Histopathologic evaluation of cervical intraepithelial neoplasia (CIN) in tissue biopsy samples: Dilemmas and solutions Sigurd Lax Department of Pathology, General Hospital Graz West Cervical carcinoma is a global burden that occurs more frequently in developing countries of Africa, Asia and Latin America compared to the industrialized world of Europe and North America. The incidence of cervical carcinoma correlates with the socioeconomic status of a population and has been strongly influenced by the implementation of cytological screening that allow the early detection of cervical precursor lesions. Organized screening programs such as in Scandinavian countries, the Netherlands and U.K. are associated with a lower incidence and mortality of cervical cancer compared to opportunistic screening like in Austria. The etiology and pathogenesis of cervical cancer is strongly linked to HPV. Almost all types of cervical cancers have HPV integrated within their genome. For the most frequent types of cervical carcinoma, squamous cell carcinoma and adenocarcinoma distinctive precancers are known. Squamous cell carcinoma is considered to develop from a precancer designated cervical intraepithelial neoplasia (CIN). CIN is graded into 3 groups, CIN 1−3 according to the degree of proliferation of atypical basal cells and the presence of mitotic figures. For grading of CIN the squamous epithelium is divided into 3 thirds. The atypical basal and parabasal cells involve the basal third in CIN 1, the basal and the middle third in CIN 2 and more than two thirds in CIN 3. In particular, CIN 1 and 2 are not well defined since the presence of mitosis as well as koilocytotic changes are considered further diagnostic criteria. For the diagnosis of CIN 3 the presence of mitoses in the superficial third of the epithelium is considered helpful. In contrast to CIN 1 and also CIN 2, CIN 3 lacks a significant amount of koilocytes. The Bethesda system, which was originally established for cervical cytology uses only two categories of HPV associated lesions, low and high grade squamous intraepithelial lesions (LSIL and HSIL) that are considered biologically distinctive. LSIL is characterized by extensive koilocytosis and a proliferation of immature, undifferentiated cells within the lower third of the epithelium. Abnormal mitosis are rare. Any anogenital HPV type may occur in LSIL and the lesions are usually diploid or polyploid. In contrast, HSIL shows a proliferation of 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 24 undifferentiated, immature cells that involves at least the middle third of the epithelium and koilocytes are scant. HSIL are usually aneuploid and associated with high risk HPV such as HPV 16. Compared to the WHO system LSIL correlates with CIN1, whereas CIN2 and 3 are summarized under HSIL. There are several problems with CIN lesions: 1. It is not clear how frequently CIN of various grade progress and show regression, respectively; 2. In particular, CIN 2 shows poor agreement among observers, even among experts; 3. There are further problems with other lesions, particularly the distinction between CIN 1 and reactive and CIN 3 and metaplastic, respectively; 4. The role of adjunct methods, in particular biomarkers for diagnostic purposes needs to clearly defined. The tendency of progression and regression of the various lesions varies between different studies. It ranges between 10 and 70% for CIN 1 and CIN 3, respectively. However, larger studies such as the Toronto Long term Follow up study for abnormal cytology revealed a more than 4-fold increased progression rate for CIN 3 compared to CIN 1. However, it is likely that even CIN 2 show a greater progression rate than expected. 