CD10 for the distinct differential diagnosis of basal cell carcinoma and benign tumours of cutaneous appendages originating from hair follicle

Introduction

Firstly Headington classified benign tumours of cutaneous appendages originating from the hair follicle (BTCOHF) as germ layer hamartomas of the hair follicle, tumours of the hair follicle originating from the germ layer, tumours originating from the external layer of the hair follicle (trichilemmoma) and tumours originating from perifollicular mesenchyme in 1976 [1]. Ackerman et al. [2, 3] criticised those that were included in eight textbooks of dermatopathology in 2001. Recently, the World Health Organization (WHO) classified BTCOHF as two main subgroups: benign and malignant ones in the year of 2003 [1].

Differential diagnosis between BTCOHF such as trichoepithelioma (TE), trichoblastoma (TB), trichofolliculoma (TF), trichoadenoma (TA) and basal cell carcinoma (BCC) may be very difficult and bothersome for the clinician and the pathologist. The characteristic histopathology of TE is multiple horn cysts including concentrically laminated keratin or a less structured hyaline material. But the histological features of slowly growing, fresh-coloured, solitary papules, commonly reticulated cyst in the dermis surrounded by a mantle of cells or partly nodules usually located on the face which are very similar in both TE and BCC. Histological texture of both lesions consists of nests of basaloid cells inside the dermis [4, 5]. Trichoepithelioma and TB involve specialized tumour-associated stroma separating the nodules from the tumour and segregating the tumour epithelium from the enclosing dermis. Inversely, a specialized mesenchymal component is deficient for BCC. Besides, grooving of tumour epithelium and reticular dermis is present for that malignant tumour [6]. Trichofolliculoma is a rare tumour in any part of the skin [4] and a follicularly differentiated hamartoma mostly appearing during adulthood [7] without any sex predilection [8, 9]. Trichoadenoma is also a rare tumour, first determined in 1958 by Nikolowski [10] and seen as a nodular lesion usually on the face and buttocks [11, 12].

Basal cell carcinoma accounts for 65-75% of all skin tumours [13] and is the most common malignancy worldwide in white people [14]. The keratinous pattern of BCC resembles external stem sheath of the hair follicle rather than epidermis. The tumour may also have a follicular origin [15, 16]. Basal cell carcinoma is a low grade malignant tumour [17] and rarely metastasizes having the incidence of 0.0028% – 0.55% [18, 19]. Its recurrence for 5-year survival is 5% [19] and may show local invasion, especially the infiltrative and morpheic type [20]. So, it requires total excision and follow-up [5].

CD10 is a 100 kDa type II cell surface zinc metalloproteinase that is known as neutral endopeptidase, enkephalinase and common acute lymphoblastic leukaemia antigen (CALLA) which is included in inactivation of various biologically active peptides [21, 22]. Firstly it is determined as CALLA antigen and expressed on neoplastic cells such as Burkitt's and follicular lymphomas, lymphoblastic cell and chronic myelocytic leukaemias. It is also expressed on the surface of a wide variety of normal cells, such as fibroblasts, brush border of renal epithelial cells and enterocytes, glomeruli, myoepithelial cells of breast, bile canaliculi, hair follicles, eccrine and sebaceous glands [21, 23, 24]. For the skin, dermatofibromas, dermatofibrosarcoma protuberans, and melanomas as well as tumours and periadnexal mesenchymal cells in normal dermis are detected as the immunoreactivity areas of CD10 [25, 26].

CD10 focally stains papillary mesenchymal body, peritumoral stroma, and peripherally palisading cells. However, CD10 diffusely stains basaloid epithelium but not peritumoral stroma in BCC [5, 27, 28].

In the differential diagnosis of BTCOHF and BCC, histopathological character may not be enough and may be troublesome for pathologists, particularly in small biopsy specimens. So, immunohistochemical techniques can be useful for distinct diagnosis [29].

For these reasons, we compared staining patterns of BTCOHF (TE, TB, TF, TA) and BCC by using an immunohistochemical marker of CD10 in the current study which is the first one in literature to our knowledge as a group of BTCOHF.

Material and methods

Case selection

The investigation conforms to the principles outlined in the appropriate version of 1964 Declaration of Helsinki and approval of the present study was received from the Ethics Committee of Ankara Education and Research Hospital.

