Key Points
Question
Do outcomes of primary high-stage cutaneous squamous cell carcinoma (cSCC) differ after treatment with Mohs surgery compared with wide local excision (WLE)?
Findings
In this cohort study including 216 patients with high-stage cSCC, Mohs surgery was associated with improved outcomes in the treatment of primary high-stage cSCC compared with WLE, which showed increases in the 3-year cumulative incidences of local recurrence, nodal metastasis, distant metastasis, any recurrence, and disease-specific death.
Meaning
These findings suggest that patients with primary high-stage cSCC should be offered first-line treatment with Mohs surgery or alternative methods of peripheral and deep en face margin assessment whenever possible.
This cohort study assesses the outcomes of primary high-stage cutaneous squamous cell carcinoma among patients treated with Mohs surgery compared with those treated with wide local excision.
Abstract
Importance
High-stage cutaneous squamous cell carcinoma (cSCC) has an increased risk of recurrence, metastasis, and mortality. Studies investigating the outcomes of high-stage cSCC among patients treated with Mohs surgery compared with those treated with wide local excision (WLE) are limited.
Objective
To assess the outcomes of primary high-stage cSCC among patients treated with Mohs surgery compared with those treated with WLE.
Design, Setting, and Participants
This retrospective cohort study using propensity score weighting was conducted in a tertiary academic medical center in Boston, Massachusetts. Patients were included if they had primary high-stage cSCC and had been treated with either Mohs surgery or WLE between January 1, 2000, and December 31, 2019. Data analysis was performed between November 3 and 6, 2023.
Exposures
Primary surgical treatment with Mohs surgery or WLE.
Main Outcomes and Measures
Outcomes included local recurrence, nodal metastasis, distant metastasis, any recurrence (ie, a composite outcome of recurrence or metastasis), and disease-specific death. Propensity scores were estimated via logistic regression using baseline patient and tumor characteristics. Competing risk regression analysis was used to compute crude and inverse probability of treatment weighting (IPTW), cause-specific hazard ratios (HRs), and Fine-Gray subdistribution HRs and to derive cumulative incidence functions stratified by Mohs and WLE.
Results
This study included 216 patients with high-stage cSCC who had a mean (SD) age of 73.5 (13.3) years; 151 (69.9%) were men and 65 (30.1%) were women. The median follow-up time was 33.1 months (IQR, 11.3-77.6 months). After IPTW, the baseline characteristics were well balanced between the WLE and Mohs surgery treatment groups, with absolute standardized differences of less than 0.10 across all characteristics. In the IPTW competing risks model, the 3-year cumulative incidence of all adverse outcomes were greater among patients in the WLE group compared with those in the Mohs surgery group, including local recurrence (19.8% vs 9.6%; weighted cause-specific HR, 2.33 [95% CI, 1.39-3.92]; P = .001), nodal metastasis (17.9% vs 11.0%; weighted cause-specific HR, 1.80 [95% CI, 1.07-3.02]; P = .03), distant metastasis (8.4% vs 4.4%; weighted cause-specific HR, 2.10 [95% CI, 0.97-4.57]; P = .06), any recurrence (32.0% vs 15.8%; weighted cause-specific HR, 2.38 [95% CI, 1.57-3.61]; P < .001), and disease-specific death (17.5% vs 7.1%; weighted cause-specific HR, 2.74 [95% CI, 1.54-4.88]; P = .001).
Conclusions and Relevance
The findings of this cohort study suggest that Mohs surgery was associated with improved outcomes in the treatment of primary high-stage cSCC compared with WLE. These findings further suggest that Mohs surgery or alternative methods of peripheral and deep en face margin assessment should be offered as first-line treatment.
Introduction
Although cutaneous squamous cell carcinoma (cSCC) is generally considered a diagnosis with favorable outcomes, high-stage cSCC has been shown to have an increased risk of recurrence, metastasis, and mortality.1 The Brigham and Women’s Hospital (BWH) tumor classification system for cSCC uses 4 risk factors to confer stage: preoperative tumor diameter greater than or equal to 2 cm, poorly differentiated tumor cells, large-caliber (≥0.1 mm) nerve invasion, and tumor invasion beyond subcutaneous fat.1 Tumors with 2 or 3 of these risk factors are classified as T2b, and tumors with all 4 risk factors or bone invasion are classified as T3. Collectively, tumors classified as BWH T2b or T3 are considered high-stage cSCC.
