Conference Coverage

Mohs surgeons, dermatopathologists nearly always agree on frozen section analysis




ORLANDO – Mohs fellowship–trained surgeons are highly skilled at reading frozen section slides obtained during surgery, agreeing with the dermatopathologist’s assessment more than 99% of the time, according to a review of more than 4,000 cases.

Among 4,145 slides in five datasets, only 28 were discordant. The concordance of surgeons’ and dermatopathologists’ assessments was quite consistent, ranging from 95%-of 99.7%.

“There is absolutely no detriment to our interpretation of these frozen section slides,” Dr. James Highsmith said at the annual meeting of the American College of Mohs Surgery. “These numbers give credence as to why we have these good cure rates. We can say with 99% accuracy that yes, we are removing these tumors from our patients.”

Dr. Highsmith, a Mohs surgeon in Birmingham, Ala., reviewed 10 years of his own data (1,720 cases), as well as four published cohorts comprising another 2,425 cases. All of the studies examined the concordance of frozen section assessments between fellowship-trained Mohs surgeons and the dermatopathologists working with them. All of the series reported concordance, discordance, sensitivity, and specificity.

His own database is the largest one yet compiled, spanning the longest period of time. Of 1,720 cases, there were eight discrepancies: one in which the surgeon identified tumor that was not supported by dermatopathology, and seven in which the surgeon missed tumor that dermatopathology confirmed was present. The concordance rate was 99.5%. This worked out to a 99.4% sensitivity and 99.8% specificity, Dr. Highsmith said.

Results were similar with the other four studies.

The largest was a 2009 retrospective review of 1,156 slides collected over 10 years; they were part of a preexisting randomized, blinded study. Of these, 32 slides (2.8%) were discordant between the Mohs surgeon and the dermatopathologist, for a 99.7% concordance rate (J Am Acad Dermatol. 2009 Jan;[1]: 94-8).

The second largest cohort was published in 1989. It comprised 1,000 slides that were evaluated by the Mohs surgeon and a general pathologist; disputed slides were read by two other Mohs surgeons, two pathologists, and a dermatopathologist. The overall concordance rate was 98.9% (J Am Acad Dermatol. 1989;20[4]:670-4).

In 2013, a Mohs surgeon published a set of 170 slides that had been randomly selected from his cases as part of a quality assurance audit. There was one discordant interpretation, for an overall concordance rate of 99.4% (Dermatol Surg. 2013;39[11]:1648-52).

The smallest study, published in 2011, reviewed 99 slides assessed by a single Mohs surgeon. The concordance rate with dermatopathology was 95%.

An overall analysis of these four cohorts determined a sensitivity of 99.4%, specificity of 99.2% and negative predictive value of 99.2%, Dr. Highsmith said.

“When a fellowship-trained Mohs surgeon looks at a slide and says there’s no more tumor present, then there’s a greater-than-99% chance that is accurate,” he said.

Dr. Highsmith had no financial disclosures.

Next Article: