Development and validation of a dynamic prognostic tool for prostate cancer recurrence using repeated measures of posttreatment PSA: a joint modeling approach

Abstract

Prostate-specific antigen (PSA) is a biomarker routinely and repeatedly measured on prostate cancer patients treated by radiation therapy (RT). It was shown recently that its whole pattern over time rather than just its current level was strongly associated with prostate cancer recurrence. To more accurately guide clinical decision making, monitoring of PSA after RT would be aided by dynamic powerful prognostic tools that incorporate the complete posttreatment PSA evolution. In this work, we propose a dynamic prognostic tool derived from a joint latent class model and provide a measure of variability obtained from the parameters asymptotic distribution. To validate this prognostic tool, we consider predictive accuracy measures and provide an empirical estimate of their variability. We also show how to use them in the longitudinal context to compare the dynamic prognostic tool we developed with a proportional hazard model including either baseline covariates or baseline covariates and the expected level of PSA at the time of prediction in a landmark model. Using data from 3 large cohorts of patients treated after the diagnosis of prostate cancer, we show that the dynamic prognostic tool based on the joint model reduces the error of prediction and offers a powerful tool for individual prediction.

Fig. 1.

(A) Predicted mean evolution and (B) survival function in the 5 latent classes of the selected JLCM on WBH data (N = 1268). Predictions given for a subject with T-stage = 2, Gleason = 7, and iPSA = 10 n/mL.

Fig. 2.

Individual prediction of prostate cancer recurrence for 2 patients from UM. On the left, the patient experienced a recurrence 3.8 years after RT. Updated individual predictions are given every 6 months from 1 to 3.5 years. On the right, the patient did not experience any recurrence within the first 6 years after RT. Updated individual predictions are given every year from 1 to 6 years after RT. The x are the PSA measures used for the prediction, the vertical solid line is the time s of prediction, and the vertical dashed line is the time of recurrence.

Fig. 3.

Weighted average absolute error of prediction (WAEP) over 3 years of forecast and EP at a forecast of, respectively, 1 year and 3 years using the absolute loss function for UM cohort (on the left) and RTOG cohort (on the right).

Fig. 4.

Absolute EP for UM cohort (on the left) and RTOG cohort (on the right) based on information at s = 1,2,3 and for a forecast up to 3 years in the future.4