Mohs surgery and wide local excision (WLE) are the 2 primary surgical approaches used in treating cSCC. Both can be used with curative intent to achieve clear margins, but they differ in the method by which the excised tissue is processed and analyzed. Mohs surgery uses the fresh-frozen technique with horizontal sections of the base and en face sections of the perimeter, which allows for 100% margin assessment. Sections are examined by the resecting surgeon who maps the location of residual tumor to the wound bed and re-excises these regions until clear margins are achieved, typically on a single day.2,3 In regions where Mohs surgery is unavailable owing to a lack of necessary resources, technology, and/or personnel, an alternative method involving equivalent margin visualization and mapping through paraffin sections, which are analyzed by a pathologist (such as in the Tübingen technique), is often utilized. In contrast, WLE uses the bread-loafing technique with vertical sections, which results in histologic visualization of less than 1% to 2% of the marginal surface.2,3,4
Current National Comprehensive Cancer Network (NCCN) guidelines recommend Mohs surgery or the Tübingen technique (collectively termed peripheral and deep en face margin assessment [PDEMA]) as the preferred treatment for cSCC with 1 or more of the following risk factors: 4 cm or greater diameter, poor differentiation, invasion past fat, desmoplasia, lymphatic invasion, vascular invasion, and invasion of nerves deep to dermis or 0.1 mm or greater in caliber.5 The NCCN definitions of Mohs surgery and alternative non-Mohs PDEMA techniques have been recently described to facilitate guideline compliance.6
The existing literature has demonstrated the benefits of Mohs surgery compared with WLE in the treatment of cSCC. Large cohort studies, systematic reviews, and meta-analyses involving cSCC of all stages have identified recurrence rates of 3.0% to 3.1% following Mohs surgery and 5.3% to 8.0% following WLE, reflecting the predominance of low-stage cSCC and its overall good outcomes.7,8,9 However, studies evaluating outcomes for high-stage cSCC have been limited to single-arm cohorts. Cohort studies that evaluated outcomes after Mohs surgery alone for high-stage cSCC found rates between 3.8% and 15.9% for local recurrence, 5.5% and 13.6% for metastatic disease, and 1.9% and 4.5% for disease-specific death.10,11 A cohort study from the head and neck literature on advanced cSCC treated with WLE alone showed rates of 23.1% for local recurrence, 8.5% for nodal metastasis, and 1.4% for distant metastasis.12
The ability to directly compare treatment modalities across these studies is constrained by considerable heterogeneity in their study designs and cohort characteristics. This includes differences in patient demographics, distribution of tumor risk factors, inclusion of recurrent or metastatic tumors, and duration of follow-up. Furthermore, an inherent challenge in comparing Mohs surgery and WLE is that certain patient or tumor characteristics can affect both the treatment choice and the outcomes observed (ie, confounding by indication), making it difficult to determine what outcomes are solely due to the treatment.
Because no randomized clinical trials have been conducted for cSCC treated with Mohs surgery or WLE, we undertook this retrospective cohort study at a single tertiary academic medical center over 20 years and used propensity score methods to balance the treatment groups. Given prior data, we hypothesized that the technique of complete margin assessment of Mohs surgery would be associated with improved outcomes for high-stage cSCC. We chose to focus this study on high-stage cSCC because the baseline risk of recurrence, metastasis, and death is substantial enough to show differences between treatment modalities that are clinically meaningful.
Methods
This retrospective cohort study was approved by the Massachusetts General Brigham Human Research Committee and Institutional Review Board. Informed consent was waived due to minimal risk to participants. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Cohort
Patients with cSCC tumors diagnosed at BWH (Boston, Massachusetts) between January 1, 2000, and December 31, 2019, were identified via an electronic institutional database. Medical records were abstracted for patient characteristics, tumor characteristics, treatment history, and outcomes of interest, including local recurrence, nodal metastasis, distant metastasis, disease-specific death, and a composite outcome of any recurrence (local, regional, or distant). The follow-up period for all outcomes of interest concluded on the date of the event, date of last health care visit, or the study’s end date (October 31, 2023). Self-reported race and ethnicity data were collected along with age and sex so that demographic characteristics were balanced after propensity score weighting.