40 % of underdiagnosed CIN 2 seem to regress but progression seems to depend on the type of HPV, in particular HPV 16. The reproducibility of CIN 3 is significantly better compared to CIN 2 as shown by various groups. CIN 3 seems to be more frequently associated with oncogenic HPV and abnormal cytology compared to CIN 2. CIN 2 seems to be both undercalled as CIN 1 and overcalled as CIN 3. Therefore, suggestions have been made to request a second opinion for histological diagnosis of CIN 2. CIN 2 is frequently associated with HPV 16 but it is unclear which CIN 2 show progression to CIN 3 and invasive carcinoma, respectively. Recently, p16 immunhistochemistry was suggested as diagnostic adjunct for histological diagnosis of CIN2. The implementation of p16 immunohistochemistry in daily routine seems to be of great diagnostic value in the grading of cervical lesions on biopsies. There is evidence, that the combination of HE and p16 immunhistochemistry is as good as an expert second opinion. A 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 25 diffuse strong staining is typical for CIN 3 and most CIN 2, the latter possibly also confined to the basal two thirds of the epithelium. In contrast, CIN 1 shows a focal, patchy staining. Reactive epithelium is often completely negative. P16 is also diffusely positive in adenocarcinoma in situ (AIS) and most invasive adenocarcinomas. Ki-67 may be used in combination with p16 but obviously does not add any advantage. L1 protein immunohistochemistry has not been widely accepted for routine diagnosis, so far. 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 26 Histopathologic Evaluation of Cervical Intraepithelial Neoplasia (CIN) in Tissue Biopsy Samples: Dilemmas and Solutions Sigurd Lax Department of Pathology, General Hospital Graz West Graz, Austria Pathology Teaching Hospital of the Medical University Graz Global Incidence of cervical carcinoma (2000) Lax, ZORA Meeting 2012, Brdo Epidemiology of cervical carcinoma • HPV • Socio-economic status • Dramatic decrease of the incidence by cervical cytology screening Lax, ZORA Meeting 2012, Brdo Incidence and Mortality of Cervical Carcinoma in Finland WHO: Tumours of the Female Genital Tract, 2003 Lax, ZORA Meeting 2012, Brdo Incidence and Mortality of Cervical Carcinoma in Austria 1983-2001 Statistik Austria, 2004 2007: <6 2007: 2 Lax, ZORA Meeting 2012, Brdo Cervical carcinoma and precursor Adenocarcinoma Squamous cell carcinoma CIN AIS HPV Lax, ZORA Meeting 2012, Brdo 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 102 Classification of Squamous Precursor Lesions Traditional WHO Bethesda Mild dysplasia CIN1 LSIL Moderate dysplasia Severe dysplasia Carcinoma in situ CIN2 CIN3 HSIL Lax, ZORA Meeting 2012, Brdo Cervical Intraepithelial Neoplasia 1 (CIN 1) Lax, ZORA Meeting 2012, Brdo Cervical Intraepithelial Neoplasia 2 (CIN 2) Lax, ZORA Meeting 2012, Brdo Cervical Intraepithelial Neoplasia 3 (CIN 3) Lax, ZORA Meeting 2012, Brdo HPV DNA demonstrated by in situ Hybridisation Lax, ZORA Meeting 2012, Brdo First exposure to HPV16 and development of high grade CIN Woodman et al., Lancet 2001 Lax, ZORA Meeting 2012, Brdo 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 103 Woodman et al., Nature Rev 2007 Lax, ZORA Meeting 2012, Brdo DNA damage p53 p21 Cyclin / cdk-complexes pRB G1- cell cycle arrest DNA sythesis HPV E7 HPV E6 Apoptosis Telomerase E2F p16 cdk4 Cyclin E Molecular mechanisms of HPV on the host cell Lax, ZORA Meeting 2012, Brdo Basal and stem cells Committed cells Terminally differentiated cells Basal and stem cells Committed cells Transformed terminally differentiated cells Transformed basal/stem and committed cells Lack of terminal differentiation LSIL / CIN 1 HSIL / CIN 2&3 Modified after Crum C et al. Mod Pathol 2000 Cell Transformation in Squamous Intraepithelial Lesions Lax, ZORA Meeting 2012, Brdo 2 lesions from the same patient CIN3 / HSIL CIN2 / HSIL Lax, ZORA Meeting 2012, Brdo Distinguishing Features of L/HSIL Wright et al., in: Blaustein, 6th ed., 2011 Feature LSIL HSIL HPV type Any anogenital High risk Koilocytosis Frequent Rare Ploidy Diploid or polyploid Most aneuploid Abnormal mitosis Absent Frequent Location undifferentiated cells/mitosis Lower third Upper 2 thirds Lax, ZORA