To study the DNA ploidy pattern of uterine cervical squamous intraepithelial lesions (SILs) and its diagnostic and prognostic significance.Methods:
The study included 31 cases of SIL: 11 low-grade (LSIL) and 20 high-grade (HSIL). Feulgen–pararosaniline staining was performed on previously Papanicolaou-stained smears and a DNA image cytometric study was performed. An internal reference was used to calibrate the samples.Results:
All 31 cases of SIL, either LSIL or HSIL, were non-diploid. Of the 11 cases of LSIL, four were tetraploid and seven were aneuploid, whereas, of the 20 cases of HSIL, four were tetraploid and 16 were aneuploid. Stemline aneuploidy was not a significant discriminator between LSIL and HSIL (P = 0.32). Based on single-cell analysis, HSIL cases had significantly higher DNA content than LSIL cases (P < 0.01). When a mean of 30% or more was used for the 6c-exceeding event (6cEE) value, the sensitivity and specificity to indicate HSIL were 83% and 64%, respectively, with a positive predictive value (PPV) of 81% and negative predictive value (NPV) of 65%. All HSIL cases were cervical intraepithelial neoplasia grade 2 or worse (CIN2+) on biopsy. In addition, cases which showed recurrence had more DNA content by single-cell analysis than those with an indolent clinical behaviour: P = 0.04 and P = 0.03 for LSIL and HSIL, respectively.Conclusions:
Image cytometric DNA analysis is a useful technique for diagnostic and prognostic purposes in uterine cervical SIL when appropriate ‘c’ values are used in single-cell analysis. We propose that a >6c DNA content of 30% is useful as a cut-off level for predicting cases with CIN2+ in DNA image cytometry of cervical smears.