Editorial: What have we learned from the Partin table update?
The controversies surrounding a physician’s best treatment strategy advice to an individual patient with clinically localized prostate cancer create a continuing need for advanced statistics. Historically, the Partin tables [1] were one of the first statistical tools that physicians and patients found readily usable. The tables have been updated and always focused on prediction of pathologic stage from standard clinical variables. The next commonly cited/used tool was the Kattan nomogram [2] that carried the prediction the next step to the endpoint of biochemical relapse. By 2008, Shariat et al catalogued over 100 predictive tools published from 1966 to 2007 on various endpoints of prostate cancer [3].
What have we learned from this update of the Partin tables?
- The pre-operative grade distribution has shifted up slightly with no change in prostatectomy grade/stage distribution. The authors discuss possible causes such as changes in interpreting the Gleason scoring system, shifts in selection for surgery away from lower grade patients, and a possible plateau in stage migration.
- The tables have split off Gleason 3+4, 4+3, 8, and 9–10, and found the latter significantly more aggressive, while Gleason 4+3 and 4+4 are more similar. Gleason 9–10 must have a pattern 5 component >5% and may therefore have more aggressive biology. On the other hand, two cases of prostate cancer may have identical volumes of 4 pattern, but if one adds additional 3 pattern, that additional tumour foci paradoxically lowers the sum to 7, but perhaps not the risk of non-organ confined stage.
- In the past, the tables were commonly used to predict pT3 stage, with possible change in management away from surgery as that risk increased. Clearly the literature on surgery for higher risk disease has matured, and augmented by the adjuvant/salvage radiation literature such that it is less likely to use the tables for this reason any more. On the other hand, prediction of N1 disease for the purpose of omitting a lymph node dissection remains a useful tool. In this update, using a <2% cut-off you would essentially omit all node dissections in Gleason 6 with PSA < 10 and cT1c/cT2a, while continuing with a dissection for any dominant Gleason 4 pattern. It is noteworthy that this experience was largely based upon standard templates, and those advocating extended templates will find these N1 rates too low. Indeed, when our center adopted the extended template using a robotic technique, the N1 rate for high-risk disease was 39% and 9% for intermediate risk [4]. Moving forward, what tools do we need to provide useful statistics to our patients? Updating old tools with more contemporary patient cohorts is certainly a worthy exercise. Multicentre study based tools will be required for endpoints such as positive surgical margins, quality of life, biochemical recurrence, and other endpoints that may be significantly affected by the experience of the treating physician. Beyond this, the next step should be adaptive nomograms that update in real time rather than en masse every 4–5 years [5].
John W. Davis
Department of Urology, The University of Texas MD Anderson Cancer Center, Houston, TX, USA
References
1 Eifler JB, Feng Z, Lin BM et al. An updated prostate cancer staging nomogram (Partin tables) based on cases from 2006 to 2011. BJU Int 2013; 111: 26–33
2 Kattan MW, Eastham JA, Stapleton AM et al. A preoperative nomogram for disease recurrence following radical prostatectomy for prostate cancer. J Natl Cancer Inst 1998; 90: 766–71
3 Shariat SF, Karakiewicz PI, Roehborn CG, Kattan MW. An updated catalog of prostate cancer predictive tools. Cancer 2008; 113: 3075–99
4 Davis JW, Shah JB, Achim M. Robot-assisted extended pelvic lymph node dissection (PLND) at the time of radical prostatectomy (RP): a video-based illustration of technique, results, and unmet patient selection needs. BJUI 2011; 108: 993–8
5 Vickers AJ, Fearn P, Scardino PT et al. Why can’t nomograms be more like Neflix? Urology 2010; 75: 511–3
