In the June 21st issue of JNCI, Shahinian and colleagues from the University of Texas in Galveston report on an analysis from SEER data evaluating the driving factors for choosing hormone ablation. Using a SEER-Medicare linked database, the authors identified 61,717 men diagnosed with prostate cancer between 1992 and 1999. A total of 1802 urologists were identified to assess the importance of physician preference as the catalyst for starting androgen ablation within 6 months of diagnosis.

Multivariate analysis was performed to quantify the variation in androgen ablation use between patient characteristics, tumor characteristics (grade or stage), and urologist preference. The patient characteristics evaluated included age, ethnicity, socioeconomic status, co morbid illnesses, year of diagnosis, and state of residence. One-third of patients had received androgen ablation within 6 months of diagnosis.

Interestingly, the percentage of the total variance in the use of hormone ablation attributed to the urologist (23%) was significantly higher that what could be assigned to tumor (9.7%) or patient characteristics (4.3%).

While this study suggests that the patient's choice of urologist may be a stronger determinant of whether he will receive early androgen ablation (over tumor characteristics), several potential flaws of the study should be addressed. As correctly stated by Dr. Schellhammer in the accompanying editorial, serum PSA was not available from the database to ascertain patient risk. Certainly a urologist would be more likely to start a patient on androgen ablation if his PSA was above 20 ng/ml. Furthermore, patients with Gleason score 7 were grouped with patients with Gleason scores 5 and 6, potentially diluting the impact of advanced tumor grade as a trigger for androgen ablation.

Certainly we must proceed with caution when narrow conclusions are drawn from broad databases. Nevertheless, as urologists we must accept that the use of androgen ablation therapy in low-grade, low stage prostate cancer is highly variable and perhaps efforts should be made for its standardization based on medical evidence.

J Natl Cancer Inst. 2006 Jun 21;98(12):839-45

read other UroToday.com Editors Picks

Written by Ricardo Sånchez-Ortiz, MD, a Contributing Editor with UroToday.

Please log-in or register in order to submit comments.

Powered by AkoComment!