Skip to main content

Table 4 Prediction of CS-PCa (as determined by GS) by individual markers, clinical variables, and ProCUrE

From: A urine-based DNA methylation assay, ProCUrE, to identify clinically significant prostate cancer

A
Univariable 1st quartile 3rd quartile Difference OR 95% CI. p value
 ProCUrE − 2.0006 − 1.0828 0.91789 1.58*** 1.28 1.96 < 0.0001
 APC 0 0.37652 0.37652 1.24*** 1.10 1.39 0.0003
 HOXD3 0.43965 5.2043 4.7646 1.54*** 1.24 1.90 < 0.0001
 GSTP1 0 0.1216 0.1216 1.06** 1.02 1.10 0.0013
 KLK10 0 0.12665 0.12665 1.07** 1.02 1.11 0.0048
 TGFβ2 0 0.309 0.309 1.03* 1.00 1.06 0.0481
 TBX15 0 0.21295 0.21295 1.16*** 1.08 1.24 < 0.0001
 PSA 4.565 10.48 5.915 1.98*** 1.46 2.68 < 0.0001
 Age 59 71 12 1.66** 1.13 2.45 0.0105
 PSA density 0.09 0.2 0.11 2.35*** 1.51 3.68 0.0002
 Prostate volume 40 70 30 0.70 0.47 1.06 0.0926
B
Multivariable OR 95% CI. p value
 ProCUrE 1.358* 1.051 1.754 0.0194
 PSA 0.816 0.373 1.785 0.6108
 Age 2.718** 1.295 5.707 0.0082
 PSA density 2.878** 1.455 5.694 0.0024
  1. Using univariable and multivariable logistic regression, the ability of individual methylation markers, ProCUrE, and clinical variables to differentiate CI-PCa and CS-PCa as determined by GS was assessed in the training cohort. Since the scale of each variable is different, interquartile range odds ratios were estimated (logistic regression model *p < 0.05, **p < 0.01, ***p < 0.001)