Prediction of pathological stage at radical prostatectomy: do commonly used prostate cancer nomograms apply to men from Far North Queensland?

We aimed to assess whether common prostate cancer nomograms are valid for prostate cancer patients in Far North Queensland. Partin's tables accurately predicted organ‐confined disease, extra‐prostatic extension and seminal vesicle invasion, whilst more patients in our cohort had organ‐confined disease than was predicted by the MSKCC (Memorial Sloan Kettering Cancer Centre) tool. Both tests performed well on logistic regression modelling.

Background

Clinical nomograms are routinely used by urologists to predict pathological and clinical outcomes. Commonly used prostate cancer nomograms include Partin's tables and Memorial Sloan Kettering Cancer Centre (MSKCC) nomograms which were developed in high‐volume centres in the United States. We aimed to assess whether these tools are valid for prostate cancer patients in Far North Queensland.

Methods

All patients undergoing radical prostatectomy in Cairns between August 2014 and September 2017 were identified. Preoperative data were entered into the online nomogram tools. The predicted probability of organ‐confined (OC) disease, extra‐prostatic extension (EPE) and seminal vesical invasion was compared to the observed outcomes.

Results

Preoperative clinical information was available for 290 patients. Partin's tables accurately estimated OC disease, EPE and seminal vesical invasion with the observed outcome plot overlying the ideal correlation curve. More patients in our cohort had OC disease than was predicted by the MSKCC nomogram; fewer patients had EPE that was predicted by the MSKCC nomogram. On logistic regression modelling, the area under the curve for MSKCC and Partin's were 0.751 and 0.706, respectively, suggesting both tests have good performance in predicting final pathological outcome for our population of patients with no statistical difference between the two nomograms (P = 0.29).

Conclusion

The MSKCC preoperative nomogram and Partin's tables were both able to accurately predict pathological outcomes from preoperative clinical information in men from Far North Queensland, despite likely differences in population genetics and environmental exposures.

https://ift.tt/2EwFdjJ