Additional review of Mohs slides to optimize Mohs micrographic surgery

Weakly Supervised Classification of Mohs Surgical Sections Using Artificial Intelligence

Basal cell carcinoma (BCC) is the most frequently diagnosed form of skin cancer, and its incidence continues to rise, particularly among older individuals. This trend puts a significant strain on health care systems, especially in terms of histopathologic diagnostics required for Mohs micrographic surgery (MMS), which is used to treat BCC in sensitive locations to minimize tissue loss. This study aims to address the challenges in BCC detection within MMS whole-slide images by developing and evaluating a deep learning model that bridges weakly supervised learning with interpretable segmentation-based methods through attention maps. Utilizing data sets from 2 medical centers, the model demonstrated an average area under the receiver operating characteristic curve (AUC) of 0.958 on internal testing and an AUC of 0.934 on an independent third external data set despite no fine-tuning or preprocessing for the latter. Attention maps provided insights into the model's decision making, highlighting critical regions for slide-level classification. The sensitivity of the attention maps in localizing tumor regions was 0.853 when no filtering was applied and gave 8 revision false positives per slide on average and was reduced to an average of 2 false positives per slide with a sensitivity of 0.873 when detections smaller than 200 μm were removed from the attention maps. These findings indicate that the deep learning model is highly effective in detecting BCC in MMS whole-slide images, with robust performance across different data sets and conditions. The use of attention maps enhances the model's interpretability, making it a promising tool for aiding dermatopathologists and MMS surgeons.

Mohs micrographic surgery and dermatopathology concordance: An analysis of 1421 Mohs cases over 17 years

BACKGROUND

The success of Mohs micrographic surgery depends on the surgeon's ability to correctly interpret intraoperative frozen sections.

OBJECTIVE

This retrospective study analyzed the rate of concordance between Mohs surgeons and dermatopathologists in reading slides from Mohs surgery cases.

METHODS

A dermatopathologist reviewed all the frozen sections and the corresponding Mohs map for every 30th Mohs case at a practice employing 6 different Mohs surgeons during 2001-2017. Cases in which the dermatopathologist and the Mohs surgeon disagreed on the interpretation were noted.

RESULTS

The concordance rate between Mohs surgeons and dermatopathologists was 99.79%. The 3 discordant cases included a case of squamous cell carcinoma, a case of superficial basal cell carcinoma, and a case of hypertrophic squamous cell carcinoma in situ.

LIMITATIONS

This analysis is limited to fellowship-trained Mohs surgeons and, therefore, might not be applicable to all physicians who perform Mohs.

CONCLUSION

Fellowship-trained Mohs surgeons show high concordance with board-certified dermatopathologists in the accurate and precise interpretation of histology slides in the setting of Mohs micrographic surgery.

Histopathologic upgrading of cutaneous squamous cell carcinomas during Mohs micrographic surgery: A retrospective cohort study

BACKGROUND

Initial biopsies of cutaneous squamous cell carcinomas (cSCCs) may not reveal aggressive histologic features, which would otherwise inform appropriate surgical management and patient education.

OBJECTIVE

To assess the incidence of, and risk factors for, histopathologic upgrading of cSCC during Mohs micrographic surgery (MMS).

METHODS

This was a retrospective cohort study of invasive cSCCs treated with MMS between 2017 and 2019 at 1 academic institution. An "upgrade" was defined as a lesser degree of differentiation (poor or moderate) and/or bony or perineural invasion identified during MMS that was not reported in histopathologic evaluation of the initial biopsy.

RESULTS

Of the 1558 tumors studied, 115 (7.4%) were upgraded during MMS. In multivariate logistic regression analysis, male sex, prior field treatment, location on the ear/lip, rapid growth of cSCC, and tumor diameter ≥2 cm were significant predictors of tumor upgrading. Upgraded tumors were more likely to require ≥3 MMS stages to clear, complicated closure (flap or graft), or outside (referral) repairs.

LIMITATIONS

Single-center study, retrospective, and inter-rater variability.

CONCLUSIONS

A significant proportion of cSCCs is histopathologically upgraded with more aggressive features during MMS. Routinely documented patient and tumor characteristics can predict tumor upgrading and assist clinicians in directing the management of potentially high-risk cSCC patients.

Mohs Micrographic Surgery Dermatopathology Concordance in Canada: A Single-Institution Experience

BACKGROUND:

Mohs micrographic surgery (MMS) is a surgical modality that achieves high cure rates of nonmelanoma skin cancers but is dependent on accurate histologic examination of surgical margins. Therefore, quality assurance is essential to ongoing assessment of histological margins.

OBJECTIVES:

To prospectively determine the concordance rate between a Mohs surgeon (MS) and dermatopathologist (DP) with respect to tumour status (ie, present or absent) and tumour type. Secondary end points were to determine the relationship between discordant interpretations and slide quality and to assess the feasibility of using an electronic webform for data collection.

METHODS:

Ten percent (10%) of the planned MMS cases between January 2015 and March 2016 were randomly selected by a histotechnologist at the start of each month. The MS and DP were blinded to the chosen cases, and slides were reviewed independently at the beginning of the following month. Data were collected using an online webform. A blinded third party determined if there were discrepancies in interpretation, and any discordant slides were reviewed together and a consensus was reached.

RESULTS:

A total of 270 slides from 54 total cases were reviewed. The overall tumour status concordance rate was 93.6%. Cohen's κ was 0.86. Tumour type concordance was 98.9%. No discrepancy required a change in patient care. All discrepant slides were from cases that required multiple stages.

CONCLUSIONS:

This is the first study looking at MS-DP concordance in Canada, and our findings support the MS acting as his or her own pathologist.

Reliability of diagnosis from Mohs slides: interpersonal and intrapersonal agreement on basal cell carcinoma presence and histological subtype

BACKGROUND

The success of Mohs micrographic surgery (MMS) depends partly on the correct diagnosis of slides.

OBJECTIVES

To determine reliability of diagnosis from Mohs slides.

METHODS

This was a prospective study evaluating the reliability of diagnosis from Mohs slides of basal cell carcinoma (BCC) presence, BCC location on the slide and BCC subtype among six raters who independently assessed 50 Mohs slides twice with a 2-month interval. Slides were randomly selected whereby difficult-to-diagnose slides were oversampled. For each slide, a reference diagnosis was established by an expert panel. Cohen's kappa (κ) was calculated to determine levels of agreement interpersonally (rater vs. reference diagnosis) and intrapersonally (rater at T1 vs. T2). Multivariable logistic regression was used to determine independent risk factors for slides with interpersonal discordant diagnosis. The variables studied were BCC presence, whether a slide was scored as easy or difficult to diagnose, review duration of the 50 slides, profession and years of experience in diagnosis from Mohs slides.

RESULTS

Interpersonal and intrapersonal agreement were substantial on BCC presence (κ = 0·66 and 0·68) and moderate on BCC subtype (κ = 0·45 and 0·55). Slides that were scored as difficult to diagnose were an independent risk factor for interpersonal discordant diagnosis on BCC presence (odds ratio 3·54, 95% confidence interval 1·81-6·84).

CONCLUSIONS

Reliability of diagnosis from Mohs slides was substantial on BCC presence and moderate on BCC subtype. For slides that are scored difficult to diagnose, a second opinion is recommended to prevent misinterpretation and thereby recurrence of skin cancer.