Summary: Basal Cell and Squamous Cell

Carcinoma

•Basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) are easily recognized in frozen section with H & E or toluidine blue during Mohs, but in cases where there is extensive inflammation, fibrosis, or perineural and muscle invasion, immunostaining may be necessary to locate residual tumor.

•Cytokeratin stains, such as AE1/AE3, can be used to identify epidermal-derived tumor cells within the background of stromal cells and collagen.

•Cytokeratin immunostains generally stain both BCC and SCC, while Ber-EP4 staining is mostly for BCC.

15.4Basal Cell and Squamous Cell Carcinoma

Determination of margins of resection of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) with Mohs is generally straightforward and based entirely on H & E staining on FS material. In a 2001 survey, dermatologic surgeons reported the use of H & E in 87% of BCC and 95% of SCC, and toluidine blue was used for 13% of BCC and 5% of SCC [31].

There may be, however, a role for FS immunostaining in difficult or indeterminate cases (Fig. 15.6). In a study of 400 patients with both BCCs and SCCs, 36 frozen sections were deemed difficult to interpret due to a dense inflammation, perineural invasion, or subtle morpheaform infiltration. The authors utilized the broad-spectrum antibody cytokeratin AE1/AE3 and cytokeratin 14 to assist in determination of margins. The processing time took an hour, but they found it useful in identification of morpheaform cords of tumor and perineural infiltration; furthermore, it was particularly useful in unmasking the tumor from the dense inflammation [32].

Immunostains commonly used for BCCs and SCCs include cytokeratin (CK) stains and Ber-EP4. Cytokeratins are intermediate-filament protein polymers that are present in nearly all animal cells. They are responsible for maintenance, communication between adjacent epithelial cells, and protection from trauma

[29, 33]. Currently, there are 20 CKs, classified by their molecular weight or pH. Cytokeratin 9–20 are Type 1, acidic keratins that are encoded by chromosome 17q12–21, and CK 1–8 are Type 2, basic keratins regulated by genes located on chromosome 12q11–13 [33]. Cytokeratin antibodies are useful in differentiating between epidermal and mesodermal tumor cells [29]. BCCs express CK 5, 14, 15, and 17, and SCCs express 5, 6, 8, 14, 17, and 18 [6]. The CKs that immunostain AE1 detects are high molecular weight (HMW) CK 10, 14, 15, 16 and low molecular weight (LMW) CK 19; while AE3 detects HMW CK 1, 2, 3, 4, 5, 6 and LMW CK 7 and 8 [2, 6]. Combining AE1 and AE3 will help stain most SCCs and BCCs regardless of depth and differentiation, as the keratin expression is still present [2].

The initial studies with cytokeratin immunostaining for Mohs conducted in the mid 1980s were promising, but the processing, took several hours [4]. A 1-h protocol was developed years later utilizing AE1. Twenty patients with extensive, recurrent, or aggressive SCC underwent MMS. In the final stage, all slides were confirmed negative by H & E and underwent staining with AE1. Eight of twenty patients were found to have tumor hidden in either dense inflammation or perineural/muscle invasion. The authors noted that because adnexal structures are also cytokeratin positive, they can be a source of confusion when interpreting margins [34]. In another study, the AE1/AE3 immunostaining protocol was shortened to only 19 min. The results in FS were comparable to permanent sections stained with H & E and AE1/AE3. The authors found that cytokeratin immunostaining was particularly useful in areas of dense inflammation and perineural involvement, and avoided unnecessary additional layers and time [2]. An anticytokeratin stain, MNF116, which detects CK 5, 6, 8, 17, and 19, along with p63, may be helpful in cases of poorly differentiated SCCs that present with an atypical, infiltrative single-cell pattern. In one study, MNF116-stained slides were also compared to matched H & E “cleared” slides. Only 1 of 143 slides was mismatched, a recurrent aggressive growth BCC, which stained positive for MNF 116 while negative for H & E. They concluded that H & E is sufficient for the majority of BCC, but cytokeratin immunostaining may be a useful adjunctive for more aggressive subtypes of tumor [35]. Another possible use for cytokeratin staining may be for Mohs in patients with hematological malignancies such as chronic lymphocytic leukemia (CLL). Due to the

Fig. 15.6 (a) H & E–stained frozen section of a margin during Mohs surgery for BCC with scar tissue and dense inflammation within the panniculus that could mask tumor cells (40×). (b) Frozen section stained with AE1/AE3 CK immunostain confirms that no

tumor is buried within the infiltrate (40×), indicating a negative margin for basal cell carcinoma. In the control specimen, AE1/ AE3 CK immunostaining demonstrates lobules of tumor within the dermis in frozen section at (c) 10× and (d) 20× magnification

abundant leukemic infiltrate, it may be difficult to find the tumor [36].

