Abstract
The accuracy of 3 different ultrastructural methods used to confirm clinically proved cases of mycosis fungoides (MF) from [human] skin biopsy specimens was compared. The 1st method is a qualitative method and confirmed the diagnosis of MF in only 44% of the patients with MF. The 2nd method involved image analysis of lymphocytic nuclei and confirmed the diagnosis of MF in only 67% of the patients with MF. The most sensitive ultrastructural method for confirming the diagnosis of MF was a simple scoring of the number of sharply angled nuclear invaginations in 100 lymphocytes (method 3). Control biopsy specimens had many more lymphocytes (19-55%) with no sharply angled nuclear invaginations compared with those from patients with MF (3-15%). The clinical diagnosis of MF was confirmed in 100% of the patients with MF using the 3rd method.