Tumors were categorized according to the BWH tumor classification system for cSCC as previously described. Missing data for tumor characteristics were addressed using standard dermatopathology reporting conventions: (1) tumors were assumed to be well differentiated if differentiation was not mentioned on the pathology or operative report; (2) tumors were assumed to be invasive to the dermis if there was no mention of depth of invasion on the pathology or operative report; and (3) tumors were assumed to be absent of perineural invasion or lymphovascular invasion if not mentioned on the pathology or operative report. All eligible tumors were reviewed in the electronic medical record by 1 investigator (D.M.W.) to confirm risk factors, other relevant tumor characteristics, and outcomes.
Inclusion criteria were primary high-stage cSCC tumors (BWH T2b or T3) treated with either Mohs surgery or WLE. Exclusion criteria were noncutaneous, anal, in situ, metastatic at presentation, and recurrent tumors; tumors treated initially with a combination of Mohs surgery and WLE; tumors that received primary treatment with palliative surgery or nonsurgical treatment; and tumors treated with radical surgery (eg, amputation, bone resection, exenteration, orbitotomy, parotidectomy, lymphadenectomy, auriculectomy, rhinectomy, or penectomy) since these procedures are ineligible for Mohs surgery due to lack of general anesthesia services in the Mohs surgery center and non-Mohs PDEMA has not been routinely performed at our institution.
Statistical Analysis
A propensity score for identifying whether a tumor was likely to receive treatment with Mohs surgery or WLE was estimated using a logistic regression model that included patient characteristics, including age, sex, race and ethnicity, immune status (ie, immunocompromised due to organ transplant, autoimmune disease, or immunosuppressive medications), and tumor characteristics, including location, preoperative diameter, depth of invasion, differentiation, and presence of large-caliber perineural invasion or lymphovascular invasion. The inverse probability of treatment weighting (IPTW) method was used to balance baseline patient and tumor characteristics between the treatment groups, thereby controlling for confounding by indication. The balance in baseline covariables and outcomes was assessed using standardized differences (calculated as the difference in means or proportions divided by the standard error) before and after IPTW, with values greater than 0.10 suggesting imbalance between the treatment groups.
Competing risk regression analysis was used to compute crude and IPTW cause-specific hazard ratios (HRs) and Fine-Gray subdistribution HRs and to derive cumulative incidence functions for each treatment group and outcome. Cause-specific hazard models were chosen for analysis of treatment effects in the presence of competing risks, whereas Fine-Gray subdistribution hazard models were chosen to estimate cumulative incidence in the presence of competing risks. Statistical analyses were performed using Stata, version 18.0 (StataCorp LLC). Statistical significance was set at P < .05 (calculated via 2-tailed Wald χ2 test). Data analysis was performed between November 3 and November 6, 2023.
Results
A total of 6465 cases of cSCC treated initially with Mohs surgery or WLE at BWH between January 1, 2000, and December 31, 2019, were identified. Of these, 332 cases were classified as high-stage cSCC (BWH T2b or T3). Exclusions were made for 37 cases due to metastasis at presentation and 7 cases due to recurrence of a prior tumor. Additionally, 16 patients were excluded because they underwent a preplanned combination of Mohs surgery and WLE for initial treatment, 7 were excluded because they had received palliative treatment, and 33 were excluded because they had undergone radical surgery. Another 16 cases were excluded due to insufficient information about the primary tumor or treatment. Ultimately, 216 high-stage cSCC cases were included in the final analysis and were followed up for a median time of 33.1 months (IQR, 11.3-77.6 months).
The patients had a mean (SD) age of 73.5 (13.3) years. There were 151 men (69.9%) and 65 women (30.1%). In terms of race and ethnicity, 206 patients (95.4%) identified as White, 2 (0.9%) identified as other race or ethnicity, and 8 did not report their race or ethnicity (3.7%). A total of 19 patients required re-excision due to positive margins after the initial surgical treatment, including 16 who initially underwent WLE (3 of whom then underwent Mohs surgery) and 3 who underwent Mohs surgery (1 of whom then underwent WLE).
