Cytologic Features of Squamous Cell Carcinoma in Conventional Smears: Comparison of Cases That Performed Poorly With Those That Performed Well in the College of American Pathologists Interlaboratory Comparison Program in Cervicovaginal Cytology

Cervicovaginal cytology is associated with a significant error rate for both low- and high-grade squamous intraepithelial lesions and carcinoma.1 Previous studies have suggested that cases of high-grade squamous intraepithelial lesion that are missed on conventional smears have characteristic cytologic features, including relatively few abnormal cells, relatively small size, and hypochromatic chromatin.2,3 Recent studies from the College of American Pathologists Interlaboratory Performance Program in Cervicovaginal Cytology (PAP) suggest that cases of high-grade squamous intraepithelial lesion in both conventional smears and ThinPrep Pap Test specimens that are missed also have significantly fewer abnormal cells than cases that are reliably identified.4,5 This trend is also found for low-grade squamous intraepithelial lesion,6 although the absolute number of cells necessary to make a case regularly and reliably identifiable is only half that of cases of high-grade squamous intraepithelial lesion.

Previous studies have also documented that cases of squamous cell carcinoma in ThinPrep specimens that are always identified are less likely to have Trichomonas or features of repair than cases that are sometimes missed.7 We sought to determine if characteristic features could be identified that would distinguish cases of squamous cell carcinoma in conventional smears that are regularly and reliably identified from those that are sometimes missed.

The PAP program is a quarterly, glass slide quality improvement program. The College of American Pathologists Laboratory Accreditation Program requires that all laboratories evaluating gynecologic cytology enroll in the PAP program or an equivalent glass slide program. Cytology laboratories of all types participate, with the largest number (approximately 60%) being hospital laboratories. In addition, independent laboratories, federal and government laboratories, university laboratories, and others such as those associated with a group practice or physician's office participate.

Participant laboratories generously contribute slides to the program. Submitted slides with a diagnosis of squamous cell carcinoma must be biopsy confirmed. After receipt and accessioning into the program, the slides are reviewed by at least 3 experienced cytopathologists from the College of American Pathologists Cytopathology Resource Committee. Before acceptance into the program, each slide must be judged to be of good technical quality and an excellent example of the reference diagnosis. All 3 reviewers must agree on the exact target diagnosis, and this must agree with the submitted and biopsy diagnosis prior to accepting a slide for circulation into an educational set.

The PAP program consists of 5 glass slides of cervicovaginal material mailed 4 times per year. The coded answer sheets have diagnostic menus using terminology modified from the Bethesda System. Referenced slides are placed into 1 of 3 selection series: the 000 series for unsatisfactory slides; the 100 series for normal, infectious, and reparative conditions; and the 200 series for epithelial cell abnormalities and carcinoma.

Conventional smears in the program from 1989 to 2003 with a reference diagnosis of squamous cell carcinoma and at least 15 individual observations by pathologists or cytotechnologists were identified. Cases with poor performance and excellent performance were identified on the basis of the distribution of results in the program. Cases that performed poorly were defined as those that were misidentified as negative, repair, or unsatisfactory (000 or 001 series) by at least 2 separate laboratories and at least 10% of individual participants. The group of cases that performed well were those that the participants correctly assigned to the correct 200 series 100% of the time as well as classified specifically as squamous cell carcinoma at least 75% of the time.

Cases were originally screened and reviewed by one author (A.A.R.) to evaluate possible cytologic features that might be different between the 2 groups. A variety of factors were considered, including tumor diathesis, the presence of nucleoli, and bizarre nuclei. Features that might be significant were than evaluated by 3 authors (M.R.H., J.H.H., G.G.B.) independently without knowledge of the case performance and the consensus diagnosis used for analysis. The numbers of abnormal cells were categorized as less than 100, 101 through 250, 251 through 500, 501 through 1000, or greater than 1000 cells per slide. Big cell size was defined as cells with a nuclear diameter twice that of an intermediate squamous cell, similar to that seen in cases of moderate dysplasia (Figure, A and B). Cases were considered keratinized if they had extensive keratinization throughout the slide (Figure, A through C). Obscuring inflammation was defined as abundant acute inflammation covering a significant portion of the slide.

The statistical significance of the various cytologic features was evaluated using a 2-tailed Fisher exact test. A P value of .05 was used as a threshold for significance.

A total of 25 cases were evaluated in this report (see the Table). Among them, 2387 individual interpretations were recorded, 1676 for those that performed well and 711 for those that performed poorly. Of the 86 incorrect responses for those that performed poorly, 6.2% were for repair, and 0.8% were for Trichomonas. There was no incorrect response for cases that performed well.

Cases that performed well were significantly more likely to have greater than 1000 dysplastic cells (16/17 vs 4/8, P = .02) and be keratinized (13/17 vs 1/8, P = .007). Obscuring inflammation (7/17 good cases, 4/4 poor cases, P = .39) and large cell size (12/17 good cases, 4/4 poor cases, P = .99) were not significant.

The College of American Pathologists Interlaboratory Comparison Program in Gynecologic Cytology presents an excellent opportunity to evaluate the significance of cytologic findings in a series of extremely well-characterized cases. The diagnoses are confirmed by biopsy in every case of squamous intraepithelial lesion or above, are reviewed by 3 cytopathology members of the committee, and, most importantly, are then evaluated by their performance in the program on the basis of the interpretation of practicing cytologists. As such, the performances we document in these cases may most accurately reflect the interpretive abilities of practicing cytotechnologists and pathologists.

Previous studies have shown that cases of high-grade squamous intraepithelial lesion that are misinterpreted on conventional smears may have characteristic cytologic features, including relatively few abnormal cells, relatively small cells, and hypochromatic chromatin.2,3 Previous studies from this program also suggest that cases of high-grade squamous intraepithelial lesion in ThinPrep specimens that are often misinterpreted have significantly fewer cells than those that are reliably identified4 and, in conventional smears, have both significantly fewer cells and a smaller cell size.5 

In contrast to this, previous studies of the cytologic features of squamous cell carcinoma in ThinPrep specimens suggest that distracting factors, such as Trichomonas and features of repair, are the most common cause of misinterpreting cases,7 rather than factors that are more clearly related to screening error. The results obtained in this study suggest that in cases of squamous cell carcinoma in conventional smears, factors related to screening rather than distracting factors are more important. As a result, these cases perform in a manner that is more similar to high-grade squamous intraepithelial lesion than to squamous cell carcinoma in ThinPrep specimens. Specifically, the number of cells and prominent keratinization were significant factors in causing a case to be missed, suggesting that the most common cause of missing a case was screening error. Trichomonas was an uncommon incorrect response in this study, unlike in cases of squamous cell carcinoma in ThinPrep specimens that performed poorly. On the other hand, reparative change was a relatively frequent misdiagnosis, suggesting that misinterpretation of some of the features of squamous cell carcinoma as repair remains a source of error in conventional smears as well as ThinPrep specimens. Taken together, while there are similarities between the performance of cases of squamous cell carcinoma in conventional and ThinPrep specimens, these results at least suggest that there are differences in how cases of squamous cell carcinoma are identified and interpreted in conventional smears and ThinPrep specimens. Further studies of this area appear warranted.

In conclusion, unlike in cases of squamous cell carcinoma in ThinPrep specimens, in conventional smears, screening errors related to the number of abnormal cells and prominent keratinization are cytologic features that are highly correlated with the performance of individual cases.