Ber-EP4, a monoclonal antibody directed against two 34 kDa and 39 kDa glycopeptide chains, stains BCC but not SCC [37]. It is expressed in all normal epithelial tissues, preferentially at the basement membrane in a basolateral location, in both cutaneous and noncutaneous tumors [37]. It also stains adnexal epithelium, such as eccrine and apocrine glands and ducts,

and the base of the hair bulb, but does not stain the keratinocytes of the remainder of the follicular unit [38]. In one study, 43 cases of BCCs and SCCs were stained with Ber-EP4, and Ber-EP4 was positive in all of the BCCs, regardless of differentiation, while staining none of the SCCs. In addition, the authors noted that the general lack of hair follicle staining was useful in differentiating certain hair follicles that may otherwise be confused with a BCC [38].

Some authors found Ber-EP4 useful in unmasking tumors surrounded by inflammation. In one study, 27 slides of BCCs were stained with H & E and Ber-EP4. All slides stained with Ber-EP4, and 13/27 slides provided an enhanced view of the tumor compared to H & E. In addition, 2 of 27 slides revealed a morpheaform BCC while the H & E interpretation was negative [39]. Conversely, a group found that the use of Ber-EP4 actually resulted in cases of overdiagnosis of BCC where there were dense inflammatory infiltrates. They found that in 48% of cases that were Ber-EP4 positive, no tumor could be identified when the slides were reviewed. In fact, when tumors were surrounded by dense inflammation, the tumors could be revealed by deeper sections. Infiltrative BCC were found to be most closely associated with dense infiltrates (73%), while nodular BCC were least associated (30%). Because the authors were able to detect all cases of tumor surrounded by dense inflammation with deeper H & E sections, it was concluded that Ber-EP4 was not necessary in “unmasking” the tumor [40].

While Ber-EP4 stains primarily the base of the hair bulb, many authors have detected “patchy” staining in other portions, making it difficult to distinguish BCC from folliculocentric basaloid proliferation (FBP) [39– 41]. One study explored the use of monoclonal antidesmoglein antibody, 33–3D, which recognizes the cytoplasmic domain of desmoglein, which is conducted in a 60-min protocol. In 18 biopsy-proven BCC patients where it was difficult to differentiate FBP from BCC, the 33–3D stain was utilized. In 4/18 patients, the perimembranous staining pattern typical of normal follicles was absent and instead revealed a confluent cytoplasmic staining pattern leading the authors to take another layer. A follow-up period of 6–24 months resulted in no recurrence [41].

Differentiating BCC from trichoepithelioma (TE) is also a diagnostic challenge. Both tumors are comprised of basaloid strands of epithelium embedded in fibrocellular stroma. BCC is associated with characteristic peripheral palisading, stromal retraction, and an increased mitotic rate, among other diagnostic findings, while trichoepithelioma is more lobular and associated with papillary mesenchymal bodies. Twenty cases each of BCC and TE were stained with the immunostains Bcl-2, p53, Ki67, and PCNA. Bcl-2 and p53 were not useful in differentiating both tumors, while Ki67 and PCNA demonstrated more diffuse and frequent staining in BCC and less frequent and peripheral staining of TEs [42].

Summary: Dermatofibrosarcoma Protuberans

•Dermatofibrosarcoma protuberans (DFSP) is a locally aggressive fibrohistiocytic tumor with a high recurrence rate following wide local excision.

•CD34 is commonly used to differentiate DFSP from dermatofibroma, keloids, and sarcomas such as malignant fibrous histiocytoma and atypical fibroxanthoma. It is used in Mohs to find the edge of the DFSP for margin assessment.

15.5Dermatofibrosarcoma Protuberans

Dermatofibrosarcoma protuberans (DFSP) is a locally aggressive fibrohistiocytic sarcoma that accounts for approximately 0.1% of all malignancies [43]. This tumor can be a challenge to remove because it can have projections of tumor that can extend in all directions at the periphery, between collagen bundles, and even invading muscle and fascia. This tumor has recurrence rates as high as 43–50% after wide local excision [43, 44]. MMS may provide an opportunity for tissue sparing, while identifying areas of extension. Recurrence rates in the literature after MMS have been reported to be 0–6.6% [43, 45].

