OBJECTIVE: To improve ploidy analysis in the detection of high-grade cervical dysplasias by combining it with anti-Ki-67 immunocytochemistry in a double staining procedure to distinguish between cells with abnormal DNA content and normal cycling cells. STUDY DESIGN: Cervical cytology specimens from 49 patients with various diagnoses, mostly dysplasias, from a previous study were used. Samples were double stained with Feulgen-thionin and anti-Ki-67 immunocytochemistry. Ki-67-negative cells were noncycling, so nondiploid Ki-67-negative cells were likely truly abnormal cells. RESULTS: The area under the receiver operating characteristic curve for the ability to identify high-grade dysplasias was 0.73 for double staining and 0.74 for thionin-only ploidy analysis on Cytospin specimens. At 90% specificity, sensitivities for double staining and thionin alone were 45% and 32%, respectively, but the difference was not statistically significant. CONCLUSION: Double staining with Feulgen-thionin and anti-Ki-67 immunocytochemistry does not improve the ability of ploidy analysis of cervical cytology specimens to separate high-grade and low-grade dysplasias, but our insights into the technical aspects of double staining, especially the effects of antigen retrieval, give hope that this technique could be applied to other immunocytochemical stains that would have a greater ability to improve ploidy analysis.
OBJECTIVE: To improve ploidy analysis in the detection of high-grade cervical dysplasias by combining it with anti-Ki-67 immunocytochemistry in a double staining procedure to distinguish between cells with abnormal DNA content and normal cycling cells. STUDY DESIGN: Cervical cytology specimens from 49 patients with various diagnoses, mostly dysplasias, from a previous study were used. Samples were double stained with Feulgen-thionin and anti-Ki-67 immunocytochemistry. Ki-67-negative cells were noncycling, so nondiploid Ki-67-negative cells were likely truly abnormal cells. RESULTS: The area under the receiver operating characteristic curve for the ability to identify high-grade dysplasias was 0.73 for double staining and 0.74 for thionin-only ploidy analysis on Cytospin specimens. At 90% specificity, sensitivities for double staining and thionin alone were 45% and 32%, respectively, but the difference was not statistically significant. CONCLUSION: Double staining with Feulgen-thionin and anti-Ki-67 immunocytochemistry does not improve the ability of ploidy analysis of cervical cytology specimens to separate high-grade and low-grade dysplasias, but our insights into the technical aspects of double staining, especially the effects of antigen retrieval, give hope that this technique could be applied to other immunocytochemical stains that would have a greater ability to improve ploidy analysis.