Baseline patient and tumor characteristics before and after IPTW are presented in Table 1. In the original (pre-IPTW) cohort, treatment with Mohs surgery occurred more often than WLE among male patients (97 [74.0%] vs 54 [63.5%]; standardized difference, 0.23) and immunosuppressed patients (45 [34.4%] vs 25 [29.4%]; standardized difference, 0.11). Additionally, tumors treated with Mohs surgery vs WLE were more often located on the head or neck, ear, and lip (82 [62.6%] vs 41 [48.8%], 16 [12.2%] vs 6 [7.1%], and 8 [6.1%] vs 4 [4.8%], respectively; standardized difference, 0.51); of smaller mean preoperative diameter (25.1 mm vs 32.8 mm; standardized difference, 0.40); invasive beyond subcutaneous fat (74 [56.5%] vs 36 [42.4%]; standardized difference, 0.29); well to moderately differentiated (54 [41.2%] vs 19 [22.4%]; standardized difference, 0.41); and positive for large-caliber perineural invasion (39 [29.8%] vs 13 [15.3%]; standardized difference, 0.35). After IPTW, the distribution of baseline characteristics was well balanced between the WLE and Mohs surgery treatment groups, with absolute standardized differences of less than 0.10 across all characteristics.
Table 1. Patient Characteristics Before and After Propensity Score Weightinga.
| Characteristic | Patient group, No. (%) | |||||
|---|---|---|---|---|---|---|
| Original cohort | Resampled cohort with IPTWb | |||||
| WLE (n = 85) | Mohs surgery (n = 131) | Standardized difference | WLE | Mohs surgery | Standardized difference | |
| Follow-up, mean (SD), mo | 49.7 (48.3) | 49.5 (46.1) | 0.005 | 46.3 (44.8) | 48.9 (46.4) | 0.06 |
| Age, mean (SD), y | 73.0 (12.4) | 73.8 (13.8) | −0.07 | 73.6 (11.5) | 73.7 (14.3) | −0.01 |
| Race and ethnicity | ||||||
| White | 81 (95.3) | 125 (95.4) | 0.25 | 197 (97.0) | 203 (96.2) | 0.05 |
| Otherc | 2 (2.4) | 0 | 0 | 0 | ||
| Unknown | 2 (2.4) | 6 (4.6) | 6 (3.0) | 8 (3.8) | ||
| Sex | ||||||
| Female | 31 (36.5) | 34 (26.0) | 0.23 | 64 (31.5) | 65 (30.8) | 0.02 |
| Male | 54 (63.5) | 97 (74.0) | 139 (68.5) | 146 (69.2) | ||
| Immune status | ||||||
| Not immunosuppressed | 60 (70.6) | 86 (65.6) | 0.11 | 141 (69.5) | 148 (70.1) | 0.01 |
| Immunosuppressed | 25 (29.4) | 45 (34.4) | 62 (30.5) | 63 (29.9) | ||
| Tumor location | ||||||
| Head or neck | 41 (48.8) | 82 (62.6) | 0.51 | 116 (56.9) | 121 (57.4) | 0.05 |
| Ear | 6 (7.1) | 16 (12.2) | 20 (9.8) | 21 (10.0) | ||
| Lip | 4 (4.8) | 8 (6.1) | 9 (4.4) | 11 (5.2) | ||
| Arm or hand | 9 (10.7) | 4 (3.1) | 14 (6.9) | 14 (6.6) | ||
| Trunk | 11 (13.1) | 13 (9.9) | 23 (11.3) | 22 (10.4) | ||
| Leg or foot | 12 (14.3) | 7 (5.3) | 20 (9.8) | 20 (9.5) | ||
| Genitalia | 1 (1.2) | 1 (0.8) | 2 (1.0) | 2 (1.0) | ||
| Preoperative tumor diameter, mean (SD), mm | 32.8 (24.1) | 25.1 (11.9) | 0.40 | 27.2 (16.1) | 26.3 (12.9) | 0.06 |
| Tumor depth | ||||||
| Dermis or subcutaneous fat | 49 (57.6) | 57 (43.5) | 0.29 | 104 (51.2) | 108 (51.2) | 0.001 |
| Beyond subcutaneous fat | 36 (42.4) | 74 (56.5) | 99 (48.8) | 103 (48.8) | ||
| Tumor differentiation | ||||||
| Well to moderate | 19 (22.4) | 54 (41.2) | 0.41 | 62 (30.5) | 69 (32.7) | 0.05 |
| Poor | 66 (77.6) | 77 (58.8) | 141 (69.5) | 142 (67.3) | ||
| Tumor invasiveness | ||||||
| Large-caliber perineural invasion | ||||||
| No | 72 (84.7) | 92 (70.2) | 0.35 | 166 (81.4) | 164 (77.7) | 0.09 |
| Yes | 13 (15.3) | 39 (29.8) | 38 (18.6) | 47 (22.3) | ||
| Lymphovascular invasion | ||||||
| No | 80 (94.1) | 126 (96.2) | 0.10 | 194 (95.6) | 202 (95.7) | 0.01 |
| Yes | 5 (5.9) | 5 (3.8) | 9 (4.4) | 9 (4.3) | ||
Abbreviations: IPTW, inverse probability of treatment weighting; WLE, wide local excision.