CD34 immunostaining may be helping in discerning tumor margins in difficult cases of DFSP. CD34 is a 115 kDa single chain transmembrane glycoprotein found on normal human hematopoietic progenitor cells, endothelial cells, and a characteristic population of dermal dendritic cells [46, 47]. It is a helpful stain in differentiating DFSP from dermatofibroma (DF), malignant fibrous histiocytoma (MFH), atypical fibroxanthoma (AFX), and keloids. The sensitivity of utilizing CD34 in some reports for FFPES is 84–100%, as authors have found it does not uniformly stain all DFSP, and on occasion can stain DFs leading to false positives [44].

An anti-CD34 mouse monoclonal antibody immunostain was reportedly utilized in MMS in a difficult case in order to discern the tumor from the adjacent stroma to provide reassurance that the margins were negative. The authors recommended that a positive control, such as tissue from the first stage, be used initially to determine CD34 positivity [47]. Interestingly, in

another case where the FFPES was negative for CD34, the frozen section slide revealed strong positive staining. The authors caution that while it is a useful stain, there is considerable variability of staining for CD34, therefore recommending that CD34 negative slides be sent for paraffin-embedded sections as well [46].

Summary: Extramammary Paget’s Disease

•Extramammary Paget’s disease (EMPD) is an adenocarcinoma usually located in the skin of the anogenital and axillary region.

•Complete excision of EMPD may be difficult, and despite numerous surgical procedures and significant morbidity, the recurrence rate is still high.

•CK7 and CEA are useful markers during MMS for EMPD and differentiate Paget’s cells from atypical keratinocytes and melanocytes.

visualize 100% of the margins through his technique provided a reliable alternative to radical surgery to allow for tissue sparing. In addition, he noted that Paget’s cells in the adnexal structures should not be considered invasive malignancy [49].

Subsequent studies have shown that recurrence after MMS for EMPD has been at least 23%, compared to 33% for standard surgical excision [51]. One study noted that the recurrence rate for primary and recurrent tumors is 16% and 50%, respectively. The average margin required to clear 97% of patients in this study was 5 cm [50]. The mean time to recurrence after MMS has been reported to be 2.5 years; therefore, long-term monitoring of the patient is necessary to assess efficacy [51].

Paget’s cells can stain with carcinoembryonic antigen (CEA), epithelial membrane antibody (EMA), gross cystic disease fluid protein (GCDFP), and LMW cytokeratins such as CK 7, 8, 18, and 19 [48, 52]. These cytokeratins are found in the simple and glandular nonstratified squamous epithelium of eccrine and apocrine structures [52]. In a study conducted on EMPD and mammary Paget’s disease patients, FFPES

Extramammary Paget’s disease (EMPD) is a rare cutaneous adenocarcinoma with potential for metastases. It is associated with the apocrine glands and is found primarily in the anogenital or axillary area [48]. Paget’s cells, usually restricted to the epidermis, are large, vacuolated cells with pale staining cytoplasm and a central or laterally displaced large reticulated nucleus [48, 49]. Radical surgeries such as local excision, wide local excision, vulvectomy, or abdominoperineal resection are associated with increased morbidity. Despite these sometimes disfiguring surgeries, recurrence rates following these procedures may still be as high as 31–61% [50, 51]. In some cases, EMPD can be difficult to differentiate from a squamous cell – in situ or melanoma histologically [52].

In 1979, Frederic Mohs reported the use of his fresh tissue technique for the treatment of EMPD. He studied five cases, four of which utilized the fresh tissue technique and one utilized the fixed tissue technique. Three of the fresh tissue cases were recurrent and one was newly diagnosed. All showed no signs of recurrences with follow-ups ranging from 4 months to 4 years; consequently, he concluded that the ability to

antigen (CEA), Ber-EP4, and CAM 5.2 (CK 8 and 18). Anti-CK 7 was found to be the most sensitive stain and produced the strongest staining in each case with the lowest amount of background staining [52].

CK7 was also studied on FS in 4/12 patients, while CEA was used in two of these four patients treated with MMS [48]. The investigators reported that CK7 is the intraoperative immunostain of choice, but cautioned that distinguishing between an ordinary eccrine coil and dermal EMPD can be difficult. It may be prudent to send tissue from the debulking layer during MMS for FFPES preparation and CK immunostaining for confirmation. It should be noted, however, that one of the four patients enrolled in the study went on to have a local recurrence in 1 year [48].

Another marker found in eccrine and apocrine glands is CEA, which also has been shown to be sensitive and specific in differentiating Paget’s cell from keratinocytes or melanocytes. It may be difficult to distinguish Paget’s disease from Bowen’s disease. CEA immunostaining was utilized in a patient with EMPD with adjacent keratinocytic intraepidermal dysplasia. Negative staining in keratinocytes and melanocytes and variable cytoplasmic staining intensity in the