Prevention and Screening for Cancer in people with HIV – who and when?

Despite an increased risk of cancer in people with HIV, the only current HIV specific cancer-screening guideline is for cervical-cancer-screening, to which our research-group has earlier contributed.

Clearly, there is a need for HIV specific screening guidelines for other cancers; breast, lung, anal, and colon cancer, where an increased risk, earlier debut, or poorer prognosis among people with HIV has been identified.

The aim of this project is to firstly develop predictive models for cancer risk in people with HIV, using unsupervised machine learning techniques on datasets based on demographic, socioeconomic, HIV-related, and comorbidity factors.

Secondly, we aim to evaluate these prediction models through counterfactual prediction to justify the development of  a clinical user-friendly tool to help guide individualized cancer prevention and screening for the population with HIV.

To our knowledge this is the first project using machine-learning to create cancer-prediction models in an HIV-population.

In summary, our aim is to create a clinical tool "HIVigilant: Cancer Screening Guide”, for individualized cancer prevention and screening of the population with HIV. This clinical tool can be integrated in the HIV register for easy access and use for all HIV clinicians in Sweden. This will save lives and reduce the burden of cancer among people with HIV.

The knowledge that this project contributes can also be useful in the work of cancer prevention in the general population.