% marked cells), 1 (6–25% marked cells), 2 (26–50% marked cells), 3 (51–75% marked cells) and 4 (>75% marked cells). Marking intensity was divided into four categories: 0 (absent), 1 (low), 2 (moderate) and 3 (high). The interpretation of the immunostaining was achieved by a composite score (CS), resulting from multiplying the intensity by the percentage. For statistical analysis, the CS was considered: negative for values of 0, low positive for values between 1–4 and high positive for values between 6–12.

For the Beta-catenin immunoexpression, both low expression or loss of membrane expression and ectopic nuclear/cytoplasmic expression have been proposed as altered expression. Immunoreactivity for Beta-catenin was assessed using CS of P16. Beta-catenin membrane composite score (M-CS) was calculated by the positive score of membranous stained cells × staining intensity score. Also, according to this method, the nuclear and/or cytoplasmic score (NC-CS) of Beta-catenin was determined in the studied cases.

The assessment of the reactions was done in parallel by two pathologists (C.M. and B.C.A.), the results being later compared and adjusted. The images were obtained by using the Motic Panthera DL microscope, equipped with Motic Images Plus 3.0 ML software.

For the statistical analysis were used comparison tests represented by χ2 (chi-squared) and Pearson within the Statistical Package for the Social Sciences (SPSS) 10 software, the results being considered significant for values of p<0.05. In this study, for the calculation of average values and standard deviations, there were used numerical values of the obtained immunostains for all the cases, including the negative ones. In the scientific research, the ethical aspects were respected, based on the informed consent of the patients, the study being approved by the Local Ethics Committee (Approval No. 269/29.11.2023).

Results

Most cases of AK were diagnosed in female patients, with 27 (67.5%) cases and the average age of diagnosis was 68.63 years. The most common location for CSCC was the head (85%), followed by the neck (two cases, 5%), the upper limbs (two cases, 5%), lower limbs (one case, 2.5%) and trunk (one case, 2.5%) (Table 2). Analysis of the histological grade of AK cases revealed the presence of all three severity categories of AK. The most frequent were keratinocyte intraepithelial neoplasia (KIN) I (15 cases, 37.5%) and KIN II (15 cases, 37.5%) lesions, followed by KIN III (10 cases, 25%). KIN I lesions were histologically characterized by cellular atypia of the basal keratinocytes, the change being limited to the lower third of the epidermis. In KIN II lesions, the atypical keratinocytes occupied the lower two thirds of the epidermis, sometimes with the presence of dysplastic buds in the papillary dermis, and in the case of KIN III lesions, the atypical keratinocytes occupied the entire epidermis.

Table 2.

Precancerous lesions, epidemiological data and Ki67, P16, Beta-catenin immunoexpression

lesion	No. of cases	gender	Topography	ki67 PI [%]	p16 CS	Beta-catenin
Male	Female	Head	Neck	Upper extremities	Lower extremities	Trunk	M-CS	NC-CS
KIN I	15	3	12	34	2	2	1	1	9.53±8.33	1.66	8.06	0.06
KIN II	15	5	10	19±8.37	1.80	5.66	3.8
KIN III	10	5	5	19.5±11.55	2.30	2.2	7.8
Bowen’s disease	3	1	2	1	0	2	0	0	83.3±12.58	9.66	1.33	9.66
Seborrheic keratosis	3	2	1	2	0	1	0	0	8.33±2.33	1.33	10.66	0
p-value									<0.05	<0.05	<0.05	<0.05

Ki67 PI: Ki67 proliferative index; KIN: Keratinocyte intraepidermal neoplasia; M-CS: Membrane composite score; NC-CS: Nuclear–cytoplasmic composite score; P16 CS: P16 composite score

Two of the three cases of BD were diagnosed in female patients and were identified at the level of the upper limb, and the average age of diagnosis was 60 years (Table 2). Histologically, the BD was characterized by the presence of keratinocytes with marked atypia, with cellular and nuclear pleomorphism, with enlarged nuclei, hyperchromia and intense mitotic activity, throughout the thickness of the epithelium. In addition, we observed the loss of polarity and maturity, giving the epidermis a disordered histological aspect.

