Abstract
Background/Aim: Mohs micrographic surgery (MMS) is a specialized procedure for removing skin tumors. Intraoperative assessment of the resection margin (RM) status using frozen section examination is a crucial component of MMS. This study aimed to identify significant clinicopathological characteristics that could help surgeons determine the optimal surgical extent. Patients and Methods: One hundred and fifty-one patients with primary skin tumors were included. The relationship between RM involvement and the clinico-pathological characteristics was analyzed for each histological type. Results: Basal cell carcinoma (BCC) was significantly more likely to exhibit positive RMs and required additional excision during MMS compared to squamous cell carcinoma. In addition, the probability of RM involvement was significantly higher in high-risk BCC subtypes. Conclusion: When planning MMS, considering the histological type and presence of high-risk morphology may help surgeons perform more effective procedures.
- Skin
- epidermal carcinoma
- Mohs micrographic surgery
- resection margin involvement
- frozen section examination
Mohs micrographic surgery (MMS) is a specialized procedure for skin tumors. MMS was first developed in the 1930s and has been increasingly used to remove various skin tumors (1, 2). Mohs proposed removing the horizontal layer of skin tissue to evaluate the status of the entire deep and peripheral resection margins (RMs) under a microscope (3). This process can be repeated using a layer-by-layer technique until the RM is completely negative (4, 5). Therefore, MMS has the advantage of being able to evaluate all RMs and minimize the loss of normal tissue. MMS enables comprehensive analysis of the deep and peripheral RMs of the tumor on horizontal frozen section examination (FSE), ensuring complete tumor removal without the need for additional wider excision once clearance has been demonstrated (6-8). Most tumors treated with MMS are removed in one or two stages and repaired with primary closure, which is associated with fewer complications and less postoperative pain than other repairs (9). Patients generally have low peak pain levels following MMS and reconstruction and are highly satisfied with the pain control (7, 10, 11). However, despite these important benefits, including a higher success rate and lower amount of normal tissue loss compared to other procedures, MMS has the disadvantage of requiring a long operation time because it proceeds while the RM status is checked using FSE. In particular, FSE is a crucial component of MMS because intraoperative FSE interpretation determines the need for additional excision or operation termination (5, 12).
Many studies have compared the prognosis and clinical outcomes of patients who underwent extensive surgical resection with those of MMS-treated patients. Kimyai-Asadi et al. (13) compared the safety of MMS and related surgical repairs in 3,937 consecutive patients who underwent MMS and found that these procedures were equally safe. Ran et al. (14) compared the frequencies of RM involvement, local recurrence, and repeat surgeries between MMS and wide local excision (WLE). They showed that patients treated with WLE had higher rates of positive RMs and local recurrence, resulting in repeat surgeries. Regarding the histological type of skin tumor, previous studies on MMS have mainly focused on cutaneous malignant melanoma (MM) (15-21). In contrast, relatively few studies have described the characteristics of MMS-treated patients with non-melanoma skin tumors, including basal cell carcinoma (BCC), squamous cell carcinoma (SCC), and dermatofibrosarcoma protuberans (DFSP) (22). For example, a previous study by Tchanque-Fossuo et al. (23) included 28 Native American patients treated with MMS for BCC or SCC. Furthermore, limited research has been published on the association between the effectiveness of MMS and the pathological features of non-melanoma skin tumors.
Our institution launched MMS in 2017 and has successfully performed this procedure in more than 150 patients. Most of the patients were diagnosed with SCC or BCC. Approximately 75% of all non-melanoma skin tumors are BCC, whereas cutaneous SCC represents approximately 20%; the remainder are MMs (4%) and other rare tumors (6). In this study, we examined the association between RM status and various clinicopathological characteristics in patients who underwent MMS for primary skin tumors. Through these analyses, we aimed to identify significant characteristics that can predict a positive RM during MMS and help surgeons determine more appropriate and effective surgical procedures.
Patients and Methods
Case selection and clinicopathological data collection. This study was reviewed and approved by the Institutional Review Board of the Kangbuk Samsung Hospital (protocol number: 2022-09-036; approval date: October 6, 2022). We enrolled 151 patients diagnosed with several histological types of primary skin tumors, including BCC, SCC, MM, DFSP, and extramammary Paget disease (EMPD). All the patients underwent MMS at our institution between April 2017 and August 2021. Two board-certified pathologists specialized in dermatopathology (J.C. and Y.C.) examined all available hematoxylin and eosin-stained slides using light microscopy to confirm the histological types and final pathological diagnoses of the 151 skin tumor cases. The following clinical information and pathological characteristics were obtained from electronic medical records and pathology reports: age at the time of initial diagnosis; sex; location and size of the tumor; and RM status at the first, second, third, and fourth excisions. The tumor locations were classified as low-risk (trunk and extremities, excluding the hands, feet, nail units, pretibial regions, and ankles), medium-risk (cheeks, forehead, scalp, neck, and pretibial regions), and high-risk (central face, eyelids, eyebrows, periorbital regions, nose, lips, chin, mandible, preauricular regions, postauricular regions, temple, ear, genitalia, hands, and feet) (7). Figure 1 depicts representative photomicrographs showing the typical histological features of BCC (Figure 1A-D), SCC (Figure 1E), MM (Figure 1F and G), DFSP (Figure 1H), and EMPD (Figure 1I).
