Joint models for dynamic prediction in localised prostate cancer: a literature review.

BACKGROUND: Prostate cancer is a very prevalent disease in men. Patients are monitored regularly during and after treatment with repeated assessment of prostate-specific antigen (PSA) levels. Prognosis of localised prostate cancer is generally good after treatment, and the risk of having a recurrence is usually estimated based on factors measured at diagnosis. Incorporating PSA measurements over time in a dynamic prediction joint model enables updates of patients' risk as new information becomes available. We review joint model strategies that have been applied to model time-dependent PSA trajectories to predict time-to-event outcomes in localised prostate cancer. METHODS: We identify articles that developed joint models for prediction of localised prostate cancer recurrence over the last two decades. We report, compare, and summarise the methodological approaches and applications that use joint modelling accounting for two processes: the longitudinal model (PSA), and the time-to-event process (clinical failure). The methods explored differ in how they specify the association between these two processes. RESULTS: Twelve relevant articles were identified. A range of methodological frameworks were found, and we describe in detail shared-parameter joint models (9 of 12, 75%) and joint latent class models (3 of 12, 25%). Within each framework, these articles presented model development, estimation of dynamic predictions and model validations. CONCLUSIONS: Each framework has its unique principles with corresponding advantages and differing interpretations. Regardless of the framework used, dynamic prediction models enable real-time prediction of individual patient prognosis. They utilise all available longitudinal information, in addition to baseline prognostic risk factors, and are superior to traditional baseline-only prediction models.

A Personalized Clinical Dynamic Prediction Model to Characterize Prognosis for Patients With Localized Prostate Cancer: Analysis of the CHHiP Phase 3 Trial.

PURPOSE: The CHHiP trial assessed moderately hypofractionated radiation therapy in localized prostate cancer. We utilized longitudinal prostate-specific antigen (PSA) measurements collected over time to evaluate and characterize patient prognosis. METHODS AND MATERIALS: We developed a clinical dynamic prediction joint model to predict the risk of biochemical or clinical recurrence. Modeling included repeated PSA values and adjusted for baseline prognostic risk factors of age, tumor characteristics, and treatment received. We included 3071 trial participants for model development using a mixed-effect submodel for the longitudinal PSAs and a time-to-event hazard submodel for predicting recurrence of prostate cancer. We evaluated how baseline prognostic factor subgroups affected the nonlinear PSA levels over time and quantified the association of PSA on time to recurrence. We assessed bootstrapped optimism-adjusted predictive performance on calibration and discrimination. Additionally, we performed comparative dynamic predictions on patients with contrasting prognostic factors and investigated PSA thresholds over landmark times to correlate with prognosis. RESULTS: Patients who developed recurrence had generally higher baseline and overall PSA values during follow-up and had an exponentially rising PSA in the 2 years before recurrence. Additionally, most baseline risk factors were significant in the mixed-effect and relative-risk submodels. PSA value and rate of change were predictive of recurrence. Predictive performance of the model was good across different prediction times over an 8-year period, with an overall mean area under the curve of 0.70, mean Brier score of 0.10, and mean integrated calibration index of 0.048; these were further improved for predictions after 5 years of accrued longitudinal posttreatment PSA assessments. PSA thresholds <0.23 ng/mL after 3 years were indicative of a minimal risk of recurrence by 8 years. CONCLUSIONS: We successfully developed a joint statistical model to predict prostate cancer recurrence, evaluating prognostic factors and longitudinal PSA. We showed dynamically updated PSA information can improve prognostication, which can be used to guide follow-up and treatment management options.

Elucidating Prostate Cancer Behaviour During Treatment via Low-pass Whole-genome Sequencing of Circulating Tumour DNA.

BACKGROUND: Better blood tests to elucidate the behaviour of metastatic castration-resistant prostate cancer (mCRPC) are urgently needed to drive therapeutic decisions. Plasma cell-free DNA (cfDNA) comprises normal and circulating tumour DNA (ctDNA). Low-pass whole-genome sequencing (lpWGS) of ctDNA can provide information on mCRPC behaviour. OBJECTIVE: To validate and clinically qualify plasma lpWGS for mCRPC. DESIGN, SETTING, AND PARTICIPANTS: Plasma lpWGS data were obtained for mCRPC patients consenting to optional substudies of two prospective phase 3 trials (FIRSTANA and PROSELICA). In FIRSTANA, chemotherapy-naïve patients were randomised to treatment with docetaxel (75 mg/m2) or cabazitaxel (20 or 25 mg/m2). In PROSELICA, patients previously treated with docetaxel were randomised to 20 or 25 mg/m2 cabazitaxel. lpWGS data were generated from 540 samples from 188 mCRPC patients acquired at four different time points (screening, cycle 1, cycle 4, and end of study). OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: lpWGS data for ctDNA were evaluated for prognostic, response, and tumour genomic measures. Associations with response and survival data were determined for tumour fraction. Genomic biomarkers including large-scale transition (LST) scores were explored in the context of prior treatments. RESULTS AND LIMITATIONS: Plasma tumour fraction was prognostic for overall survival in univariable and stratified multivariable analyses (hazard ratio 1.75, 95% confidence interval 1.08-2.85; p = 0.024) and offered added value compared to existing biomarkers (C index 0.722 vs 0.709; p = 0.021). Longitudinal changes were associated with drug response. PROSELICA samples were enriched for LSTs (p = 0.029) indicating genomic instability, and this enrichment was associated with prior abiraterone and enzalutamide treatment but not taxane or radiation therapy. Higher LSTs were correlated with losses of RB1/RNASEH2B, independent of BRCA2 loss. CONCLUSIONS: Plasma lpWGS of ctDNA describes CRPC behaviour, providing prognostic and response data of clinical relevance. The added prognostic value of the ctDNA fraction over established biomarkers should be studied further. PATIENT SUMMARY: We studied tumour DNA in blood samples from patients with prostate cancer. We found that levels of tumour DNA in blood were indicative of disease prognosis, and that changes after treatment could be detected. We also observed a "genetic scar" in the results that was associated with certain previous treatments. This test allows an assessment of tumour activity that can complement existing tests, offer insights into drug response, and detect clinically relevant genetic changes.