Expression Of P16, P53and IMP3and Their Clinical Significance In Pre-Malignant And Malignant Lesions Of Uterine Cervix

Objective:1. To assess the diagnostic value of P16, P53and IMP3in cervical carcinogenesis by their immunohistochemical expression and their correlation with clinicopathological parameters.2. To determine if P16or P53in combination with IMP3can increase diagnostic accuracy for detecting cervical precancer and cancer.3. To know the relation between HPV infection and IMP3expressionMethods:A total of120cases among which32CIN I,35CIN Ⅱ-Ⅲ and53SCC, were selected retrospectively with additionally,20benign cases obtained for the control group for the study. Among SCC29were FIGO stage I,12were for each stage Ⅱ and Ⅲ and unfortunately there was no stage IV cases. Formalin fixed paraffin embedded cervical tissue blocks were obtained for each cases and were analyzed by immunohistochemistry to study the expression of p16, p53and IMP3. The tissue samples were obtained from the specimen of hysterectomy, cervical biopsy, cone excision and LEEP. The clinical parameter including age, histologic diagnosis, FIGO staging, LN status, vessel invasion and HC2result were obtained by retrospective chart review of the selected cases. The age of the patients ranged from24to63years old. The mean age was42.7±7.964. Among SCC patients the oldest age was61years and the youngest age was27years with the mean age44.6±7.8. For CIN Ⅱ-Ⅲ the oldest age was51years and the youngest age was25years with the mean age39.7±6.8while for CIN I the oldest age was63years and the youngest was24years with the mean age of40.1±9.2. Among normal cases the youngest age was41and the oldest was56years with the mean age of47.2±3.443. Statistical analysis was performed using the SPSS for Windows Software, version19.0. All statistical calculations were set to95%confidence intervals and significance levels of5%i.e. a P value of less than0.05was considered significant. Immunostaining was scored semi quantitatively. Pearson X2was used to determine the statistical significance of difference in expression of p16, p53and IMP3among the study groups. Pearson correlation coefficient was used to know the associations between the overexpression of p16, p53and IMP3with various clinicopathological parameters as well as to know the relation between markers. Based on the co-expression of the markers, sensitivity and specificity was calculated and ROC curve was plotted for coexpression of IMP3and p16.Result.1. The observed differences between the frequency of p16, p53and IMP3overexpression among the study groups was found to be statistically significant (P=0.003).2. P16staining was detected in both cytoplasm and nucleus. None of the control case showed P16overexpression. The difference in frequency of P16expression among the study groups was statistically significant (P=0.024). There was statistically significant difference in frequency of p16expression among FIGO (P=0.001). P16overexpression was positively correlated with the study groups (p=0.211, P=0.021).3. P53staining was found to be exclusively nuclear. None of the control case was positive for P53overexpression. The difference in the overexpression of p53among the study groups was found to be statistically significant (P-0.007). Within SCC, the differences in the frequency of p53expression among FIGO stages was statistically significant(P=0.012).Overexpression of P53was positively correlated with the study groups and FIGO stage (p=0.275, P=0.002and p=0.358, P=0.008, respectively).4. IMP3staining was universally cytoplasmic, with diffuse staining of strong intensity in all SCC, whereas staining was typically of moderate or weak intensity in CIN. None of the control case showed IMP3expression. The differences in the frequency of IMP3overexpression was observed to be proportional with the severity of cervical lesion and the difference was statistically significant (P0.000). Within SCC the differences in the frequency of IMP3overexpression among FIGO stage was statistically significant (P=0.030). Overexpression of IMPS was positively correlated with FIGO stage (p=0.292, P=0.036), vascular invasion (p=0.328, P=0.017), lymph node positivity (p=0.280, P=0.042) and HR-HPV status (p=0.175, P=0.219).5. The relation between p16and IMP3expression among the study groups was positively correlated and the correlation was statistically significant (p=0.339, P=0.000)6. The sensitivity and specificity of p16as an independent marker to detect cervical precancerous and cancerous lesion was90%and50%respectively. The PPV was75.2%, NPV was74.3%positive likelihood ratio was1.8and negative likelihood ratio was0.2.7. The sensitivity and specificity of IMP3as an independent marker to detect cervical precancerous and cancerous lesion was68.2%and92.3%respectively. The PPV was93.7%, NPV was63.1%positive likelihood ratio was8.5and negative likelihood ratio was0.35.8. The sensitivity and specificity of the co-expression of IMP3and p16to detect the cervical precancerous and cancerous lesion was87%and87.5%respectively. Likewise, the PPV was93.7%, NPV was75.7%, positive likelihood ratio was6.69and negative likelihood ratio was0.15. The AUC in ROC curve was0.969.Conclusion:1. P16was found to be a robust biomarker of cervical carcinoma. The expression of p16did not show much difference between CIN lesions.2. Increased proportion of p53expression was seen with increase in histologic severity of cervical lesions. Presence of immunoreactive p53only in a small proportion of cells in most cases support that in cervical cancer, p53is mostly wild type having short half-life and hence difficult to detect immunohistochemically. P53immunoreactivity can be helpful to decide a neoplastic lesion, but the absence of P53does not exclude neoplasia3. IMP3over expression was predominantly found in SCC and was positively correlated with FIGO stage, lymph node positivity and vascular invasion which are the indicators of tumor aggressiveness. Thus, IMP3can be presumed to be a marker of enhanced tumor aggressiveness.4. Positive correlation was observed between IMP3expression and HR-HPV status.5. P16as an independent marker to detect the precancerous and cancerous lesions had high sensitivity but low specificity while IMP3as an independent marker had high specificity but low sensitivity. Both sensitivity and specificity increases with co-expression of IMP3and p16. Likewise, both PPV and NPV are higher with co-expression of IMP3and p16compared to their independent expression.6. On the basis of obtained AUC of0.969in the ROC curve for the co-expression of IMP3and p16it can be said that the diagnostic accuracy to detect cervical precancerous and cancerous lesion is excellent with co-expression of IMP3and p16.