European Urology

1. Introduction

The long-expected final report of the Reduction by Dutasteride of Prostate Cancer Events (REDUCE) trial appeared on April 1, 2010, in the New England Journal of Medicine together with a commentary entitled, “Chemoprevention of prostate cancer,” by Dr Patrick Walsh. Both documents are of clinical importance and warrant discussion.

2. The REDUCE trial

This trial investigated the value of the dual-action 5α-reductase inhibitor dutasteride (GlaxoSmithKline, Middlesex, UK) in 8231 men with prostate-specific antigen (PSA) values between 2.5 and 10 ng/ml who had undergone previous negative biopsies. Of these men, 6729 had at least one biopsy and were analyzed and presented as what the authors called the “restricted crude rate,” which, at all end points, was taken into consideration. The study resulted in a reduction of positive biopsies over a 4-yr period of 5.1%, which translates into a relative risk reduction of 22.8%. Favorable results reported from earlier studies with respect to parameters related to benign prostatic hyperplasia (BPH), such as significant reductions in acute urinary retention, BPH-related surgery, and urinary tract infections, were confirmed [1].

A number of issues have become subject to public discussion and will be addressed.

3. Is the risk reduction of positive biopsies after 4 years clinically relevant?

The study addressed a group of men and a PSA range in which biopsies are indicated in routine clinical practice: Most urologists worldwide would carry out periodic biopsies in men presenting with PSA values between 2.5 and 10 ng/ml. The chance of a positive biopsy in this setting was 24.9% in the placebo group and 19.9% in the dutasteride group: a 5% absolute difference. These detection rates are high compared with data of previously biopsied men from the European Randomized Study of Screening for Prostate Cancer (ERSPC), Rotterdam, where the probability of a positive biopsy after 4 yr was 11.8% in 872 men [2]. These two figures, however, can be considered to be in the same order of magnitude if one takes into account the difference in risk factors between the two populations. These are, mainly, the larger number of 10 biopsy cores, the high average PSA of 5.9 ng/ml, the large proportion of men with a positive family history, and the smaller prostate volumes.

Because of the initially elevated PSA values, the biopsies carried out after 2 yr and 4 yr within the REDUCE study cannot be compared with the “end of study” biopsies, which were taken at random and without consideration of clinical indications in men who initially presented with PSA values of 2.5–10 ng/ml after 7 yr of participation in the Prostate Cancer Prevention Trial (PCPT) trial [3], as claimed by Walsh in his editorial [4]. Also, the set up of the protocol avoids an influence of PSA change on biopsy indications in the dutasteride arm. This led to an equal number of biopsies in both arms of the study (an average of 1.41 per participant) and to very few biopsies outside the protocol, which showed very similar detection rates of 16.6% and 16.7% with only a small difference in the proportion of Gleason 7–10 cancers. In a person who requires periodic biopsies anyway, a reduction of the chance of a positive biopsy of 23% is clinically relevant. Obviously, the mechanism of action of dutasteride is of great importance in this respect. A decrease of negative biopsies may limit over diagnosis and unnecessary anxiety, but the effect on more distant end points remains unknown.

4. How does dutasteride act in reducing positive biopsies?

The authors and the reviewer, Dr Walsh, agree on this point. The mechanism is not primary prevention, but the reduction in cancer volume and the inhibition of growth of cancers that react to the intracellular reduction of 5α-dihydrotestosterone (DHT) in spite of the simultaneous increase in testosterone (T), which may increase to 40 times higher levels [5]. Considering the commonly applied definitions of secondary and tertiary prevention as the application of preventive measures to high-risk groups or to preexisting disease, dutasteride achieves a mixture of both. As the authors stated, it is likely that most of the cancers detected by protocol biopsies were present at the time of randomization. Dutasteride most likely acts by inhibiting their progression to larger lesions and by facilitating the detection of faster growing tumors because of the 30% volume reduction of the prostate in the dutasteride group.

