MONTREAL Regression of a thin melanoma should not be the sole criterion to justify a sentinel lymph node biopsy, according to a retrospective study.
Regression alone is insufficient in the absence of other recognized high-risk predictors of sentinel lymph node (SLN) involvement, Dr. Pierre-Luc Dion said. More established risk factors include Breslow thickness greater than 1 mm, Clark level IV or V, and lesion ulceration.
Prognosis generally is better with thinner melanoma. Identification of the minority of thin melanoma patients who are at high risk for metastasis, however, remains a clinical challenge.
"There is an ongoing debate if regression is good news or bad news for patients. Some think it shows an immune response against tumor cells, but others say it can lead to an underestimation of the real thickness," Dr. Dion said at the annual conference of the Canadian Dermatology Association.
There is no consensus in the literature. One study of 65 thin melanomas with regression found that only 2 lesions (3%) had a positive SLN biopsy result (Ann. Surg. Oncol. 2003;10:55861).
Another study found only 1 patient with a positive SLN biopsy among 344 who had thin melanomas (mean Breslow thickness of 1.1 mm) that had shown histologic regression (Ann. Surg. Oncol. 2008;15:31622).
Other investigators have proposed that tumor regression predicts a higher risk of sentinel node involvement in thin melanomas (Br. J. Dermatol. 2003;149:6623).
Dr. Dion and his colleagues assessed 693 patients treated at the melanoma clinics for Le Centre Hospitalier Universitaire de Québec-L'Hôtel-Dieu de Québec in Saint-Nicolas and Québec City. All of the patients underwent sentinel lymph node biopsy between 1996 and 2007. The median Breslow thickness was 2.28 mm and mean age was 55 years.
A total of 653 patients had a lesion that was greater than 1 mm. Their prognoses were compared with those of a group of 40 others with thinner melanoma and regression. Regression was determined by a pathologist, who found at least a 15% reduction in the lesion size on multiple slides.
Among the patients with regression, none had a positive lymph node, compared with 146 (22%) in the control group, suggesting regression alone is not a reliable predictor, said Dr. Dion of the hospital in Saint-Nicolas.
Of the 40 patients, 1 experienced complete regression. Another six patients in this group developed in situ melanoma. The median Breslow thickness among the remaining 33 patients was 0.6 mm. At a mean follow-up of almost 4 years, recurrence occurred in two patients, including one local recurrence and one transit metastasis, Dr. Dion said.