Woolson and Folberg disagree with our interpretation of their results on the prognostic value of vascular patterns in uveal melanoma.1 They misquote us by omission of the following highlighted words from our article,2 "Although in multivariate analysis, Folberg et al12 found mitotic counts to be the most significant single predictor of melanoma metastases, this was not emphasized, and they believed that vascular patterns were more important." They performed a stepwise procedure to develop a multivariate model. In their final model, mitotic counts was the most statistically significant factor with the smallest P value. Given our wording restricted to that result, it is unclear why they took issue with our comment.
Kroll S, Char DH. A Response to Uveal Melanoma: Growth Rate and Prognosis—Reply. Arch Ophthalmol. 1998;116(3):399-400. doi: