Two new papers have been published describing research relating to the prognostic powers of nomograms in patients treated for localized prostate cancer. Brockman et al. have developed a new nomogram for the prediction of prostate-cancer-specific mortality (PSCM) rates in men with biochemical recurrence following radical prostatectomy. Lee et al. have found that, following treatment using surgery or radiotherapy, similar risks of biochemical recurrence—determined by treatment-specific nomograms—are associated with different rates of PCSM.

Definitive treatment for localized prostate cancer does not always lead to permanent eradication of the disease. Approximately a quarter of men treated using radical prostatectomy experience biochemical recurrence, which is the reappearance in serum of above-threshold levels of PSA. However, even following biochemical recurrence, the natural history of prostate cancer varies, leading to a 15-year PCSM of 32–45%.

Credit: Bruce Rolff/Hemera/Thinkstock

As a result of the surgical removal of all prostate tissue, serum PSA is undetectable shortly after radical prostatectomy, and biochemical recurrence can be strictly defined. In the study by Brockman and colleagues, the definition included measurement of three successive PSA rises of >0.1 ng/ml, with a final PSA >0.2 ng/ml, or a PSA level ≥0.4 ng/ml ≥6 weeks postoperation, or PSA >0.1 ng/ml leading to administration of secondary therapy. With these definitions, 2,254 consecutive patients experiencing biochemical recurrence after radical prostatectomy without adjuvant therapy were identified at five US academic medical centres between 1987 and 2011. Data from these patients were used to develop a nomogram incorporating nine clinical variables, to predict 5-year, 10-year, and 15-year PCSM. Among these variables, preoperative PSA, pathological Gleason score, extraprostatic extension, seminal vesicle invasion, time to biochemical recurrence, and PSA level at recurrence were all significantly associated with PCSM. The nomogram had a concordance index of 0.774.

Brockman and colleagues also developed a nomogram that included PSA doubling time (PSADT). However, inclusion of PSADT did not improve predictive accuracy. The information provided by PSADT seems largely redundant when appropriate standard clinical variables are considered. Furthermore, lengthy observation can be necessary to calculate PSADT, whereas the nomogram based on standard variables can be used to guide treatment decisions at the time of biochemical recurrence.

Although radical prostatectomy removes all PSA-producing tissue, both brachytherapy and external-beam radiation therapy (EBRT) are targeted to prostate tumours. Use of these radiotherapy techniques to treat localized disease results in complex and variable PSA responses. Rather than applying a strict PSA threshold, determination of recurrence following radiotherapy relies on evaluation of the PSA nadir, which in some cases is only reached years after treatment. The standard definition of biochemical recurrence after EBRT is a PSA measurement ≥2 ng/ml above the PSA nadir. Application of this definition to patients treated surgically delays determination of recurrence by 5 years.

Lee et al. have now compared the risks of biochemical recurrence following treatment of localized prostate cancer using radical prostatectomy, EBRT, or brachytherapy, to see if they are equivalent in terms of predicting PCSM. Data were collected for 13,803 men treated at two high-volume, US hospitals between 1995 and 2008. Validated, treatment-specific nomograms were used to calculate the 5-year progression-free probability (5Y-PFP) for each patient, which was compared with PCSM. Patients treated using EBRT had higher 10-year PCSM than surgically treated patients (5Y-PFP >75%, 3.0% versus 0.9%; 5Y-PFP 51–75%, 6.8% versus 5.9%; 5Y-PFP 26–50%, 12.2% versus 10.6%; and 5Y-PFP ≤25%, 26.6% versus 21.2%). No differences were observed between patients treated by brachytherapy and surgery, although selection of limited numbers of generally lower-risk patients for brachytherapy reduced the power of the comparison.

This analysis demonstrates that biochemical recurrence is not equivalent in patients treated using radical prostatectomy and EBRT, and should not be used for direct comparison when choosing between these treatments.