Development and evaluation of prostate cancer risk prediction models for use in the community

  • Mohammad Aladwani

Student thesis: Phd


Prostate cancer is one of the most common cancers in men, and the incidence is increasing around the world. Unlike breast cancer in women, there are no effective early detection programs such as screening. This is partly due to lack of an adequate biomarker with ability to detect clinically significant prostate cancer and to be specific to it. The Prostate-specific antigen test has been used in addition to the digital rectal examination (DRE) to determine prostate cancer risk. These tests can produce false positive or false negative results. This challenging issue of inaccuracy in detecting prostate cancer can be improved by using a risk prediction model. Many researchers have tried to develop a predictive model to improve the performance of prostate cancer detection by combining several factors and tests that are related to prostate cancer. However, the majority of the existing risk prediction models are not suitable to be implemented in primary care settings either due to incorporating inappropriate invasive tests or issues relating to the study design and methodology at the development stage. Therefore, there is a need to develop a risk prediction model for prostate cancer that consists of readily available, easy to measure, and low-cost so that it can be implemented in primary care and community settings. In this thesis, I investigated the existing risk models for prostate cancer that can be implemented in primary care by conducting a systematic review. The findings suggested that there is a paucity of such models. I also reviewed the literature on prostate cancer risk factors. To date, there is some emerging evidence that suggests causal relationships with the disease such as physical activity and some genetic factors, in particular single nucleotide polymorphism. This new knowledge can help advance cancer prevention. I examined the relationship between body size and body shape and the risk of prostate cancer. The findings suggest an "apple" body shape indicative of central obesity was protective against prostate cancer. The changes in body size did not show any association. Next, I developed a risk prediction model that consists of age, prostate-specific antigen (PSA), and free-to-total PSA ratio (%fPSA) using multivariate logistic regression analysis. I compared the results in two different outcomes. The first outcome was prostate cancer detected by multiparametric reasoning imaging targeting biopsy, and the second outcome was by the conventional biopsy needle. The results demonstrate the usefulness of our model in detecting prostate cancer which outperformed PSA alone in detecting prostate cancer. Also, it shows that by using the model, fewer clinically significant cases are missed. Furthermore, my results also showed that there is a demand for testing by men to increase their awareness and knowledge of their risk of having the disease so they can treat it as early as possible. Lastly, I explored men's perspectives on home testing for prostate cancer using a PSA kit. The findings suggested men were happy with the home testing kit. In sum, prostate cancer incidence is increasing worldwide. A good risk prediction model offers a way forward to aid the early detection of prostate cancer as its performance is better than the PSA test alone. Risk prediction model can be applied to detect prostate cancer either with conventional needle biopsy or MRI-guided biopsy. PSA home testing kit is considered as a future proof to enhance the number of men taking the test, in particular, men with ethnic minority and harder to reach groups.
Date of Award31 Dec 2022
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorKenneth Muir (Supervisor) & Artitaya Lophatananon (Supervisor)

Cite this