Meeting 2012, Brdo LSIL Lax, ZORA Meeting 2012, Brdo 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 104 HSIL Lax, ZORA Meeting 2012, Brdo Problems and Questions • Diagnostic uncertainty due to poor reproducibility, in particular of CIN 2 • Further diagnostic weaknesses: Reactive versus CIN1 Reactive/metaplastic versus CIN3 • Uncertain potential of progression of the various lesions • Biomarkers for improved diagnostic accuracy Lax, ZORA Meeting 2012, Brdo Natural History of Various CIN % Regression % Persistence % Progression CIN1 57 32 11 CIN2 43 35 22 CIN3 32 56 12 Lax, ZORA Meeting 2012, Brdo Mitchell et al., JNCI 1996 Toronto Long Term Follow up of Abnormal Cytology Degree Dysplasia % Progression 2/10 y % Regression 2/10 y Mild 0.6 / 12 44 / 88 Moderate 1.5 / 17 33 / 83 Severe 2.8 /21 Holowaty et al., JNCI 1999 Lax, ZORA Meeting 2012, Brdo Interobserver Agreement for CIN • Varies among different studies • Substantial disagreement due to problems with CIN1 and CIN2 • Improved results by using weighted k-statistics Lax, ZORA Meeting 2012, Brdo The Natural History of CIN2 Castle et al., Gynecol Oncol 2009 • 40% of undiagnosed CIN2 seem to regress within 2 years • CIN2 containing HPV16 seems to progress more likely compared to CIN2 with other HPV types Lax, ZORA Meeting 2012, Brdo 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 105 Reproducibility of CIN between experts Carreon et al., Int J Gynecol Pathol, 2007 Lax, ZORA Meeting 2012, Brdo • Population-based study in Costa Rica • Comparison of local pathologists and 2 experts • Diagnosis correlated with HPV and cytology • CIN3: 81/84%, CIN2: 13/31% agreement • Oncogenic HPV: 94% (CIN3) and 72% (CIN2) • Abnormal cyto: 81% (CIN3) and 61% (CIN2) • CIN3 is better reproducible and can better validated by HPV test and cytology than CIN2 Reproducibility of CIN between experts Carreon et al., Int J Gynecol Pathol, 2007 Lax, ZORA Meeting 2012, Brdo •4 experts; QC slide panel (n=185) •Both inter- and intraobserver variability good among experts if weighted k-values used (0.75-0.86 and 0.74- 0.94, respectively) •Best for non-CIN and CIN3 •Worst for CIN2 •Disagreement more frequent between neighbouring categories Interobserver Agreement for CIN Cai et al., AJSP 2007 Lax, ZORA Meeting 2012, Brdo Interobserver Reproducibility of CIN2 Cai et al., AJSP 2007 Lax, ZORA Meeting 2012, Brdo The diagnostic problem of CIN2 • Mixture of CIN1 (1/3) and undercalled CIN3 (2/3) (Castle et al. 2007) • Overcall of CIN1 more likely?! (Galgano et al., 2008) • Unclear, which CIN2 progress • High prevalence of HPV16 (43%) • 2nd opinion for CIN2 recommended to increase diagnostic accuracy • p16 immunohistochemistry as aid? (Dijkstra et al., 2010) Lax, ZORA Meeting 2012, Brdo HPV16 Genotyping a Benchmark for Cervical Biopsy Interpretation ? Galgano et al. AJCP , 2008 • ALTS population (ASCUS+LSIL); n=5060 • 10 centers+expert panel • Hybrid Capture 2 and HPV genotyping • % of HPV16 positivity correlated with severity of lesion • But significant discrepancy between centers • Agreement between centers and experts weak regarding CIN2 Lax, ZORA Meeting 2012, Brdo 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 106 HPV16 Genotyping a Benchmark for Cervical Biopsy Interpretation ? Galgano et al. AJCP, 2008 Lax, ZORA Meeting 2012, Brdo HPV16 Genotyping a Benchmark for Cervical Biopsy Interpretation ? Galgano et al. AJCP, 2008 Lax, ZORA Meeting 2012, Brdo P16 immunostaining as an alternative to histology review for reliable grading of CIN Dijkstra et al., J Clin Pathol 2010 • Combined use of HE and p16 ImHC significantly improves accuracy of interpreting and grading cervical lesions on biopsies • Accuracy of CIN grading of a single pathologist with p16 adjunct comparable to expert panel • P16 staining of cervical lesions should be implemented in daily routine Lax, ZORA Meeting 2012, Brdo P16 immunostaining as an alternative to histology review for reliable grading of CIN Dijkstra et al., J Clin Pathol 2010 Lax, ZORA