Multi-parametric MRI to guide salvage treatment of recurrent prostate cancer

Dinis Fernandes, C. (2019)

Doctoral Thesis

Prostate cancer (PCa) is frequently treated with radiotherapy. However, depending on the aggressiveness of the disease, the risk of recurrence can be up to 35% within five years of the initial treatment. Patients with localised recurrent PCa are candidates for curative (i.e. salvage) treatment. To overcome the toxicity associated with whole-gland approaches, focal salvage treatments target the index lesion while sparing the surrounding tissue. The studies described in this thesis elaborate on the use of quantitative multi-parametric MRI (mp-MRI) for the detection and localisation of locally recurrent PCa after radiotherapy. Pre-treatment radiomic imaging features were found to have potential to improve recurrence-risk prediction models for high-risk PCa patients treated with radiotherapy. In this thesis, the mp-MRI properties of irradiated benign tissue and recurrent tumour were characterised, with access to pathological samples. These findings can be used as a foundation to establish guidelines (which are currently absent) on how to assess and score MRI scans after radiotherapy. Improving radiological knowledge in the recurrent setting can lead to improved staging and result in better patient selection for salvage treatments. Lastly, this thesis provides evidence on how best to define the region to target, leading to a refinement of focal salvage strategies.

• Recurrent prostate cancer
• Multi-parametric MRI
• Radiotherapy
• Histopathology
• Machine learning
• Radiomics

• All authors: Dinis Fernandes, C.
• Supervisor: Heide, U.A. van der
• Co-supervisor: Smolic, M.
• Committee: Marijnen, C.A.M.; Pelger, R.C.M.; Haustermans, K.; Leeuwen, A.G.J.M. van
• Qualification: Doctor (dr.)
• Awarding Institution: Faculty of Medicine, Leiden University Medical Center (LUMC), Leiden University
• Date: 2019-09-25
• ISBN (print): 9789463237567
• Sponsorship: KWF Kankerbestrijding