Appendix 4.

The middle ground

The effect of 5ARIs in the prevention of prostate cancer has been studied in two pivotal trials, PCPT and REDUCE. PCPT compared finasteride 5 mg/day to placebo in normal men > age 55 with PSA < 3.0. REDUCE studied men with an elevated PSA and a prior negative biopsy, considered at higher risk for prostate cancer than the PCPT population, who were randomized between dutasteride 0.5 mg/day vs placebo. In both studies, the end point was a prostate biopsy, performed at 7 years in PCPT and at 2 and 4 years in REDUCE. In both studies, patients were offered a ‘for cause’ biopsy if they developed a further rise in PSA or prostate nodule. These studies showed the following findings: PCPT and REDUCE showed a reduction in the likelihood of prostate cancer diagnosis of 25–30% This reduction was in Gleason 6 prostate cancer For men on 5ARIs, PSA performs better as a marker for PCa detection Both studies showed a small increase in the rate of diagnosis of high-grade prostate cancer (Gleason 8–10). This was both an absolute and a relative increase (280 vs 237 in PCPT, i.e., 43 more cases on the finasteride arm out of 4368 men biopsied, i.e., an increase of 1%); and an increase from 19 to 29 cases in REDUCE, i.e., 10 more cases out of 3298 on Avodart, an increase of 0.3%. On reanalysis of the REDUCE cases using the modified Gleason criteria, the Gleason 8–10 cancers increased from 0.5 to 1.0%, an absolute increase of 0.5%. The observation of an increase in absolute numbers of high-grade cancers seen in both PCPT and REDUCE is consistent. It is likely that the reasons for this increase are multifactorial. As demonstrated in multiple modelling studies, the effects of cytoreduction of the gland and increased performance of PSA at identifying men with high-grade cancer biased the studies towards finding more high-grade cancer in the 5ARI arms. These studies also suggest plausibly that the true effect of 5ARIs is a decrease in high-grade cancer. However, the possibility that, in rare patients, high-grade cancers may be induced cannot be dismissed completely. This effect is likely to occur in at most one in 200 patients, and likely in considerably fewer than this. Further, there are considerable benefits of avoiding a diagnosis of low-grade prostate cancer, including avoiding the ‘survivor’ label, and avoiding a risk of radical treatment. 5ARIs also have other benefits, in terms of reduced sequelae of BPH progression. While it is not possible to exclude with certainty the possibility that a small increased risk of highgrade cancer exists associated with 5ARI use, the preponderance of evidence suggests that most or all of the increased high-grade cancers seen in the two trials are related to ascertainment bias. Patients at high risk for prostate cancer thus benefit from the risk reduction associated with 5ARI treatment. Such patients should be counselled as to the risks and benefits of 5ARI treatment, including both the substantial reduction in the risk of diagnosis, and the slight increased risk of high-grade cancer. Such patients include those with a strong family history, racial predisposition, and persistently elevated PSA. The use of 5ARIs for men with BPH/LUTS should not be significantly influenced by this concern. Men who are not at increased risk for prostate cancer and have no BPH/LUTS but are concerned about reducing prostate cancer risk should be offered dietary and lifestyle modification (discontinue smoking, exercise, obesity avoidance, and dietary modification) rather than 5ARIs.

5ARIs: 5-alpha reductase inhibitors; PSA: prostate-specific antigen; REDUCE: REduction by DUtasteride of prostate Cancer Events; PCPT: Prostate Cancer Prevention Trial; RR: relative risk; BPH: benign prostatic hyperplasia; LUTS: lower urinary tract symptoms.