RISKS AND BENEFITS OF PROSTATE BIOPSY FOLLOWING AN ABNORMAL PSA SCREEN IN GENERAL AND IN A SPECIFIC PATIENT: DECISION ANALYSES

Purpose:  Although not ideal, discussion about outcomes after prostate cancer treatment sometimes occurs only after an elevated PSA screening test is found.  The choice of whether or not to proceed to biopsy may depend on patient preferences.  We performed decision analyses to determine if men with an elevated PSA should undergo a prostate biopsy.

Method:  For 65-year-old men with an elevated PSA and no other prostate cancer risk factors, we built a Markov model comparing five strategies: no biopsy (Bx); Bx with active surveillance (AS) for low grade cancer, followed by either radiation (R) or radical prostatectomy (RP) for subsequent progression; Bx with RP at diagnosis; and Bx with R at diagnosis. For higher grades of cancer found on biopsy, we made the assumption that patients would receive therapy (R) at diagnosis, and these results are incorporated into life expectancy (LE) and quality adjusted life years (QALYs) for all biopsy strategies.  We used a previously validated nomogram for predicting prostate cancer on biopsy and published literature for all other estimates including progression, mortality (SEER) and utilities.  We also considered a specific patient referred to our consultation service, and modeled his utilities using constant risk aversion (RA), reflecting his response to both standard gamble and time trade-off questions, and discounted his QALYs accordingly.

Result:

	LE (yr)	QALY General	QALY Index Patient
No Bx	11.8	8.9	5.9
Bx Low Grade AS then R	12.5	10.3	6.7
Bx Low Grade AS then RP	12.5	10.3	6.7
Bx Low Grade RP	12.6	10.2	6.9
Bx Low Grade R	12.6	10.3	7.0

In sensitivity analyses, no Bx was equivalent to definitive radiation therapy (for both low and high risk prostate cancer), and preferred over other strategies only if quality of life without biopsy was perfect.  For quality of life with no Bx of 0.99 or less, all biopsy strategies were preferred. In the index patient, no Bx was similarly the least favored strategy, with overall lower QALYs reflecting risk aversion.

Conclusion:  These analyses support performing a prostate biopsy given an elevated PSA for its benefit in LE and QALYs.  Foregoing biopsy would only be a viable option in patients who are worry-free and feel their health would be perfect.