Numbers may not sum to group totals and percentages may not sum to 100% owing to rounding and missing data.
Because resampling involved fractional contributions from individual patients, rounded values approximate the effective contribution of patients in the weighted cohort but do not correspond to discrete individuals in the dataset.
Categories are not specified to protect patient privacy owing to the small number of patients.
Table 2 shows outcomes and the cause-specific hazard models. In the crude (unadjusted) cause-specific hazard models, the rate of developing all adverse outcomes was greater for WLE compared with Mohs surgery for local recurrence (crude cause-specific HR, 1.72 [95% CI, 0.83-3.56]; P = .15), nodal metastasis (crude cause-specific HR, 1.59 [95% CI, 0.71-3.53]; P = .26), distant metastasis (crude cause-specific HR, 3.51 [95% CI, 1.08-11.41]; P = .04), any recurrence (crude cause-specific HR, 1.97 [95% CI, 1.10-3.50]; P = .02), and disease-specific death (crude cause-specific HR, 2.47 [95% CI, 1.07-5.71]; P = .03). In the IPTW cause-specific hazard models, the rate of developing all adverse outcomes was greater for WLE compared with Mohs surgery for local recurrence (weighted cause-specific HR, 2.33; 95% CI, 1.39-3.92; P = .001), nodal metastasis (weighted cause-specific HR, 1.80; 95% CI, 1.07-3.02; P = .03), distant metastasis (weighted cause-specific HR, 2.10; 95% CI, 0.97-4.57; P = .06), any recurrence (weighted cause-specific HR, 2.38; 95% CI, 1.57-3.61; P < .001), and disease-specific death (weighted cause-specific HR, 2.74; 95% CI, 1.54-4.88; P = .001).
Table 2. Outcomes and Cause-Specific Hazard Models.
| Outcome | No. of outcomes/No. of patients | Patient group | |||
|---|---|---|---|---|---|
| Original cohort | Resampled cohort with IPTW | ||||
| Crude cause-specific HR (95% CI) | P value | Weighted cause-specific HR (95% CI) | P value | ||
| Local recurrence | |||||
| Mohs surgery | 14/131 | 1 [Reference] | NA | 1 [Reference] | NA |
| WLE | 15/85 | 1.72 (0.83-3.56) | .15 | 2.33 (1.39-3.92) | .001 |
| Nodal metastasis | |||||
| Mohs surgery | 12/131 | 1 [Reference] | NA | 1 [Reference] | NA |
| WLE | 12/85 | 1.59 (0.71-3.53) | .26 | 1.80 (1.07-3.02) | .03 |
| Distant metastasis | |||||
| Mohs surgery | 4/131 | 1 [Reference] | NA | 1 [Reference] | NA |
| WLE | 9/85 | 3.51 (1.08-11.41) | .04 | 2.10 (0.97-4.57) | .06 |
| Any recurrence | |||||
| Mohs surgery | 21/131 | 1 [Reference] | NA | 1 [Reference] | NA |
| WLE | 25/85 | 1.97 (1.10-3.50) | .02 | 2.38 (1.57-3.61) | <.001 |
| Disease-specific death | |||||
| Mohs surgery | 9/131 | 1 [Reference] | NA | 1 [Reference] | NA |
| WLE | 14/85 | 2.47 (1.07-5.71) | .03 | 2.74 (1.54-4.88) | .001 |
Abbreviations: HR, hazard ratio; IPTW, inverse probability of treatment weighting; NA, not applicable; WLE, wide local excision.
