These 11 risk factors contained 8 traditional risk factors and 2 novel ones — NLR and time since diagnosis. 

The authors noted that the “model did not include 12 variables due to lack of statistical significance or model improvement, including type of progression, number of prior secondary hormonal therapies, PSA doubling time, age, baseline corticosteroid use, or prior prostatectomy.”  

For each of the 11 variables significantly associated with OS, the hazard ratio (HR) and 95% confidence interval was computed and a risk score was developed for each using regression coefficients. These risk scores were placed on a nomogram, which was used for prognostication.

Based on the risk scores from the training set, the test set was divided into either low-risk and high-risk groups based on the median cut-off points. In this model, the 11-variable model provided significant separation between low- and high-risk prognostic groups (HR: 0.35; 95% CI: 0.27-0.46).

Related Articles

Using the model, the effect of enzalutamide treatment was determined for low-and high-risk prognostic groups. For the low-risk group, median OS was not reached with both enzalutamide and placebo, but for the high-risk group, median OS was 27.4 and 24.9 months, respectively.

The researchers then determined that it was possible to separate patients into 3 risk categories based on tertiles of risk factors. Patients with more than 6 risk factors were placed into the high-risk tertile, those with 4 to 6 risk factors were placed into the intermediate-risk tertile, and those with 0-3 risk factors were in the low-risk tertile. This model provided significant separation of the OS curves for the low-risk group (HR: 0.20; 95% CI: 0.14-0.29) and intermediate-risk group (HR: 0.40; 95% CI: 0.30-0.53) compared with the high-risk group.

The model also provided discrimination for key secondary endpoints, such as progression-free survival, PSA decline, and radiographic response.