Out of the three cases of SK, two cases were diagnosed in male patients, with localization on the head and the average age of diagnosis was 71 years (Table 2). The microscopically analyzed cases showed central nodular formation formed by a proliferation of keratinocytes, some with a basaloid appearance, with acanthosis, papillomatosis, and hyperkeratosis and with formation of corneous pseudocysts in the thickness of the lesion. Skin appendages (sebaceous glands, hair follicles) were present, with normal microscopic appearance.

Immunoexpression of Ki67

Ki67 immunostaining showed positivity in 80% of KIN I cases, 86.66% of KIN II cases and 90% of KIN III cases. The immunostaining was nuclear positive in some of the cells in the basal layer and was rarely present in the upper layers of the epidermis. The mean value of Ki67 PI in these cases was 9.53±8.33% for KIN I, 13±8.37% for KIN II and 19.5±11.55% for KIN III (Table 2; Figure 1A).

Figure 1.

Immunoexpression of Ki67: (A) AK, ×100; (B) BD, ×100; (C) SK, ×100; (D) Ki67 immunostaining values distribution. AK: Actinic keratosis; BD: Bowen’s disease; KIN: Keratinocyte intraepidermal neoplasia; SK: Seborrheic keratosis.

The three cases of BD were positive for Ki67. The immunoexpression of Ki67 was nuclear, with distribution throughout the thickness of the epidermis and a mean PI of 83.3±12.58% (Table 2; Figure 1B).

The three cases of SK were positive for Ki67. Ki67 immunoexpression analysis revealed nuclear immunostaining in rare cells, with distribution throughout the thickness of the epidermis and a mean PI of 8.33±2.33% (Table 2; Figure 1C).

The increased Ki67 immunoexpression was significantly associated with CSCC precursor skin lesions, noting an increase in the percentage of immunolabeled cells in cases with KIN III and BD lesions (Figure 1D).

Immunoexpression of P16

The investigation of P16 immunoexpression revealed the presence of reaction in 80% of KIN I cases, 86.66% of KIN II cases and 80% of KIN III cases, being present at nuclear and cytoplasmic levels, with the variation of intensity depending on KIN. The average value of the P16 CS was 1.66 for KIN I, 1.80 for KIN II and 2.30 for KIN III (Table 2; Figure 2A).

Figure 2.

Immunoexpression of P16: (A) AK, ×100; (B) BD, ×100; (C) SK, ×100; (D) P16 immunostaining scores distribution. AK: Actinic keratosis; BD: Bowen’s disease; P16 CS: P16 Composite score; KIN: Keratinocyte intraepidermal neoplasia; SK: Seborrheic keratosis

All BD cases analyzed were positive for P16 with immunostaining throughout the thickness of the epithelium at the nuclear and cytoplasmic level and an average value of the P16 CS of 9.66 (Table 2; Figure 2B).

All analyzed cases of SK were positive for P16, with cytoplasmic and nuclear immunolabeling with an average CS of 1.33 (Table 2; Figure 2C).

Statistical analysis revealed significant associations between high P16 CS and BD (p<0.05) (Figure 2D).

Immunoexpression of Beta-catenin

The membrane immunoexpression of Beta-catenin was modified in all cases of AK, with average M-CS of 8.06 for KIN I, 5.66 for KIN II and 2.2 for KIN III. Cytoplasmic and/nuclear Beta-catenin immunoexpression was identified in 23 (57.5%) cases of AK, of which one (2.5%) case of KIN I, 12 (30%) cases of KIN II and 10 (25%) cases of KIN III (Table 2; Figure 3A), with an average NC-CS of 0.06 for KIN I, 3.8 for KIN II and 7.8 for KIN III.

Figure 3.

Immunoexpression of Beta-catenin: (A) AK, ×100; (B) BD, ×100; (C) SK, ×100; (D) Beta-catenin immunostaining membrane scores distribution. AK: Actinic keratosis; BD: Bowen’s disease; KIN: Keratinocyte intraepidermal neoplasia; M-CS: Membrane composite score; SK: Seborrheic keratosis

Beta-catenin immunoexpression in BD was altered, by loss of membrane immunoexpression or abnormal nuclear/cytoplasmic immunoexpression in all three BD cases, with a Beta-catenin M-CS of 1.33, respectively Beta-catenin NC-CS of 9.66 (Figure 3B).

