Partin Nomograms

INDIVIDUAL PREDICTION RISK

Given three input values (PSA, clinical stage, and Gleason score), we return estimated probability (likelihood) as a percentage value of being in each of the four pathological stages (Organ confined, Extraprostatic extension, Seminal Vesicle Invasion, Lymph Node Invasion). The software program below generates % individual risk and in addition, the 95% confidence intervals are also returned.  Figure 3 from the BJUI article shows the graphs from which the software program was generated for N=5730 prostate cancer cases.



Key References:
  1. Huang, Y., Isharwal, S., Haese A., chun, F. K. H., Makarov, D. V., Feng, Z., Han, M., Humphreys, E., Epstein, J. I., Partin, A. W., and Veltri, R. (2010) Prediction of patient-specific risk and percentile cohort risk of pathological stageoutcome using continuous prostate-specific antigen measurement, clinical stage, and biopsy Gleason score. BJUI 107, 1562-1569.

  2. Makarov DV, Trock BJ, Humphreys EB et al. Updated nomogram to predict pathologic stage of prostate cancer given prostate-specific antigen level, clinical stage, and biopsy Gleason score (Partin tables) based on cases from 2000 to 2005. Urology 2007; 69: 1095-101

Acknowledgements:
The statistical R-program was prepared by Dr. Ying Huang, Fred Hutch Cancer Research Center (FHCRC). Greta Stoianovici (JHH Brady Urology Webmaster) transformed the R-program into the final formatted user- friendly nomogram. Dr. Robert Veltri organized the research study and assisted in the final design of the Partin Nomogram.

RISK DISTRIBUTION ESTIMATION

For a given pathological stage (Organ confined, Extraprostatic extension, Seminal Vesicle Invasion, Lymph Node Invasion), we estimated the distribution of the risk (probability of being in this pathological stage) for patients in the entire JHH cohort (N=5730).

Figure 4 -y coordinate is p , x coordinate is v
               (0,1)=(0%,100%))
from the BJUI paper shows the risk quantile curve for each pathological stage and their 95% CI.

First, given two input values (pathological stage, and a value p in (0, 100%)), the software program below returns the estimated proportion of subjects in the JHH cohort with the probability (likelihood) of being in this pathological stage less than p, together with its 95% confidence interval. For example, if we enter p as the probability of being Organ Confined for an individual patient deter mined from the left side Table, then the output will show the proportion of subjects in the JHH cohort with smaller probability of being Organ Confined compared to this patient, in other words the placement of the patient's risk of being Organ Confined relative to others in the JHU population.

Second, given two input values (pathological stage, and a value v in (0, 100%)), the software program below returns the vth percentile of the probability of being in this pathological stage in the entire JHH cohort and its 95% confidence interval.  For example, if one select Organ Confined and a low v1=5% to get one output, and then select Organ Confined and a high v2=95% to get another output.  The two outputs will provide the 5th to 95th percentiles of the probability of being Organ Confined in JHU population.

 

© The Johns Hopkins University, The Johns Hopkins Hospital, and Johns Hopkins Health System. All rights reserved. Disclaimer
Email: webmaster@urology.jhu.edu | 600 North Wolfe Street, Baltimore, Maryland 21287

urology second opinion urology second opinion