The eTable in Supplement 1 shows the Fine-Gray subdistribution hazard models. IPTW cumulative incidence functions for each outcome stratified by treatment group are shown in the Figure. The 3-year cumulative incidence of all adverse outcomes was greater for WLE compared with Mohs surgery: local recurrence (19.8% vs 9.6%), nodal metastasis, (17.9% vs 11.0%), distant metastasis (8.4% vs 4.4%), any recurrence (32.0% vs 15.8%), and disease-specific death (17.5% vs 7.1%).
Figure. Propensity Score–Weighted Cumulative Incidence Functions.
HR indicates hazard ratio.
Discussion
Our results highlight the benefits of Mohs surgery compared with WLE for the treatment of primary high-stage cSCC. In the propensity score–weighted analysis, which fully adjusts for differences between the treatment groups for factors known to impact outcomes, the 3-year cumulative incidence of recurrence, metastasis, and disease-specific death was twice as high for tumors treated with WLE compared with Mohs surgery. In the 3 years following treatment, 32.0% of tumors treated with WLE recurred or metastasized in contrast to 15.8% of those treated with Mohs surgery, and 17.5% of tumors treated with WLE resulted in disease-specific death compared with 7.1% of those treated with Mohs surgery.
These results have important implications for clinical practice and the management of high-stage cSCC. The findings affirm the importance of complete margin assessment via Mohs surgery or alternative methods of PDEMA according to NCCN guidelines. Sampling error due to bread-loafing techniques used in WLE can result in falsely negative margins and incomplete tumor removal, resulting in the higher rates of recurrence, metastasis, and disease-specific death observed. Studies have shown recurrent cSCC to be more biologically aggressive with metastatic rates as high as 30%, and metastatic cSCC portends a grim prognosis, with mortality rates of more than 50% for immunocompromised patients with organ transplant or chronic lymphocytic leukemia who make up a substantial portion of those with high-stage cSCC.13,14 As such, it is crucial to attain surgical cure through the initial treatment whenever possible.
The observed 2-fold increase in recurrence and metastatic rate in high-stage cSCC treated with WLE compared with Mohs surgery translates into heightened demand on health care systems and the necessity for costlier subsequent therapy after failed initial treatment with WLE. According to the NCCN, management of recurrent locally advanced or metastatic cSCC involves multidisciplinary discussion and multimodality treatment, including surgery, systemic therapy, and radiation therapy.5 Existing cost-effectiveness studies have already established the higher effectiveness and lower costs associated with treatment with Mohs surgery compared with WLE in intermediate-risk, low-stage cSCC.15 It can be inferred that these results would be amplified in high-stage cSCC, wherein double the number of cases treated with WLE in comparison to Mohs surgery would necessitate expensive, downstream multimodal therapy.
There is a growing body of literature on the use of nodal imaging and adjuvant therapy following surgery with negative margins. At our institution, imaging surveillance of the draining lymph node basins is routinely pursued in high-stage cSCC, where nodal metastatic risk is thought to exceed 10%, akin to the 10% threshold for sentinel lymph node biopsy in melanoma.16 Interestingly, tumors treated with Mohs surgery in this study had a 3-year nodal metastatic rate of 11.0%, and tumors treated with WLE had a 3-year nodal metastatic rate of 17.9%, both of which are above the 10% threshold for nodal staging and surveillance. It would be expected that surgical excision by either modality may not majorly alter this risk since micrometastases may already be in transit toward nodes or have reached nodes at the time of surgery. Further research using multicenter data are needed to define optimal guidelines for imaging surveillance of high-stage cSCC with consideration of primary treatment modality. Meanwhile, we will continue our practice of radiologic surveillance for BWH T2b or higher cSCC every 6 months for 3 years. Regarding adjuvant therapy, a dual-center retrospective study showed that adjuvant radiation therapy after surgery with negative margins reduced the risk of local and locoregional recurrence by half.17 The study controlled for use of Mohs surgery on multivariable analysis but did not perform stratified analysis by surgical treatment approach. Given our findings that treatment with Mohs surgery halves the rate of local and locoregional recurrence in high-stage cSCC compared with treatment with WLE, future studies are needed to explore the extent to which adjuvant radiation therapy reduces the recurrence and metastatic risk after Mohs surgery and whether these reductions outweigh the risks and morbidity of radiation, which can be more substantial in frail patients.
