Deceiving high-grade cervical dysplasias identified as human papillomavirus non-16 and non-18 types by Invader human papillomavirus assays

Histomorphologic assessment and distribution of high-risk human papillomavirus (HPV) types in cervical high-grade squamous intraepithelial lesions with unusual histomorphologic features

In rare cases, equivocal histomorphology ('deceiving dysplasia') does not allow immediate diagnosis of cervical high-grade squamous intraepithelial lesion (HSIL). We studied whether these cases are correlated with specific high-risk human papillomavirus (hr HPV) types. During 2011-2017, 39 cases of p16-positive cervical tissue biopsies with unusual ('deceiving') dysplastic histomorphology were identified and matched with the same number of controls (typical HSIL samples). Histomorphological characteristics were reviewed blindly and HPV testing was performed using the clinically validated RealTime test (Abbott) and Anyplex HPV 28 (Seegene). HPV 16 and HPV 31 were the two most frequent HPV types in both groups, although minimum, proportional, hierarchical and any etiological attribution estimates for HPV 16 were significantly lower in the deceiving group (13.2%, 21.3%, 23.7% and 23.7%) than in the control group (32.4%, 48.1%, 48.6% and 48.6%). In addition, the distribution of other hr HPV types differed between the two study groups, with five HPV types (HPV 56, 58, 59, 73 and 82) detected only in the deceiving group. Histomorphologic review of both groups (regardless of HPV type) confirmed significant differences in nuclear atypia, maximum lesion thickness and cellularity, although these were diminished when cross-comparisons between HPV16/18 and non-HPV16/18 cases pooled from both study groups were evaluated. Different attribution estimates for HPV 16, HPV 16/18 and non-16/18 hr HPV types in deceiving and control groups were observed, in particular for HPV 16. However, an unusual (deceiving) histomorphology may also depend on unknown HPV-related molecular changes.

Bizarre cell dysplasia of the cervix

AIM

The aim of this study was the characterization of a new subtype of high-grade cervical squamous intraepithelial lesion (HSIL) with enlarged cells containing bizarre nuclei: so-called bizarre cell dysplasia (BCD).

METHODS

A total of 29 cervical cone biopsy samples of this type of dysplasia were studied. Multi-target polymerase chain reaction and in situ hybridization human papillomavirus (HPV) detection was performed in all cases. BCD was defined as a subtype of HSIL characterized by the presence of large dysplastic cells with abnormal, large pleomorphic nuclei or multinucleation causing nucleomegaly. This results in bizarre nuclear shapes. Bizarre cells are scattered throughout the whole thickness of the dysplastic squamous epithelium.

RESULTS

The BCD lesions arise within the conventional/classic high grade or "bland" type squamous dysplasia HSIL. Statistically they were significantly associated with HVP type 16. A significant association with other studied viruses (Herpes simplex virus [HSV]1, HSV2, Varicella zoster virus, Epstein-Barr virus, cytomegalovirus, human herpesvirus 6, and human polyomaviruses BK and JC) was not confirmed.

CONCLUSIONS

BCD involves cytologically characteristic morphologic changes that are recognizable, but which may pose some risk of misdiagnosis as low-grade squamous intraepithelial lesion due to the enlargement of dysplastic cells and multinucleation. Based on the unique histological, cytological and biological features of BCD including strong association with HPV 16 infection, we believe that this is a specific, and so far unrecognized variant of HSIL.

Variability of Pathologists' Utilization of p16 and Ki-67 Immunostaining in the Diagnosis of Cervical Biopsies in Routine Pathology Practice and Its Impact on the Frequencies of Cervical Intraepithelial Neoplasia Diagnoses and Cytohistologic Correlations

Context.—The use of p16 in cervical biopsies improves the accuracy of cervical intraepithelial neoplasia (CIN) diagnosis and grading and decreases its interpathologist variability.

Objective.—To determine the impact of the frequency of use of p16 immunostains in cervical biopsies on pathologists' diagnoses of CIN grade 1 and grade 2 or above (CIN1 and CIN2+) and on cytohistologic correlations.

Design.—We identified all cervical biopsy specimens with cytologic correlations subjected or not to p16 staining from January 1, 2005, to September 30, 2010; calculated each pathologist's percentage of p16 use; and correlated it with their major cytohistologic discrepancy rates, CIN2+ diagnoses, and CIN1/CIN2+ ratios.

Results.—During the study period, each of the 23 pathologists interpreted 59 to 1811 (mean, 518) of 11 850 cervical biopsy specimens, used p16 for 0% to 21.31% (mean, 10.14%) of these, had CIN2+ detection rates of 9.5% to 24.1% (mean, 18.9%), and CIN1/CIN2+ ratios of 0.7 to 4.5 (mean, 1.5). Compared to the 12 “low users” of p16, who used p16 fewer times than the institution's mean for p16 use, the 11 “high users” of p16 diagnosed more biopsies (8391 versus 3459), had a lower rate of major cytohistologic discrepancies (12.62% versus 14.92%, P < .001), a higher rate of CIN2+ diagnoses (19.9% versus 16.4%, P < .001), a lower range of CIN2+ rates (15.0%–23.1% versus 9.5%–24.1%), and lower CIN1/CIN2+ ratios (1.2 versus 2.3).

Conclusions.—We found a high intrainstitutional variability of p16 use in cervical biopsies, CIN2+ rates, and CIN1/CIN2+ ratios. Use of p16 for greater than 10% of cervical biopsies was associated with improved cytohistologic correlation rates and with lower variability in the frequencies of histologic diagnoses.