OBJECTIVE: To determine the accuracy of cytology in the intraoperative diagnosis of central nervous system (CNS) lesions in human immunodeficiency (HIV)-positive patients. STUDY DESIGN: We prospectively studied 75 cases of computed tomography- or magnetic resonance imaging-guided brain biopsies performed with stereotactic instrumentation and a Nashold biopsy cannula over the course of five years. Biopsy samples were sent for both frozen section and immediate cytologic evaluation. Crush preparations (Papanicolaou and Diff-Quik stained) were used for cytologic assessment. There were 23 cases of progressive multifocal leukoencephalopathy (PML), 8 of toxoplasmosis (toxo), 3 of herpes simplex virus, 26 of lymphoma, 3 of HIV encephalitis, 1 of melanoma, 1 of hamartoma and 10 of nonspecific changes (paraffin section). RESULTS: Using permanent hematoxylin and eosin-stained histologic sections as the "gold standard," frozen section had a sensitivity of 78% and a specificity of 90%, while cytology had a sensitivity of 88% and specificity of 90%. Most of the false negative cases in cytology and frozen section were due to the predominance of necrosis and/or gliosis, present in six cases of toxo. Two of the false positive cases in frozen section (diagnosed as lymphoma) showed toxo, while two false positive cases in cytology (diagnosed as PML) showed only gliosis with negative immunoperoxidase staining for PML in the permanent sections. CONCLUSION: First, cytology had higher sensitivity that frozen section. Second, cytology provided faster results in most instances, primarily due to the nature of specimen preparation. Third, most misdiagnoses occurred in infectious diseases, especially toxo; this should therefore be kept in mind when nonspecific changes with an atypical lymphocytic infiltrate are seen. Fourth, cytology can be an alternative to frozen section for the intraoperative diagnosis of CNS lesions in HIV-positive patients. Another advantage of cytology is elimination of the need for cutting potentially infectious fresh tissue.
OBJECTIVE: To determine the accuracy of cytology in the intraoperative diagnosis of central nervous system (CNS) lesions in human immunodeficiency (HIV)-positivepatients. STUDY DESIGN: We prospectively studied 75 cases of computed tomography- or magnetic resonance imaging-guided brain biopsies performed with stereotactic instrumentation and a Nashold biopsy cannula over the course of five years. Biopsy samples were sent for both frozen section and immediate cytologic evaluation. Crush preparations (Papanicolaou and Diff-Quik stained) were used for cytologic assessment. There were 23 cases of progressive multifocal leukoencephalopathy (PML), 8 of toxoplasmosis (toxo), 3 of herpes simplex virus, 26 of lymphoma, 3 of HIV encephalitis, 1 of melanoma, 1 of hamartoma and 10 of nonspecific changes (paraffin section). RESULTS: Using permanent hematoxylin and eosin-stained histologic sections as the "gold standard," frozen section had a sensitivity of 78% and a specificity of 90%, while cytology had a sensitivity of 88% and specificity of 90%. Most of the false negative cases in cytology and frozen section were due to the predominance of necrosis and/or gliosis, present in six cases of toxo. Two of the false positive cases in frozen section (diagnosed as lymphoma) showed toxo, while two false positive cases in cytology (diagnosed as PML) showed only gliosis with negative immunoperoxidase staining for PML in the permanent sections. CONCLUSION: First, cytology had higher sensitivity that frozen section. Second, cytology provided faster results in most instances, primarily due to the nature of specimen preparation. Third, most misdiagnoses occurred in infectious diseases, especially toxo; this should therefore be kept in mind when nonspecific changes with an atypical lymphocytic infiltrate are seen. Fourth, cytology can be an alternative to frozen section for the intraoperative diagnosis of CNS lesions in HIV-positivepatients. Another advantage of cytology is elimination of the need for cutting potentially infectious fresh tissue.