a) Atropic vaginitis: degenerate hyperchromatic nuclei and karyorrhexis in orangeophilic cells with NC ratio consistent with atrophic epithelium. This appearance could be confused with HSIL but the pattern was consistent throughout the slide and follow up confirmed as negative.(b) Endocervical polyps are associated with inflammation involving endocervical and immature metaplastic cells, which may be difficult to distinguish from each other. This cell group has some features suggesting glandular neoplasia but the variably hyperchromatic nuclear chromatin is regular and the nuclear membranes are even (unlike HSIL or CGIN). Pseudomultinucleation with abundant cytoplasm and a low nuclear/cytoplasmic is seen.(c) Non-specific endocervicitis The cells on the margin of this group are columnar but the cytoplasm is reduced in size due to mucin depletion, which is a feature of non-specific endocervicitis (and may also be seen with endocervical polyps). There is no nuclear crowding, the chromatin pattern and membranes are regular, and there is no pseudostratification to suggest CGIN.(d) Repair/regeneration This conventional Pap smear was incorrectly reported as suspicious of glandular neoplasia and is an example of extreme changes that may be seen in repair and regeneration. The cervix appeared irregular at colposcopy and was also suspicious for malignancy; a small diagnostic LLETZ was carried out. Histology showed follicular cervicitis with marked regenerative and repair changes in the overlying epithelium. In retrospect, the even nuclear membranes and fine chromatin pattern in flat rather than 3-dimensional cells confirmed this to be a false positive. Nevertheless, histological correlation was essential to provide the final diagnosis.
(a) ThinPrep LBC preparation showing dense hyperchromatic nuclei with NC ratio normal for immature metaplasia; irregular chromatin (dyskaryosis) is lacking.(b) Conventional smear incorrectly reported as moderate dyskaryosis.
Trichomonas vaginalis was noted on review (top left of cell group) when colposcopy showed cervicitis but no evidence of CIN. Prominent nucleoli and even nuclear membranes favour a reactive process.
The term ‘atypical immature squamous metaplasia’ has been used in the past for histological biopsies, which corresponds to the difficulty of distinguishing immature metaplasia from SIL in cytology.
Immunostaining for p16 and Ki67 (Duggan et al. 2006) or CK17 and p16 (Regauer & Riech 2007).) and can distinguish immature metaplasia from LSIL and HSIL in the majority of cases.
Tuboendometrioid metaplasia
Tuboendometrioid metaplasia (TEM) and, less frequently, endometriosis may be seen on the cervix especially after procedures such as LLETZ and cone biopsy (Ismail 1991; Hirschowitz et al. 1994).
Figure 10.7 (a-f) Endometrial and tubal metaplasia mimicking HSIL
(a) Typical low-power appearances of TEM help to avoid over-diagnosis as do stromal cells at high-power.(b) Stromal cells at high-power are seen as uniform oval cells with even finely granular chromatin and inconspicuous nuclear membranes.
Tubal metaplasia
Tubal metaplasia may be seen in the absence of TEM when it may mimic HSIL.
Cytologically, the cells have plump enlarged nuclei. Basal plates can usually be seen even in the absence of cilia, which are diagnostic of tubal cells.
Tubal metaplasia with inconspicuous cilia and basal plates may be a cause of false positive reports (Figure 10.8 b-c).
(a) Tubal metaplasia in LBC with conspicuous cilia and basal plates.(b) Plump round nuclei incorrectly reported on a conventional smear as moderate dyskaryosis but reviewed as tubal metaplasia when cilia and basal plates were noted (after the histological diagnosis was known).(c) Ciliated cells of tubal metaplasia on a colposcopic biopsy of the case illustrated in (b): unnecessary treatment was avoided by cytology review at a multidisciplinary meeting.
Endometrial cells
Endometrial cells may be a cause of false positive reports when they are seen outside the dates expected by date of the last menstrual period and when they do not present the classical appearances illustrated in Chapter 9b.
Two situations when endometrial cells have presented pitfalls in diagnosis are illustrated below: i) IUCD changes when the presence of a device was not known and ii) lower uterine segment sampling, especially when the endocervical canal is shortened after excisional biopsies of CIN.
(a) Typical appearance of ‘bubble gum cells’ in a patient with an IUCD that had not been mentioned on the request form. This was incorrectly reported as atypical glandular cells in a woman over 40 years of age. Colposcopy was negative and the cytological diagnosis revised on review.(b) Lower uterine segment endometrial cells sampled after cone biopsy for CIN3. This was reported as borderline, high-grade dyskaryosis not excluded (equivalent to ASC-H) but colposcopy was negative and the report revised on review. Note the even nuclear membranes and finely granular chromatin.