A total of 30 cases of BTCOHF (21 TE [Group 1a], 70%; 5 TB [Group 1b], 16.6%; 2 TA [Group 1c], 6.7%; 2 TF [Group 1d], 6.7%) and 30 cases of BCC were retrieved and analysed from the archives of the Department of Pathology at Ankara Education and Research Hospital where the cases of BTCOHF had been deposited between 2004 and 2008. The punch biopsies and incision biopsies, not enclosing the neighbouring epidermis and dermis were not included in the study.

Basal cell carcinomas were classified as nodular (18 cases, Group 2a, 60%), superficial (4 cases, Group 2b, 13.3%), infiltrative (2 cases, Group 2c, 6.7%) and mixed (6 cases, Group 2d, 20%) containing two or more types together. The staining patterns of CD10 are comparatively evaluated between both groups of BTCOHF and BCC totally and between all subgroups which belong to the same main group.

Immunohistochemistry

For the immunohistochemical evaluation, formalin (10% solution; pH 7.0-7.6) fixed, paraffin-embedded tumoral tissues were prepared. Then, a pair of 4 µm sections were placed on slides covered with poly-L lysine for each case. The original HE stained slides were detained for comparison with immunostained sections. The tissue sections were dried for 12 hours in a 37°C oven and then deparaffinized with xylene and rehydrated through graded alcohols. Antigen retrieval was performed by heating them under pressure with EDTA (ScyTek Laboratories, Logan, Utah, USA) for 9 min. The sections which were placed in the aforementioned solutions for 20 min at room temperature were then taken into phosphate-buffered saline (PBS) solution. Endogenous peroxidase was inhibited by incubation with 1% H₂O2 for 15 min. After washing of samples in PBS, they were incubated with Ultra V Block (ScyTek Laboratories, Logan, Utah, USA) for hindering non-specific binding. Each pair of the sections was incubated for 2 hours with mouse monoclonal antibody of anti-CD10/ CALLA (neutral endopeptidase) (Ab-2 clone 56C6 Neomarkers Fremont, CAS, USA) as a primary antibody at room temperature. Additional four-time washing with PBS was performed and it was followed by biotinylated UltraTek Anti-Polyvalent antibody (ScyTek) for 20 min as a second antibody. They were washed again in PBS and added to DAB (ScyTek Laboratories, Logan, Utah, USA) chromogene/substrate KIT for 5 min. The sections were counterstained with HE, then dehydrated in alcohols and cleared in xylene, and lastly, balsam was performed onto them and coverslip was mounted. Sections of the small intestine for CD10 were used for positive controls (i6000 automatic staining system Biogenex) which were performed for each case.

Evaluation and statistical analysis

All specimens were observed under a light microscope and the number of immunopositive tumour cells and stromal cells were evaluated using a scale of [0] to [2+] as follows: [0] negative (< 10% positive cells); [1+] regionally positive (10-50% positive cells); [2+] diffusely positive (> 50% positive cells). For the statistical analysis, SPSS – 13.0 was used, which is a computer statistical programme. All the data were expressed as means ± standard errors of means (SEM). Student’s t-test in the analysis of numerical variants, Pearson chi-square and Fisher's exact tests for comparing the rational data were used at the suitable areas. Pearson correlation analysis was used for the relationship between the numerical data and p-value less than 0.05 was considered as significant for all the tests.

Results

Patients with BTCOHF (14 females and 16 ma­les) ranged in age from 26 to 74 years (median 61.43 ±14.43) and patients with BCC (12 females and 18 males) ranged in age from 34 to 85 years (median 64.60 ±10.89) (Table I). While both tumour groups were localised on the head region, BTCOHFs were mostly detected on the nasal area and BCCs – on the eye circumference (Tables II, III).

