Local Recurrence, Metastasis, and Mortality Rates Among Patients With Thin and In Situ Melanomas

This study looked at outcomes among 580 patients with melanoma in situ (MIS) and 542 patients with thin non-ulcerated melanomas ≤1 mm in thickness that were not transected at the base. The patients were followed for at least 4 years, with a mean follow-up time of 6.5 years. In the MIS group, 14 patients (2.4%) had a local recurrence, 2 (0.3%) had a visceral metastasis, and 2 (0.3) died owing to melanoma. Among patients with thin non-ulcerated melanomas, 7 (1.3%) had a local recurrence, 12 (2.2%) had a metastasis, and 6 (1.1%) died owing to melanoma. The mean Breslow depth for thin melanomas was 0.45 mm, which is consistent with the low rates of metastasis and death noted here. The results also showed that 31.9% of the cohort developed a basal cell carcinoma during the follow-up period, 23.0% developed a squamous cell carcinoma, and 18.7% had a second primary melanoma. The study design is somewhat unclear, but it appears that the authors excluded patients who had a second primary melanoma; hence, the metastases theoretically came from the melanomas under question. Furthermore, they excluded patients with relevant known genetic mutations that would have predisposed them to develop an additional melanoma.

I was quite struck that 2 of the 580 patients with MIS developed a metastasis, especially because neither of those patients had a local recurrence (ie, it was not the case that the patients had a local recurrence with an invasive component that led to metastasis). The authors do not tell us whether these patients had histopathologic regression. Although the phenomenon of metastases apparently deriving from MIS has been reported many times, the rate noted in this study is higher than that I would have expected. The patient database for our pigmented lesion clinic goes back about 50 years, and there are instances from the early days of the database where patients who supposedly only had MIS developed metastases, but I do not think that there has been such a case for at least 25 years. I, personally, have never seen this happen. I should mention that I believe that the examples from our database occurred in lesions that had histopathologic regression (ie, they may not have truly been in situ).

One mechanism through which tumors labeled as MIS might metastasize is if there is invasion, but it is so focal that it is not seen on the sections examined by the dermatopathologist. There have been several studies that have found areas of invasion when supposed MIS lesions are sectioned more aggressively, but, in general, the depth of invasion in those studies is 0.2 to 0.5 mm and unlikely to have clinical significance. Nevertheless, this could explain why a seeming MIS results in metastasis.

I was also struck by the high rate of second primary melanomas in the cohort, where 18.7% of the patients had an additional melanoma within 4 years, despite the exclusion of patients with known genetic syndromes. This is a much higher number than those reported by others or those noted in my own experience. The rates of development of new squamous cell carcinoma (23.0%) and new second primary melanoma (18.7%) in the cohort are comparable, whereas my personal experience is that squamous cell carcinoma is considerably more common than a second primary melanoma. However, regardless of the exact numbers, this study is a helpful reminder that developing a second primary melanoma or non-melanoma skin cancer is not unusual for a patient with melanoma, and patients need to be vigilant about their sun protection and self-examination practices regardless of the depth of the initial tumor.