Histologic regression in melanoma is associated with a significantly lower risk of sentinel lymph node positivity and may serve as a prognostic factor when deciding whether to perform a sentinel lymph node biopsy, according to a study published online Sept. 2 in JAMA Dermatology.

Researchers conducted a meta-analysis of 14 studies in 10,098 patients who showed histologic regression of a primary melanoma and whose sentinel lymph node status was available. In the 14 studies combined, the analysis showed patients with histologic regression had a 44% lower likelihood of having a positive sentinel lymph node than patients without regression, reported Dr. Simone Ribero of the University of Turin (Italy), and coauthors (JAMA Dermatol. 2015 Sep 2. doi: 10.1001/jamadermatol.2015.2235 ).

This figure was even lower for patients enrolled in high-quality studies, compared with lower-quality studies, they said. In the nine studies considered high quality, patients with regression had a 52% lower likelihood of having a positive sentinel lymph node, while those enrolled in the five studies considered low quality had a 27% lower likelihood of having a positive sentinel lymph node.

The prognostic significance of regression in primary melanoma has been confounded by the fact that disappearance of part of the tumor can result in an underestimation of the original Breslow thickness. Some previous studies have found an increase in sentinel lymph node positivity associated with regression, while other studies suggested that regression was in fact a protective factor against sentinel lymph node metastasis.

Although the studies used different definitions of histologic regression and all were observational (counted among the limitations of the studies), the authors concluded that regression was prognostically favorable. “The results of this meta-analysis may be useful when deciding to offer SLNB [sentinel lymph node biopsy] to patients with regressions of melanomas,” they wrote, adding: “It may help clinicians make a final selection of the most appropriate patients for this procedure.”

The authors had no conflicts of interest to declare.