Limitations
The primary limitations of this study are its single-center data and retrospective design. As a tertiary academic medical center, our institution likely manages a higher proportion of high-risk cSCC cases compared with the general population, which may limit its generalizability. Still, the observed rates of adverse outcomes after Mohs surgery in our cohort, with the exception of disease-specific death, fall within the previously reported ranges for adverse outcomes at another tertiary academic medical center and a high-volume community-based practice.10,11 Given the relative rarity of high-stage cSCC and the need for a 20-year cohort to produce the data reported in the present study, a multicenter randomized clinical trial may be prohibitively expensive. It is likely infeasible given the lack of equipoise and national guidelines moving away from WLE. The use of a propensity score–weighted analysis to balance baseline confounders effectively mitigates the limitations inherent in the present study’s retrospective design. This approach enables us to mimic the outcomes of a randomized clinical trial comparing Mohs surgery and WLE for high-stage cSCC.18 Unassessed differences in the type and frequency of surveillance, including radiologic imaging, between the Mohs surgery and WLE groups may represent a limitation of this study. While it is unlikely that the binary occurrence of recurrence or metastasis was underdetected given the extensive follow-up period, differential surveillance practices could influence the time-to-event outcomes in our analysis.
Conclusions
The findings of this cohort study suggest that Mohs surgery may be superior to WLE for the treatment of high-stage cSCC. The propensity score–weighted analysis demonstrated a halving of 3-year cumulative incidence of recurrence, metastasis, and disease-specific death for tumors treated with Mohs surgery compared with WLE. This has important implications for reducing morbidity, mortality, and downstream health care resource utilization. Patients presenting with primary high-stage cSCC should be offered first-line treatment with Mohs surgery whenever possible and NCCN-compliant PDEMA if Mohs surgery is not possible or is unavailable. Future research should continue to focus on exploring the role of adjuvant therapies and surveillance for recurrence in high-stage cSCC. The evolving landscape of cSCC management demands ongoing investigation to optimize patient outcomes and health care efficiency.
eTable. Fine-Gray Subdistribution Hazard Models
Data Sharing Statement
References
- 1.Jambusaria-Pahlajani A, Kanetsky PA, Karia PS, et al. Evaluation of AJCC tumor staging for cutaneous squamous cell carcinoma and a proposed alternative tumor staging system. JAMA Dermatol. 2013;149(4):402-410. doi: 10.1001/jamadermatol.2013.2456 [DOI] [PubMed] [Google Scholar]
- 2.Tolkachjov SN, Brodland DG, Coldiron BM, et al. Understanding Mohs micrographic surgery: a review and practical guide for the nondermatologist. Mayo Clin Proc. 2017;92(8):1261-1271. doi: 10.1016/j.mayocp.2017.04.009 [DOI] [PubMed] [Google Scholar]
- 3.Rapini RP. Comparison of methods for checking surgical margins. J Am Acad Dermatol. 1990;23(2, pt 1):288-294. doi: 10.1016/0190-9622(90)70212-Z [DOI] [PubMed] [Google Scholar]
- 4.Kimyai-Asadi A, Goldberg LH, Jih MH. Accuracy of serial transverse cross-sections in detecting residual basal cell carcinoma at the surgical margins of an elliptical excision specimen. J Am Acad Dermatol. 2005;53(3):469-474. doi: 10.1016/j.jaad.2005.02.049 [DOI] [PubMed] [Google Scholar]
- 5.Squamous cell skin cancer (version 1.2024). National Comprehensive Cancer Network . 2023. Accessed January 26, 2024. https://www.nccn.org/professionals/physician_gls/pdf/squamous.pdf