Meeting 2012, Brdo P16 immunostaining as an alternative to histology review for reliable grading of CIN Dijkstra et al., J Clin Pathol 2010 Lax, ZORA Meeting 2012, Brdo “Surrogate markers” for CIN diagnosis • P16 • Ki-67 • HPV • L1 Lax, ZORA Meeting 2012, Brdo 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 107 p16INK4 • Overexpression in CIN, AIS and most carcinomas; • Association with HPV: RB Inactivation ?! • Good correlation with SIL/CIN • Detection of dysplastic cells in Pap smears (Klaes et al. Int J Cancer 2001) Lax, ZORA Meeting 2012, Brdo P16 in low and high grade CIN Lax, ZORA Meeting 2012, Brdo p16 and CIN Klaes et al., AJSP 2002, Branca et al., IJGP 2004, Tringler et al. Hum Pathol 2004 • Diffuse positive p16 immunoreactivity only in invasive carcinomas, CIN2/3 and CIN1 associated with high risk HPV • Part of CIN 1-3 negative for p16 • No predictive value for high risk HPV clearance after conisation, no prognostic value for carcinomas • Positivity also in reactive and metaplastic epithelium Lax, ZORA Meeting 2012, Brdo Ki-67 (Mib 1) • Expression during the cell cycle • In normal epithelium expression only in suprabasal and a few basal cells (hormone dependence) • HPV infection leads to activation of the cell cycle • LSIL: positive cells in the superficial third of the epithelium (not found in reactive changes) • HSIL: multple positive cells throughout the epithelium • Assist in the distinction of SIL from reactive changes ! Lax, ZORA Meeting 2012, Brdo Pap III / ASCUS-H Lax, ZORA Meeting 2012, Brdo Atypical Squamous Epithelium Lax, ZORA Meeting 2012, Brdo 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 108 Ki-67 Lax, ZORA Meeting 2012, Brdo p16 Lax, ZORA Meeting 2012, Brdo Reactive or CIN Lax, ZORA Meeting 2012, Brdo Reactive Changes p16 Ki-67 Lax, ZORA Meeting 2012, Brdo Reactive versus neoplastic Lax, ZORA Meeting 2012, Brdo P16 is overexpressed in AIS Lax, ZORA Meeting 2012, Brdo 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 109 Cervical Lesions and P16 Galgano et al., AJSP 2010 • P16 more sensitive than HE histology • For the distinction between CIN and reactive/metaplastic changes reliable • Ki-67 seems to provide no additional information ? • P16 seems to be particularly helpful for CIN2 Lax, ZORA Meeting 2012, Brdo Immunohistochemistry HPV L1 Negri, AJSP 2008, Galgano, AJSP 2010; Hoshikawa, Path Res Pract 2010 • Not widely in use; in combination with p16? • Specific proof of HPV L1 capsid protein • Indicates productive phase • Prognostic value for CIN1? L1 positive: 21-27% progression L1 negative: 80-97% progression • No distinction between CIN1 and reactive Lax, ZORA Meeting 2012, Brdo Combination of L1 and p16 Negri et al. AJSP , 2008 Lax, ZORA Meeting 2012, Brdo • Many L1-/p16- lesions seem to be reactive • Assessment of the progression potential of CIN1? L1, p16 and Progression Lax, ZORA Meeting 2012, Brdo Hoshikawa et al., Path Res Pract 2010 Caveat of high grade CIN High grade CIN in crypts/glands may resemble atypical glandular epithelium CIN3 as glandular mimic 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 110 Adenocarcinoma in situ (AIS) Lax, ZORA Meeting 2012, Brdo Adenocarcinoma in situ (AIS/ACIS) • Normal glandular or surface epithelium replaced by neoplastic epithelium • No invasion • Concomitant CIN in ca. 50% • Atypical Pap Smear only in 50% • Dysplasia not used (poor reproducibility) • CGIN: British terminology Lax, ZORA Meeting 2012, Brdo 2 lesions of the same patient Dysplasia? AIS Lax, ZORA Meeting 2012, Brdo Take Home Message • Inter-/intraobserver agreement for CIN in particular valuable for negative and CIN3 • CIN2 is a problematic lesion • 2nd opinion or p16 adjunct suggested • P16 most important surrogat marker for HPV • Add of Ki-67 not necessary but may be helpful • HPV L1 Protein immunohistochemistry not widely used (informs about CIN1 progression) Lax, ZORA Meeting 2012, Brdo 3. izobraževalni dan programa ZORA - ZORA 2012 ____________________________________ 111