Beta-catenin immunoexpression at the nuclear/cytoplasmic level was absent in all cases of SK, and membrane immunoexpression showed an M-CS of 10.66 (Figure 3C).

Statistical analysis revealed significant associations between increased of Beta-catenin immunoexpression at the nuclear/cytoplasmic level in KIN III type lesions and BD (p<0.05), respectively between increased membrane immunoexpression in early precancerous KIN I lesions (Figures 3D, 4).

Figure 4.

Beta-catenin immunostaining nuclear/cytoplasmic scores distribution. BD: Bowen’s disease; KIN: Keratinocyte intraepidermal neoplasia; NC-CS: Nuclear–cytoplasmic composite score; SK: Seborrheic keratosis.

In this study, for cases of AK, we reveal a statistically significant positive linear correlation for Ki67 and P16 (p=0.026, Pearson’s test), a statistically significant negative linear correlation for Ki67 and Beta-catenin with membrane immunoexpression (p=0.013, Pearson’s test) and a statistically significant negative linear correlation for P16 and Beta-catenin with membrane expression (p=0.007, Pearson’s test) (Figure 5).

Figure 5.

Distribution of the percentage values of the expression of Ki67, P16 and membranous Beta-catenin in AK. AK: Actinic keratosis

For BD, we noticed positive linear relationships between P16 and Ki67 immunoexpression, and also negative linear relationships with nuclear/cytoplasmic Beta-catenin immunoexpression, but they were not statistically significant (p>0.05, Pearson’s test).

Discussions

CSCCs are most often the consequence of the malignant transformation of preneoplastic skin lesions as a result of the accumulation of genetic mutations, following multiple exposures to type B ultraviolet radiation. The cellular mechanisms underlying the initiation and progression of squamous cell carcinoma (SCC) are of great importance for understanding its prognosis and evolution. KA, BD, and SK are precursor lesions of CSCC, the first two being considered prophase lesions of CIS, with the possibility of evolution towards frank CSCC. AK is the most frequent precursor lesion of CSCC, estimates from the literature indicate that 20% of lesions evolve malignantly [9]. The risk of progression to SCC of BD is 3–5%, being higher in the case of lesions located in the skin of the neck [10].

Like other types of cancer, the development of CSCC is a multistep process involving the sequential acquisition of genetic alterations. Ki67, a high molecular weight non-histone protein, is usually accepted as the most reliable marker of cell proliferation [11], which is correlated with tumor growth and metastatic potential [11]. In our study, Ki67 expression was identified in all categories of studied lesions, thus we observed the positivity of the reaction in 34 (85%) cases of AK, in all three cases of BD and all three cases of SK. Also, the percentage of immunolabeled cells was significantly higher in cases of BD and KIN III type lesions, compared to KIN I and SK. An increase in the number of immunolabeled cells is also observed with the increase in the degree of keratinocyte dysplasia and the appearance of malignancy, which suggests a greater proliferative activity. Another study in the literature obtained significant increases in Ki67 immunoexpression in cases of CSCC and AK compared to normal skin, but without significant differences in immunoexpression between CSCC and AK [12]. Other studies have obtained results similar to ours, with significant increases in Ki67 immunoexpression in advanced precursor cancer lesions, such as BD [13]. Contrary to our results, Talghini et al. found no significant differences between the three pathological conditions (BD, SCC, AK) in terms of Ki67 expression [14].