Stromal, peripheral and central expression of CD10 in both tumours were graded from [0] to [2+] for each case. 19 of 30 cases (63.3%) of BCC and 11 of 30 cases of BTCOHF (36.7%) had stromal [0] immunoreactivity. 8 of 30 cases (26.7%) of BCC and 4 of 30 cases of BTCOHF (13.3%) had stromal [1+] immunoreactivity. 15 of 30 cases of BTCOHF (50%) and 3 of 30 cases (10%) of BCC had stromal [2+] immunopositivity. All degrees of stromal immunoreactivity of BTCOHF were significantly stronger than BCC concerning both numerical and degree of expression (p = 0.003) (Figures 1, 2). 14 of 30 cases of BTCOHF (46.7%) and 3 of 30 cases (10%) of BCC had peripheral [1+] immunoreactivity. The peripheral expression of CD10 for BCC was significantly stronger than BTCOHF regarding as [1+] immunopositivity (p = 0.03). On the contrary, 8 of 30 cases (26.7%) of BTCOHF and 5 of 30 cases of BCC (16.6%) had peripheral [2+] immunoreactivity. The peripheral [2+] immunoreactivity of BTCOHF was stronger than BCC, but not significant (Figures 3, 4). Central expression of CD10 was not significant for both groups.

CD10 expression of the subgroups of both groups is summarized in Tables IV and V and there was no significant difference between the subgroups.

Discussion

A large number of and various contributory laboratory techniques have been researched on the purpose of the correct differentiation between BCC and BTCOHF because their treatment and prognosis are also dissimilar. But they are continuing to be a diagnostic challenge, especially in a small and superficial biopsy.

CD10 is a very useful marker for the differential diagnosis of leukaemia and lymphoma [21] and has been used in the differential diagnosis of various epithelial and mesenchymal neoplasias recently. Renal cell carcinoma, hepatocellular carcinoma, urothelial carcinoma, and prostate carcinoma; endometrial stromal sarcoma, schwannoma, and stromal cells of ovarian epithelial cells express CD10 [30-32]. Ogawa et al. detected CD10 reactivity with p53 in stromal cells of colorectal tumours and Iwaya et al. assigned CD10 positivity in breast tumours, and then evaluated it for an important prognostic factor for recurrence and survival [32-34]. Similarly, Kazuhiro et al. decided in their study that CD10 positivity may be assigned as a marker for estimation of recurrence or invasion in BCCs. They observed CD10 immunoreactivity in tumour cells for less aggressive subtypes and in stromas for more aggressive subtypes [35, 36]. In the study of Pham et al. stromal expression of CD10 in TE and cellular expression of CD10 in BCC were significant. They correlated it with stromal features, patterns of the tumour development and discrepancy in host reactions [5, 37]. However, Costache et al. suggested that CD 10 was useless for displaying stromal reactivity [38].

In the current study, we evaluated the stromal, peripheral and central immunoreactivity of CD10 in BTCOHF and BCC grading from [0] to [2+] for each case. Stromal expression of CD10 for BTCOHF was significantly stronger than that of BCC regarding both the numerical and the degree of expression. Contrary to this, peripheral expression in BCC was significantly stronger than that of BTCOHF concerning [1+] immunopositivity. However, peripheral [2+] reactivity of CD10 in BTCOHF was stronger than the other, but not significant. Central expression of CD10 was not significant for both groups. In our study, stronger stromal staining in BTCOHF and stronger peripheral staining in BCC with CD10 were observed and these results were similar to Pham et al.’s ones [39]. Additionally, the outcome of staining of the same localizations supported the hypothesis of follicular differentiation of both groups, despite the degree of immunopositivity was different from each other.

There was no significant difference between the subgroups of both groups concerning CD10 immunoreactivity. When we analysed the subgroups of each group, we detected that the mixed type of BCC developing in nodular [+] infiltrative pattern had a tendency to obtain the stromal staining more densely according to the other subtypes of BCC. This result was supporting the study of Yada et al. [32], i.e. when capability of invasion increases in BCC, the tendency of stromal staining with CD10 may become possible.

In conclusion, we have determined in this study that CD10 may be useful for the differential diagnosis between BTCOHF and BCC as an immunohistochemical marker and it may solve a dilemma for the clinicians and the pathologists particularly in the small and superficial biopsies. In our opinion, studying with CD10 may even be a life-saving method for the selected cases. For all that, a limited number of the cases was the handicap of our study. However varied markers, dissimilar techniques, and a larger multicentred series are necessary for the distinct discriminations.

