Vertical Growth Phase as a Prognostic Factor for Sentinel Lymph Node Positivity in Thin Melanomas: A Systematic Review and Meta-Analysis.

BACKGROUND

The 2010 American Joint Committee on Cancer guidelines recommended consideration of sentinel lymph node biopsy for thin melanoma (Breslow thickness <1.0 mm) with aggressive pathologic features such as ulceration and/or high mitotic rate. The therapeutic benefit of biopsy-based treatment remains controversial. The authors conducted a meta-analysis to estimate the risk and outcomes of sentinel lymph node positivity in thin melanoma, and examined established and potential novel predictors of positivity.

METHODS

Three databases were searched by two independent reviewers for sentinel lymph node positivity in patients with thin melanoma. Study heterogeneity, publication bias, and quality were assessed. Data collected included age, sex, Breslow thickness, mitotic rate, ulceration, regression, Clark level, tumor-infiltrating lymphocytes, and vertical growth phase. Positivity was estimated using a random effects model. Association of positivity and clinicopathologic features was investigated using meta-regression.

RESULTS

Ninety-three studies were identified representing 35,276 patients with thin melanoma who underwent sentinel lymph node biopsy. Of these patients, 952 had a positive sentinel lymph node biopsy, for an event rate of 5.1 percent (95 percent CI, 4.1 to 6.3 percent). Significant associations were identified between positivity and Breslow thickness greater than 0.75 mm but less than 1.0 mm, mitotic rate, ulceration, and Clark level greater than IV. Seven studies reported on vertical growth phase, which was strongly associated with positivity (OR, 4.3; 95 percent CI, 2.5 to 7.7).

CONCLUSIONS

To date, this is the largest meta-analysis to examine predictors of sentinel lymph node biopsy positivity in patients with thin melanoma. Vertical growth phase had a strong association with biopsy positivity, providing support for its inclusion in standardized pathologic reporting.