P16 is a tumor suppressor gene, whose inactivation by deletion, mutation, or methylation [15] is considered a major oncogenic event in the carcinogenesis of CSCC but is observed in a wide range of cancers. P16 expression is strongly associated with HPV infection, and therefore IHC expression of p16 is considered a surrogate marker of oncogenic HPV infection [1166]. In our study, the P16 analysis indicated the presence of expression in all categories of lesions. Thus, we observed the positivity of the reaction in 33 (82.5%) cases of AK and in all cases of BD and SK. Blokx et al. reported a higher P16 expression in high-grade AK lesions compared to low-grade lesions [17]. Hodges & Smoller reported P16 expression in 100% of cases of AK, the staining being weak to moderate, located in the lower half of the epidermis, as well as P16 expression in 100% of CSCC, with moderate to intense intensity [7]. On the contrary, other studies showed the absence of the P16 marker in non-Bowenoid AK [18] and positivity in 60% of CSCC [19]. However, since the majority of AK lesions do not progress to CIS or invasive carcinoma, the authors pointed out that the overexpression of P16 seems to be necessary, but not sufficient for tumor progression, this transformation also requiring the involvement of other factors [7]. In our study, P16 immunostaining indicated significant associations with BD and KIN III, which showed a predominantly high CS P16. Other studies have found higher expression of P16 in BD compared to KIN lesions [20].

In addition, activation of the Wnt/Beta-catenin pathway involves an accumulation of Beta-catenin in the cytoplasm, which subsequently leads to its nuclear translocation, then to increased Beta-catenin-dependent transcription, especially of the c-myc and cyclin D1 genes. Uncontrolled cell growth is thought to be the key step in mediating cellular malignancy. Beta-catenin comprises some additional functions, such as maintaining the normal morphotype and mediating intercellular adhesion when combined with E-cadherin to form a complex [21]. Loss of cell membrane expression of this protein will impair the intercellular adhesion system, facilitating tumor cell metastasis. This study showed that Beta-catenin is continuously expressed on the cell membrane and the continuous staining of cells in the basal layer and spinous layer to some extent reflects its adhesion property and function in normal epidermal tissues. Xuan et al. demonstrated that there was no significant difference between cells from normal keratinocytes and benign tumors (SK, common warts, or keratoacanthoma) regarding the expression of Beta-catenin on the cell membrane. By contrast, malignant cancers including BD, basal cell carcinoma, and SCC show a decrease in membrane expression and an increase in cytoplasmic immunoexpression or nuclear expression. Therefore, membrane expression of Beta-catenin may correlate with keratinocyte differentiation [21, 22].

Other studies suggest that a decrease in membrane expression of Beta-catenin is observed in CSCC, then in AK and BD [23]. Like to these results, our study indicates that the membrane expression of Beta-catenin showed high M-CS in SK and KIN I type lesions, and the nuclear/cytoplasmic immunoexpression was increased in the cases of KIN III and BD, with increased NC-CS. Based on these data, it is suggested that the abnormal expression of Beta-catenin would probably increase the degree of cellular atypia, therefore, it would have critical roles in the formation of tumor cells from epidermal keratinocytes. The dual roles, in cell adhesion and cell signal transduction respectively, demonstrate the involvement of Beta-catenin in cancer progression.

In our study, we also observed statistically significant positive linear correlations between P16 and Ki67 immunoexpression, and negative linear correlation between the membrane expression of Beta-catenin and Ki67, respectively with P16 immunoexpression. Contrary to our results, other studies showed that in carcinogenesis, tumor progression was associated with tumor proliferative fraction (Ki67-positive cells) and tumor differentiation, but without correlation with P16 expression [24].

Conclusions

In this study, we obtained significant associations between high CS of P16, Ki67 immunoexpression and nuclear/cytoplasmic immunoexpression of Beta-catenin and advanced precancerous lesions KIN III and BD. These results show the importance of the investigated markers in the evaluation of the aggressiveness and prognosis of precursor lesions of CSCC. This idea is also supported by the significant linear relationships obtained between P16, Ki67 and cytoplasmic nuclear immunoexpression of Beta-catenin in AK lesions, this panel of antibodies being useful in the stratification of precancerous lesions of CSCC.

Conflict of interests

The authors declare that they have no conflict of interests.

Source of funding

This work received financial support through the Project “Correlations between dermatoscopic, histopathological and immunohistochemical aspects in keratinocyte precancers precursors of squamous cell carcinoma”, financed by the Amaradia Polyclinic, Craiova, Romania.