Marieke van Son

148 CHAPTER 8 Nomogram The static nomograms for model 1 and 2 are depicted in Figure 2 and 3, respectively. An exemplary case is included in the figure caption. The Kaplan-Meier curves for bDFS for low-, intermediate-, and high-risk groups, as identified by model 1 (nomogram score <193, 193-222, and >222, respectively) and model 2 (nomogram score <297, 297-334, and >334, respectively) are shown in Figure 4. Estimated bDFS at 24 months for low-, intermediate, and high-risk groups was 84%, 70%, and 31% for model 1 (p<0.0001) and 100%, 71%, and 5% for model 2 (p<0.0001), respectively. Both models can be used as webtools through: https://fs-hdr-bt-prediction.shinyapps.io/model1/ (model 1) or https://fs-hdr-bt-prediction.shinyapps.io/model2/ (model 2). Figure 2 – Nomogram based on model 1 for prediction of biochemical failure among patients who underwent FS-HDR-BT. Probabilities of biochemical failure within 12, 24, and 36 months can be calculated. Instruction: Locate the patient’s GTV (cm3) of the recurrent prostate cancer lesion on the ‘GTV (cm3)’ axis. Draw a line straight upward to the ‘Points’ axis to determine the number of points based on the GTV. Repeat this process for each of the four variables. Sum the points that are received for each of the four predictors (‘Total points’). Finally, draw a line straight down from the ‘Total points’ axis to find the patient’s probability of having biochemical failure within 36, 24, and 12 months, respectively. An interactive version of the nomogram can be used online through: https://fs-hdr-bt-prediction.shinyapps.io/model1/. As an example, a 72-year-old patient with a GTV of 4.0 cm3, a PSA-level of 6.0 ng/ml, and a pre-salvage PSADT of 25 months has an estimated 12-, 24-, and 36-months bDFS probability of 95% (95% CI: 93-98%), 78% (95% CI: 70-87%) and 53% (95% CI: 40-71%), respectively.