Acknowledgements

We would like to deeply thank our valuable personnel of the Pathology Laboratory of the Department of Pathology, Ankara Education and Research Hospital, namely Sefika Mercan, Kadriye Aglamaz, Hamdi Aydin, Haydar Kayabas and the others for providing us with useful and selfless assistance.

All authors contributed equally and were involved in writing the paper and finally approved the submitted and published versions.

References

 1. Cotton D. Troublesome tumours. 1: Adnexal tumours of the skin. J Clin Pathol 1991; 44: 543-548.  

2. Weedon D, Strutton G: Tumors of cutaneous appendages. In Skin Pathology. 2nd edition. Edited by Weedon D, Strutton G. Churchill-Livingstone; New York 2002; 859-901.  

3. Ackerman AB, Reddy VB, Soyer HP. Neoplasms with follicular differentiation. Ardor Scribendi, New York 2001.  

4. Simpson W, Garner A, Collin JR. Benign hair-follicle derived tumours in the differential diagnosis of basal-cell carcinoma of the eyelids: a clinicopathological comparison. Br J Ophthalmol 1989; 73: 347-353.  

5. Pham TT. CD10 Expression in trichoepithelioma and basal cell carcinoma. J Cutan Pathol 2006; 33: 123-128.  

6. Izikson L, Bhan A, Zembowicz A. Androgen receptor expression helps to differentiate basal cell carcinoma from benign trichoblastic tumors. Am J Dermatopathol 2005; 27: 91-95.  

7. Choi EH, Ahn SK. Hair follicul nevus. Int J Dermatol 1992; 31: 578-581.  

8. Mehregan AH. Hair follicle tumors of the skin. J Cutan Pathol 1985; 12: 189-195.  

9. Gray HR, Helwing HB. Trichofolliculoma. Arch Dermatol 1962; 86: 619-625.

10. Nikolowski W. [Trichoadenoma (organoid follicular hamartoma).] Arch Klin Exp Dermatol 1958; 207: 34-45.

11. Rahbari H, Mehragan A, Pinkus H. Trichoadenoma of Nikolowski. J Cutan Pathol 1977; 4: 90-98.

12. Krishna Swaroop DS, Ramakrishna BA, Bai SJ, Shanthi V. Trichoadenoma of Nikolowski. Indian J Pathol Microbiol 2008; 51: 277-279.

13. Ionescu DN, Arida M, Jukic DM. Meatstatic basal cell carcinoma. Four case reports, review of literature and immunohistochemical evalution. Arch Pathol Lab Medicine 2006; 130: 45-51.

14. Miller SJ. Aethiology and pathogenesis of basal cell carcinoma. Clin Dermatol 1995; 13: 527-536.

15. Roewert-Huber J, Lange-Asschenfeldt B, Stockfleth, et al. Epidemiology and aetiology of basal cell carcinoma. Br J Dermatol 2007; 157 Suppl 2: 47-51.

16. Leffel DJ, Fitzgeral DA. Basal cell carcinoma. Fitzpatrick’s Dermatology in General Medicine. Freadberg IM, Eisen AZ. Wolff K, Auston KF, Goldsmith LA, Katz SI, Fitzpatrick TB. McGraw Hill 1999; 1: 857-864.

17. Murphy GF, Elder DE. Basal cell carcinoma. In: Atlas of Tumour Pathology. Non-melanocytic Tumours of the Skin. Washington, DC: Armed Forced Institude of Pathology 1990; 47-60.

18. Lear JT, Harvey I, de Berker D, et al. Basal cell carcinoma. J R Soc Med 1998; 91: 585-588.

19. Brenn T Mc Kee PH. Tumors of the surface epithelium. In: Patholoy of The Skin. Mc Kee PH, Calonje E, Granter SR (eds.) Vol. 2 Elsevier Mosby, Philadelphia. 2005; 1167-1184.

20. Saldanha G, Fletcher A, Slater DN. Basal cell carcinoma: a dermatopathological and molecular biological update. Br J Dermatol 2003; 148: 195-202.

21. McIntosh GG, Lodge AJ, Watson P, et al. NCL-CD10-270: a new monoclonal antibody recognizing CD10 in parffin-embedded tissue. Am J Pathol 1999; 154: 77-82.

22. Sezaki N, Ishimaru F, Tabayashi T, et al. The type 1 CD10/neutral endopeptidase 24.11 promoter: functional characterization of the 5'-untranslated region. Br J Haematol 2003; 123: 177-183.