Typical histological features of cutaneous basal cell carcinoma (BCC; A-D), squamous cell carcinoma (SCC; E), malignant melanoma (MM; F and G), dermatofibrosarcoma protuberans (DFSP; H), and extramammary Paget disease (EMPD; I). (A) BCC is characterized by the proliferation of basaloid tumor cells forming nests and sheets, retraction of tumor cell nests from the stroma, and occasional keratinization. (B) Infiltrating BCC consists of jagged nests and cords of basaloid tumor cells associated with obvious fibromyxoid stroma. The tumor has an irregular permeating pattern of stromal invasion. (C) Micronodular BCC consists predominantly of rounded, ovoid, or slender groups of basaloid cells scattered throughout the dermis. (D) In sarcomatoid BCC, nodules of basaloid tumor cells are separated by malignant mesenchymal stroma resembling undifferentiated pleomorphic sarcoma. (E) SCC displays variable-sized sheets of neoplastic keratinocytes with frequent keratin pearl formation. (F) MM is characterized by a confluent proliferation of neoplastic melanocytes involving the papillary dermis. (G) The neoplastic melanocytes possess enlarged, pleomorphic, hyperchromatic nuclei with inconspicuous nucleoli and mitotic figures (yellow arrows). (H) DFSP displays the proliferation of spindle cells, infiltrating the subcutaneous tissue in a honeycomb-like pattern. (I) EMPD shows large, atypical cells spreading just above the basal layer of the epidermis. They possess abundant pale, vacuolated cytoplasm and are singly dispersed or arranged in clusters. Staining method: A-I, hematoxylin and eosin staining. Original magnification: A, 40×; B, 12.5×; C, 12.5×; D, 100×; E, 40×; F, 40×; G, 200×; H, 12.5×; I, 100×.
Surgical procedure of MMS. Briefly, the surgeon marked the RMs of the clinically evident tumors (usually 1-2 mm). The area to be excised was cleansed with an antiseptic, and local anesthetics (buffered lidocaine with epinephrine) were administered. The tumor was removed (debulked) using a scalpel blade or curette, with caution not to erase the markings. The first-stage RMs were removed with the blade at a 45-degree angle to the skin (Figure 2A). Beveling the blade allows proper alignment of the peripheral RMs of the tissue during preparation for pathological examination. During the first-stage excision, the two lateral sides of each resection margin were marked with a different color for further orientation, allowing for identifying the residual tumor (if the residual tumor is detected during FSE). Once the specimen was completely removed, hemostasis was achieved. An additional long-lasting anesthetic was occasionally administered to prevent local pain while the patient awaited laboratory processing of the specimen and interpretation of the pathology slides. While the tissue was processed (20-60 min), the patient’s wound was bandaged in the operating room.
Intraoperative resection margin evaluation with frozen section examination. (A) First-stage resection margins are removed and marked with blue or yellow color. (B) An illustrated pathological record of Mohs microscopic surgery. Two microscopic areas of resection margin involvement are marked on the image with red color. (C and D) Hematoxylin and eosin staining of (C) frozen and (D) the matched permanent sections reveals the resection margins involving basal cell carcinoma, indicated by yellow and green arrows, respectively. Staining method: C and D, hematoxylin and eosin staining. Original magnification: C, 4×; D, 4×.
Intraoperative RM evaluation using FSE. The resected specimen was flattened to ensure the beveled peripheral RMs were on the same plane as the deep RM. The markings were in different colors. To correlate with the surgical defect, a two-dimensional map of the lesion and the corresponding area was drawn using the same colors to identify each marking. An illustrated pathological record of MMS is shown in Figure 2B. The specimen was then frozen using an Optimum Cutting Temperature Compound (OCT; Tissue-Tek O.C.T. Compound, Sakura Finetek, Torrance, CA, USA) and fixed to a chuck. The OCT-embedded tissue block was sectioned using a cryotome, and several horizontal histological sections were placed on glass slides to allow complete analysis of all deep and peripheral RMs. The sections were stained with hematoxylin and eosin. Slides were examined under a light microscope (BX51; Olympus, Tokyo, Japan). The procedure was repeated if a residual tumor was detected in any of the RMs (Figure 2C and D). A new 1-2 mm RM was again taken only around the compromised area and mapped positively during the histological examination. If only the deep RM was involved, the second stage was occasionally performed without enlarging the diameter of the defect. Pathological specimens were carefully handled to maintain their orientation. In addition, the full tissue around the mapped tumor-positive area was taken to limit the chance of sampling errors in the second and subsequent stages. Once all RMs were tumor-free, surgeons safely reconstructed the wound.
