Page 45 - Strategies for non-invasive managementof high-grade cervical intraepithelial neoplasia - prognostic biomarkers and immunotherapy Margot Maria Koeneman

the performance of p16 as a prognostic biomarker may be explained by the fact that it is, in fact, a surrogate marker of cell cycle processes in the development of CIN.
Lu et al. studied insulin-like growth factor II m-RNA-binding protein 3 (IMP3) as a prognostic 2 biomarker in high-grade CIN. IMP3 is normally expressed in embryogenesis and is also an oncoprotein that is expressed in malignant tumours. Lu et al. followed 710 patients with histologically confirmed CIN 2/3 during a mean follow-up of 44 days before surgical intervention.
(83) 140 of these patients were IMP3 positive and 38 patients were diagnosed with invasive carcinomaafterLEEP.AllinvasivecarcinomaswereIMP3positiveandnoIMP3-negativepatient was diagnosed with invasive carcinoma. 27% of IMP3 positive patients with CIN 2/3 developed invasive carcinoma within the time frame of the study. The possibility of the carcinoma already being existent at the time of the initial biopsies (biopsy error) cannot be excluded. The role of IMP3 in CIN was not established before and requires additional research that should precede further evaluation of IMP3 as a prognostic biomarker.
Recently, Uleberg et al demonstrated that zinc finger protein 441 and phospholipase D6 can predict regression of CIN 2-3.(84) These two proteins were identified in a learning set of 20 cases (10 regression and 10 persistence cases) and validated in an independent test set of another 20 high-grade CIN lesions (10 regression and 10 persistence cases). All 20 cases were correctly classified. Zinc finger proteins are transcription regulators functioning as tumour suppressors. Phospholipase proteins are involved in several cellular processes, including cell proliferation and may function by inducing epigenetic changes. The role of both biomarkers in CIN has not been established before. In this study, supernatants of fresh biopsy samples were used to extract water- soluble proteins, which are otherwise degraded in formalin-fixed, paraffin-embedded biopsies. Application of these promising biomarkers therefore requires additional fresh biopsy samples, making clinical applicability less likely.
Composite prognostic biomarker profiles
Baak et al. and Kruse et al. have developed prognostic biomarker profiles, by which patients can be categorized according to probability of lesion regression. The authors followed 90 patients with CIN 1 (n=25) or CIN 2 (n=65).(18, 76) Progression occurred in 17% and was effectively predicted by Ki67 testing. Furthermore, CIN progressors showed decreased Rb, CK 13, CK 14 and involucrin, and increased p21 and p27 expression. Ki67 SI90 and Rb in the deeper half of the epithelium were the strongest multivariate independent predictors of progression. Further evaluation showed that combined quantification of Ki67, Rb, CK 13 and CK 14 gives accurate information about the progression risk of CIN1/2. They developed a composite profile of these four markers, which categorizes patients into several groups with probabilities of lesion regression varying from <1% to 100% (figure 5). Munk et al. developed a predictive biomarker profile for high-grade CIN. They evaluated 162 CIN 2/3 patients over a mean of 16 weeks.(6) The overall regression rate was 21%. The percentage of pRb in the lower epithelium, lesions size, the amount of CD4+ cells in the stroma and consistent condom use were independently associated with lesion regression. A model consisting of these four markers categorized patients into a group with a low regression probability of 8% and a group with a high regression probability of 56%.
A review of prognostic biomarkers
43