Even considering the different affinity of T and DHT to the androgen receptor, it remains unclear why the treatment-related rise in T does not prevent these effects of dutasteride. While the study presents clinically relevant data, which are significant for a period of 4 yr, a number of important questions remain unanswered. What is the duration of the effect seen? Is there any effect on more relevant end points, such as clinical progression and death from prostate cancer? What is the exact mechanism of an increased number of fully differentiated cancers classified as Gleason 8–10 after 4 yr or 7 yr, respectively [3]? What is the fate of the 12 Gleason score 8–10 cancers found in the dutasteride group during years 3–4 of the study?

5. Can unnecessary biopsies be avoided by use of dutasteride?

A 19.9% positive biopsy rate was seen in the dutasteride arm after 4 yr. This means that 81.1% of all men had negative biopsies. The rate of biopsies is identical in both arms because the elevated PSA at inclusion remained the only protocol biopsy indication. This is an important difference with the PCPT trial [3]. The question of whether the degree of PSA response or the increase of PSA over time relates to cancer detection or to the detection of more aggressive disease after 4 yr is not addressed in the paper. It can be assumed that such analyses, in line with similar publications on the PCPT trial [6], [7], and [8], are in progress. If the rate of biopsies could be safely reduced by considering PSA response, as suggested in these references, this would add an important clinical indication to the use of dutasteride because the reduction of overdiagnosis and overtreatment is of great importance for future applications of screening for prostate cancer. In this respect it would be very important to further analyze those cases that present with high Gleason scores after 2 and 4 yr and to give account of their potential curability or probability to escape treatment. Hopefully, follow-up will be continued to report long-term outcome data in due time.

6. The aggressive cancers

As Walsh correctly states, there is no evidence that dutasteride prevents lethal tumors. This is a clear limitation that is inherent to the short duration of the follow-up within the REDUCE trial. In addition, the explanations given for the higher rate of poorly differentiated cases, which might have increased to an even larger difference with a 7-yr follow-up as applied in the PCPT study, remains unsatisfactory. Considering the presented evidence of proper randomization, it must be assumed that identical numbers of poorly differentiated tumors were present in both arms of the study. The selective suppression of the growth of usually well-differentiated cancers that respond to dutasteride and the 30% volume reduction of the prostate are likely to facilitate the detection of more aggressive, faster growing tumors at an earlier stage in the active treatment arm. Whether, as Walsh states, the lowering of PSA provides a false sense of security has to be investigated and discussed on the basis of these considerations. If a PSA rise under dutasteride treatment or a routine repeat biopsy every 2 or 4 yr based on the originally elevated PSA would selectively identify aggressive cancers in a curable stage, then his argument would not hold. This possibility, however, is considered by Walsh where he refers to part of the work by Thompson et al. [8] in his editorial.

7. Conclusions

In the view of these reviewers, the REDUCE study provides clinically important information. A relative reduction of the detection of small, potentially overdiagnosed cancers of 28.2% (derived from Table 3 of the paper by Andriole et al. [1]) is relevant for men who present with a clinical indication for re-biopsy because of an initially elevated PSA between 2.5 and 10 ng/ml and a previous negative biopsy. Another important aspect, addressed in this review, is that the possibility to save biopsies and to selectively identify aggressive prostate cancer has not (yet) been addressed by the REDUCE study group. The way in which this trial is reported does not take account of the effect of the reduction of biopsy indications resulting from dutasteride use. In clinical practice this is likely to result in further reductions of prostate cancer diagnoses. Obviously, this option can only be safely pursued if a rise of PSA under treatment reliably indicates a repeat biopsy. Relevant missing information, such as the stage distribution of the aggressive cancers and the applied treatment, could have been included in the original paper or in the appendix published simultaneously. It can be expected that the treasure of information embodied in the REDUCE database will be used to address the issues that could not be addressed properly because of space restriction in the present article. We are looking forward to further information.

Conflicts of interest

FH Schröder is an advisor to GlaxoSmithKline, Ferring, and Schering. MJ Roobol is an advisor to GlaxoSmithKline, Beckman, and Genprobe.