Modeling Study to Predict Progression of Anal Cancer Pre-cursor Lesions in HIV

Persons living with HIV-infection are at greatly increased risk for anal carcinoma and anal intraepithelial neoplasia (AIN). Despite the overall improvement in HIV outcomes, the incidence of anal carcinoma has not decreased with the advent of highly active antiretroviral ther-apy. Further, there is substantial morbidity and mortality associated with anal carcinoma in HIV-infected individuals. For these reasons, it is critically important to develop effective screening and treatment strategies in this population. Anal carcinoma is similar to cervical carcinoma in that they are both associated with infection with human papillomavirus (HPV) and share similar cytologic features of dysplasia. Anal cytology screening is ultimately expected to reduce the incidence of anal carcinoma similar to that seen with cervical cancer after the introduction of cervical Pap screening. Given the high frequency of abnormal cytology in patients with HIV and the need to confirm results by high-resolution anoscopy, however, the implementation of screening programs requires a substantial commitment of clinical resources. The workload and costs involved in following up on abnormal cytology reduce the cost-effectiveness of screening and pose a significant barrier to its widespread integration into routine HIV care. A model for predicting which patients are at greatest risk for progressive of anal dysplasia is needed in order to decrease the need for excessive confirmatory procedures in this population. Without such a model, anal carcinoma screening may remain cost prohibitive for many HIV clinics.

The objective of this study is to develop a predictive model to identify patients who are at greatest risk for progression of anal intraepithelial neoplastic changes. The central hypotheses are: 1) that progression of early HPV-related anal dysplasia is associated with environmental, virological, and host molecular factors and 2) that it is possible to develop a predictive statistical model with a high sensitivity and specificity for predicting disease progression. These hypotheses have been formulated on the basis of strong evidence from studies of HPV-related cervical dysplasia that smoking, HPV E2 expression, HPV E6/E7 protein expression, high-risk HPV type, HPV viral load, p16 expression, p53 expression and Ki67 expression are associated with progressive cervical dysplasia. The rationale for the proposed research is that development of a predictive model will allow clinicians to design more cost-effective screening and follow-up strategies which focus resources on intensively testing only those patients with a significant risk for progression. Further, the models developed will allow clinicians to identify a population of patients who may benefit from early treatment interventions. Finally, information learned from this research may provide information applicable to other HPV-related cancers.