Statistical analysis. Using R version 4.1.2 (R Foundation for Statistical Computing, Vienna, Austria), the independent two-sample t-test, Pearson chi-square test, Fisher’s exact tests, and linear-by-linear association tests were performed to analyze the clinicopathological differences between BCC and SCC and the relationship between RM status and the clinicopathological characteristics of patients with BCC or SCC. We also analyzed the association between high-risk histological subtypes of BCC or histological differentiation of SCC with RM involvement and clinicopathological characteristics. Statistical significance was set at p<0.05.
Results
Baseline clinicopathological characteristics. Table I summarizes the clinicopathological characteristics of 151 patients with skin tumors. The majority (132/151; 87.4%) of the patients were diagnosed with either BCC (n=76) or SCC (n=56). The median age of the patients at initial diagnosis was 75 years. Patients with DFSP (median age, 28 years) were much younger than the others. The median tumor sizes were larger in DFSP (4.2 cm) and EMPD (4.5 cm) than in BCC (0.7 cm) and SCC (1.1 cm). The first RM involved the tumor in more than half of the cases (84/151; 55.6%). In particular, approximately two-thirds of the BCC cases (48/76; 63.2%) involved the first RM, and none of the EMPD patients (0/7; 0.0%) had a negative RM at the first excision. All except one patient with EMPD achieved tumor-free RMs at the fourth (BCC, SCC, and EMPD), third (MM), and second (DFSP) excisions.
Patient characteristics.
Clinicopathological differences between BCC and SCC. Table II summarizes the differences in clinicopathological characteristics between BCC and SCC. SCC patients (median age, 81 years) were significantly older than BCC patients (median age, 75 years; p=0.004). The distribution of tumor locations was similar. The median tumor size of SCC (1.1 cm) was significantly larger than that of BCC (0.7 cm; p<0.001). Despite the smaller size, BCC had a higher rate of RM involvement in the first and second excisions (63.2% and 41.7%, respectively) than SCC (39.3% and 18.2%; p=0.011 and p=0.063, respectively). Both groups achieved negative RMs after the fourth excision.
Clinicopathological differences between basal cell carcinoma (BCC) and squamous cell carcinoma (SCC).
Association of RM involvement with the clinicopathological characteristics of BCC. Table III summarizes the relationship between a positive first RM and the clinicopathological characteristics of BCC. Forty-seven tumors (63.5%) had first RM involvement. Larger tumor size and high-risk subtype were significantly associated with first RM involvement (p=0.003 and p=0.001, respectively). Most BCC cases with high-risk histology exhibited a positive first RM (28/33, 84.8%). Table IV summarizes the differences in the clinicopathological characteristics of BCC according to the presence or absence of high-risk histology. In addition to the first RM, high-risk subtypes were significantly associated with a second positive RM (p=0.018).
Association of first resection margin (RM) involvement with clinicopathological characteristics of basal cell carcinoma.
Association of the high-risk histological subtype with resection margin (RM) involvement and clinicopathological characteristics of basal cell carcinoma.
Association of RM involvement with the clinicopathological characteristics of SCC. Table V summarizes the relationship between a positive first RM and the clinicopathological characteristics of SCC. The frequency of first RM involvement in moderately to poorly differentiated SCCs (7/11, 63.6%) was higher than that in well-differentiated tumors (15/45, 33.3%), but this difference was not significant (p=0.134). Table VI summarizes the differences in the clinicopathological characteristics of SCC according to the degree of histological differentiation. Patients with well-differentiated SCC (81 years) were significantly older than those with moderately to poorly differentiated tumors (75 years; p=0.004). The tumor size of moderately to poorly differentiated SCC (mean, 2.9 cm) was larger than that of well-differentiated SCC (mean, 1.1 cm, p<0.001). The differences in all RM statuses were not significant between the two groups, which were divided based on histological differentiation.
Association of first resection margin (RM) involvement with the clinicopathological characteristics of squamous cell carcinoma.
Associations of histological differentiation with resection margin (RM) involvement and the clinicopathological characteristics of squamous cell carcinoma.