- 6.Xu Y, Lim Y, Bordeaux J, et al. Achieving adherence with NCCN guidelines for nonmelanoma skin cancer regarding peripheral and deep en face margin assessment (PDEMA). JNCCN. 2024;22(9):e247037. doi: 10.6004/jnccn.2024.7037 [DOI] [PubMed] [Google Scholar]
- 7.Rowe DE, Carroll RJ, Day CL Jr. Prognostic factors for local recurrence, metastasis, and survival rates in squamous cell carcinoma of the skin, ear, and lip. Implications for treatment modality selection. J Am Acad Dermatol. 1992;26(6):976-990. doi: 10.1016/0190-9622(92)70144-5 [DOI] [PubMed] [Google Scholar]
- 8.van Lee CB, Roorda BM, Wakkee M, et al. Recurrence rates of cutaneous squamous cell carcinoma of the head and neck after Mohs micrographic surgery vs. standard excision: a retrospective cohort study. Br J Dermatol. 2019;181(2):338-343. doi: 10.1111/bjd.17188 [DOI] [PubMed] [Google Scholar]
- 9.Lacerda PN, Lange EP, Luna NM, Miot HA, Abbade LPF. Efficacy of micrographic surgery versus conventional excision in reducing recurrence for basal cell carcinoma and squamous cell carcinoma: a systematic review and meta-analysis. J Eur Acad Dermatol Venereol. 2024;38(6):1058-1069. doi: 10.1111/jdv.19743 [DOI] [PubMed] [Google Scholar]
- 10.Soleymani T, Brodland DG, Arzeno J, Sharon DJ, Zitelli JA. Clinical outcomes of high-risk cutaneous squamous cell carcinomas treated with Mohs surgery alone: an analysis of local recurrence, regional nodal metastases, progression-free survival, and disease-specific death. J Am Acad Dermatol. 2023;88(1):109-117. doi: 10.1016/j.jaad.2022.06.1169 [DOI] [PubMed] [Google Scholar]
- 11.Matsumoto A, Li JN, Matsumoto M, Pineider J, Nijhawan RI, Srivastava D. Factors predicting outcomes of patients with high-risk squamous cell carcinoma treated with Mohs micrographic surgery. J Am Acad Dermatol. 2021;85(3):588-595. doi: 10.1016/j.jaad.2021.01.063 [DOI] [PubMed] [Google Scholar]
- 12.Harris BN, Bayoumi A, Rao S, Moore MG, Farwell DG, Bewley AF. Factors associated with recurrence and regional adenopathy for head and neck cutaneous squamous cell carcinoma. Otolaryngol Head Neck Surg. 2017;156(5):863-869. doi: 10.1177/0194599817697053 [DOI] [PubMed] [Google Scholar]
- 13.Burton KA, Ashack KA, Khachemoune A. Cutaneous squamous cell carcinoma: a review of high-risk and metastatic disease. Am J Clin Dermatol. 2016;17(5):491-508. doi: 10.1007/s40257-016-0207-3 [DOI] [PubMed] [Google Scholar]
- 14.Lam JKS, Sundaresan P, Gebski V, Veness MJ. Immunocompromised patients with metastatic cutaneous nodal squamous cell carcinoma of the head and neck: poor outcome unrelated to the index lesion. Head Neck. 2018;40(5):985-992. doi: 10.1002/hed.25069 [DOI] [PubMed] [Google Scholar]
- 15.Udkoff J, Beal BT, Brodland DG, Knackstedt T. Cost effectiveness of intermediate-risk squamous cell carcinoma treated with Mohs micrographic surgery compared with wide local excision. J Am Acad Dermatol. 2022;86(2):303-311. doi: 10.1016/j.jaad.2021.07.059 [DOI] [PubMed] [Google Scholar]
- 16.Que SKT, Zwald FO, Schmults CD. Cutaneous squamous cell carcinoma: management of advanced and high-stage tumors. J Am Acad Dermatol. 2018;78(2):249-261. doi: 10.1016/j.jaad.2017.08.058 [DOI] [PubMed] [Google Scholar]
- 17.Ruiz ES, Kus KJB, Smile TD, et al. Adjuvant radiation following clear margin resection of high T-stage cutaneous squamous cell carcinoma halves the risk of local and locoregional recurrence: a dual-center retrospective study. J Am Acad Dermatol. 2022;87(1):87-94. doi: 10.1016/j.jaad.2022.03.044 [DOI] [PubMed] [Google Scholar]
- 18.Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399-424. doi: 10.1080/00273171.2011.568786 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable. Fine-Gray Subdistribution Hazard Models
Data Sharing Statement