23. Knapp W, Lutz D, Bettelheim P. et al. Diagnostic specificity of the monoclonal anti-CALLA antibody VIL-A1 in leukemia and malignant lymphoma. Med Oncol Tumor Pharmacother 1984; 1: 25-32.

24. Dong HY, Gorczyca W, Liu Z, et al. B-cell lymphomas with coexpression of CD5 and CD10. Am J Clin Pathol 2003; 119: 218-230.

25. Kanitakis J, Bourchany D, Claudy A. Expression of the CD10 antigen (neutral endopeptidase) by mesenchymal tumors of the skin. Anticancer Res 2000; 20(5B): 3539-3544.

26. Kanitakis J, Narvaez D, Claudy A. Differential expression of the CD10 antigen (neutral endopeptidase) in primary versus metastatic malignant melanomas of the skin. Melanoma Res 2002; 12: 241-244.

27. Carr RA, Taibjee SM, Sanders DSA. Basolid skin tumors: basal cell carcinoma. Current Diagnostic Pathol 2007; 13: 252-272.

28. Carr RA, Sanders DSA. The use of immunohistochemistry in the differential diagnosis of common epithelial tumors of the skin. Current Diagnostic Pathol 2007; 13: 237-251.

29. Kirchmann TT, Prieto VG, Smoller BR. CD34 staining pattern distinguishes basal cell carcinoma from trichoepitelioma. Arch Dermatol 1994; 130: 589-592.

30. Bahrami S, Malone JC, Lear S, et al. CD10 expression in cutaneous adnexal neoplasms and a potential role for differentiating cutaneous metastatic renal cell carcinoma. Arch Pathol Lab Med 2006; 130: 1315-1319.

31. Chu P, Arber DA. Paraffin-section detection of CD10 in 505 nonhematopoietic neoplasms.Freguent expression in renal cell carcinoma and endometrial stromal sarcoma. Am J Clin Pathol 2000; 113: 374-382.

32. Yada K, Kashima K, Daa T, et al. Expression of CD10 in basal cell carcinoma. Am J Dermatopathol 2004; 26: 463-471.

33. Ogawa H, Iwaya K, Izumi M. Expression of CD10 by stromal cells during colorectal tumor development. Hum Pathol 2002; 33: 806-811.

34. Iwaya K, Ogawa H, Izumi M. Stromal expression of CD10 in invasive breast carcinoma: a new predictor of clinical outcome. Virchows Arch 2002; 440: 589-593.

35. Schirren CG, Rütten A, Kaudewitz P, et al. Trichoblastoma and basal cell carcinoma are neoplasms with follicular difeferentiation sharing the same profile of cytokeratin intermediate filaments. Am J of Dermatopathol 1997; 19: 341-350.

36. Hjalmar K, Lorenz E. Cytokeratins as markers of follicular differentiation. Am J Dermatopathol 2001; 23: 501-509.

37. Ramos-Ceballos FI, Pashaei S, Kincannon JM, et al. Blc-2, CD34 and CD10 expression in basaloid follicular hamartoma, vellus hair hamartoma and neurofollicular hamartoma demonstrate full follicular differentiation. J Cutan Pathol 2008; 35: 477-483.

38. Costache M, Bresh M, Böer A. Desmoplastic trichoepitelioma versus morpheic basal cell crcinoma:a critical reappraisal of histomorphological and immunohistochemical criteria for diffrentiation. Histopathology 2008; 52: 865-876.

39. Abdelsayed RA, Guijarro-Rojas M, Ibrahim NA, et al. Immunochemical evaluation of basal cell carcinoma using Bcl-2, Ki67, PCNA and P53. J Cutan Pathol 2000; 27: 169-175.



Address for correspondence

Ilker Sengul , General Surgery, MD, Asst. Prof.

The Founder Chairman of the Department of General Surgery

Vice Dean’s Office

Deanery of Giresun University Faculty of Medicine

28100 Giresun

Turkey

Vice-chancellor’s office: +90 454 310 10 00

Deanery: +90 454 214 14 96

GSM: +90 530 885 00 50

fax: +90 454 310 10 16

e-mail: dr.ilker52@mynet.com