Discussion
This study examined 151 patients with primary skin tumors (76 BCCs, 56 SCCs, 9 MMs, 3 DFSPs, and 7 EMPDs) who underwent MMS. When comparing the clinicopathological characteristics of BCC and SCC, BCC was significantly more likely to involve the first RM and require additional excision than SCC. We divided the BCC cases into two groups according to the presence of high-risk morphology and investigated whether there were significant differences in clinicopathological characteristics between the two groups. We found that tumor involvement of the first RM was significantly more frequent in high-risk subtypes of BCC, including infiltrative, micronodular, and sarcomatoid BCCs. During the first excision, FSE revealed RM involvement in most patients with high-risk BCC subtypes (84.8%). In addition to the first RM, high-risk BCC morphology was significantly associated with second RM involvement. These findings indicated a significantly higher probability of residual tumors on the RMs of MMS specimens in high-risk BCC subtypes than in nodular or superficial BCCs. In the biopsy diagnosis of BCC, documentation of the presence of a high-risk morphology may help surgeons determine the extent of surgical resection and more effective surgical procedures. In contrast, we did not find any significant relationship between RM involvement and clinicopathological characteristics in patients with SCC. Regarding the degree of histological differentiation of SCC, the frequency of a positive first RM was higher in moderately to poorly differentiated SCC (63.6%) than in well-differentiated SCC (33.3%); however, the difference was not significant.
Our observation that additional RM excision is required in high-risk BCC subtypes aligns with previous data showing that micronodular, infiltrative, and sclerosing BCCs have higher recurrence rates and more frequent positive RMs than the nodular and superficial subtypes (24-28). Hendrix et al. (29) also found that micronodular BCC had more covert, clinically undetected tumor extensions than nodular BCC, making it more difficult to completely remove the tumor. They also reported that the number of stages for complete excision, the width of tissue required to remove the subclinical extension of the tumor, and the depth of the defect at the completion of MMS were all greater in micronodular BCC than in the nodular subtype. Similarly, Batra et al. (30) reported that three or more additional RM excisions were required for high-risk BCC subtypes and that the identification of these lesions likely to exhibit extensive subclinical spread can help guide management to ensure complete tumor eradication, thereby reducing the recurrence risk and its associated morbidity and cost.
We noted that all seven EMPDs involved the first RM, and four of the seven patients with EMPD did not achieve negative RMs at the second and third excisions. EMPD is a rare cutaneous malignancy with poorly described treatment strategies (31, 32). Due to its rarity, clinical trials comparing various treatment methods for EMPD are lacking. It is difficult to determine the surgical extent of EMPD, and obtaining a negative RM is challenging because of its multifocal distribution. Simond et al. (33), in their literature review of EMPD, stated that MMS appeared superior to WLE in terms of tissue preservation and disease recurrence, but the effectiveness of MMS for EMPD remains controversial. Hartman et al. (34) described in their case report that the patient underwent MMS for EMPD. However, the RMs remained positive despite several additional excisions, and the management of EMPD was complicated by the multifocal, non-contiguous nature of the tumor and the presence of clinically occult extensions. Kato et al. (35) compared preoperative mapping biopsy with MMS to assess the RM status. They reported that mapping biopsy can be useful for treating EMPD, as it is equivalent to MMS in terms of the recurrence rate. Since the boundary of EMPD is unclear to the naked eye, having RMs slightly larger than visible to the naked eye helps reduce additional resection.
This study had some limitations. First, all the study participants were treated at a single institution, which makes it difficult to generalize the findings to a larger population. Second, we could not analyze the statistical significance of clinicopathological differences in patients with MM, DFSP, or EMPD because of the relatively small number of cases. Third, since we determined the histological subtype of each BCC case as the most dominant single one, we could not analyze the clinicopathological difference according to the amount of high-risk morphology in cases showing mixed histology. Additionally, we acknowledge the possibility of artifacts that can occur if the laboratory processing steps of FSE are not adequately followed, including poor sectioning, freezing artifacts, and interpretation or technical errors that can occur when surgeons do not excise sufficient non-neoplastic tissue to ensure clearance of all tumor tissues and non-contiguous tumors.
In conclusion, we suggest that MMS for BCC should be performed with more RM than visible to the naked eye to prevent multiple additional surgeries and shorten the operation time. In addition, if BCC is found in an initial biopsy, identifying whether there is a high-risk subtype may help determine the surgical scope. Based on these results, further investigations are warranted to confirm or disprove the efficacy of MMS in various types of skin tumors. Identifying a subset of lesions showing extensive subclinical spread can help guide management to ensure complete tumor eradication, thereby reducing the risk of recurrence and associated morbidity and cost.
Acknowledgements
This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean Government (MSIT) (2023R1A2C2006223).
Footnotes
Authors’ Contributions
All Authors made substantial contributions to the conceptualization and design of this study; acquisition, analysis, interpretation, and validation of the data; preparation of the draft; critical editing and revision of the manuscript; and final approval of the version to be published.
Conflicts of Interest
None of the Authors have any conflicts of interest or financial ties to declare regarding this study.
- Received March 23, 2023.
- Revision received April 3, 2023.
- Accepted April 4, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).