**3.2 The Reduction by Dutasteride of Prostate Cancer Events (REDUCE)**

Published in the *New England Journal of Medicine* in April of 2010, the REDUCE trial, was a 4 year multicenter, randomized, double-blind, placebo-controlled, parallel-group study. 8,231 men were randomized equally to dutasteride 0.5mg daily versus placebo. This trial was begun prior to the completion of the PCPT trial. Eligible patients were randomized to receive either Dutasteride or placebo. There has never been a large randomized trial comparing finasteride to dutasteride for the prevention of prostate cancer.

An important distinction between dutasteride and finasteride is the effect on 5-alpha reductase. Unlike finasteride, dutasteride affects the expression of both type 1 and type 2 isoenzymes of 5-alpha reductase inhibitors. Animal studies demonstrate that compared to finasteride, dutasteride has an increased reduction in both DHT and tumor growth.27 Dutasteride, then, theoretically could enhance the anti-tumor effect.

Eligible participants were required to be 50-75 years old, have a serum PSA between 2.5ng/mL-10mg/mL for men aged 50-60 years or 3.0-10ng/mL for men aged >60 years, and had undergone a prostate biopsy within six months of enrollment. Men were excluded if they had more than one biopsy, had prostate cancer of any grade, HGPIN, atypical small acinar proliferation, or a prostate volume of more than 80 grams, had previous prostate surgery of any kind, or had an international prostate symptom score (IPSS) of 25 or higher.

During the trial 6726 men (82.6%) underwent at least 1 biopsy and 1516 men (22.5%) were diagnosed with prostate cancer. The primary endpoint was the presence of prostate cancer

The Changing Landscape of Prostate

Cancer Chemoprevention: Current Strategies and Future Directions 435

responsible for a trend towards increased detection of higher grade tumors found in the "for cause" biopsies for the finasteride arm of the PCPT trial. Consequently, the "end of study" biopsies showed no increased risk of high grade tumor.30 The increased sensitivity of digital rectal exam performed while a patient is on finasteride has also been shown to increase

Finally, PSA sensitivity for detecting prostate cancer increased with the use of 5-ARIs. Both finasteride and dutasteride reduce the PSA laboratory value by approximately half. Consequently, PSA performs better (is more sensitive) in detecting high grade tumors.32 For the PCPT, for cause biopsies due to a higher PSA level artificially selected for more biopsies to be taken from the treatment group. Prostate volume decreases with finasteride use thus allowed for a larger percentage of the gland to be sampled due to the ratio of the needle to the prostate. The PCPT did not stratify on prostate volume size and it is unclear if this

A reevaluation of the PCPT data adjusted for prostate sampling density and baseline PSA levels found that the finasteride group actually had a small net reduction in high grade cancers and a 53% decrease in low grade tumors.26 A subsequent analysis of the PCPT data using end point prostatectomy instead of prostate biopsy demonstrated a 16% reduction in high grade tumor. At the time of prostatectomy, significant upstaging of cancer was detected in the placebo group but not in the treatment group. In fact, the misclassification of true high-grade disease (to low-grade disease on biopsy) was significantly lower for finasteride (34.6%) than for placebo (52.6%).33 Finally, another study looking at 500 patients who underwent prostatectomy demonstrated a high-grade cancer rates of 8.2% (placebo) versus 6.0% (finasteride), a 27% risk reduction (RR, 0.73; 95% CI, 0.56-0.96; P = 0.02) with finasteride.34 The authors concluded that it finasteride probably does not cause high grade

The REDUCE trial proved that there was no overall difference for high grade tumors between dutasteride and placebo during the entire study. However subgroup analyses demonstrated a trend towards Gleason 8-10 tumors in the dutasteride group study (19 in placebo vs. 29 in dutasteride, P = .15). In addition the statistically significant increase of high grade tumors in years 3 and 4 was alarming (1 in placebo vs. 12 in dutasteride, P = .003). The authors of the study concluded that the increased number of high grade tumors was due to the increased detection bias in the dutasteride arm as well as removal of low grade cancer in years one and two in the placebo arm. Post hoc analyses similar to the ones performed on

In December of 2010, the Federal Drug Administration's Oncology Drugs Advisory Committee (ODAC) reviewed 5-ARI's indication for the prevention of prostate cancer in men at increased risk for prostate cancer. GlaxoSmithKline submitted an application to add prostate cancer prevention as an indication for dutasteride, and Merck submitted an application to alter the labeling for finasteride to reflect a more favorable safety profile with

ODAC analyzed the results from the PCPT and REDUCE trials. They concluded that the majority of cancers prevented were low risk, thus providing no evidence of prostate cancer

sensitivity for detecting nodules. Detection increased from 16% to 21% (P=.015).31

theoretical bias altered the trend towards high grade tumors.

cancers, and may in fact improve the detection of them.

**3.4 Current political environment of 5-ARI** 

regard to preventing prostate cancer.

the PCPT are pending.

detected on biopsy 2 or 4 years after treatment. Biopsies performed out of the protocol were considered protocol independent biopsies. Other important endpoints included Gleason score, tumor volume, percent of positive biopsy cores, presence of HGPIN, and presence of small acinar proliferation. BPH endpoints were also evaluated.

The dutasteride arm represented an absolute risk reduction of 5% and a relative risk reduction of 23% (857 in the placebo arm versus 659 in the dutasteride arm, P<.001). This benefit was across all subgroups including age, family history, PSA level, prostate volume, or body mass index. The odds ratio for prostate cancer, detected on biopsy, with dutasteride as compared with placebo was 0.60 for all tumors (P<.001) and 0.62 for Gleason scores of 7 to 10 (P<.001).

Low grade tumors (Gleason 5, 6) were statistically higher in the placebo group (617 in placebo vs. 437 in dutasteride, P<0.001). The evidence of premalignant lesions was also decreased. HGPIN had a relative risk reduction of 39% (p<0.001) and small acinar proliferation (ASAP) had a relative risk reduction of 21% (p=0.04).

For high grade tumors, REDUCE demonstrated no significant overall increase in Gleason score 7 to 10, high grade, cancers. This was different from the PCPT, which alarmingly showed an increase in high grade cancer. Overall, there were 220 tumors with a Gleason score of 7 to 10 among 3299 men in the dutasteride group and 233 among 3407 men in the placebo group (P=0.81).

For the Gleason grade 8-10 tumors, there was no overall statistical difference looking at biopsies from all four years: 19 in placebo vs. 29 in dutasteride (P = .15). However, focusing on the second round of biopsies during years 3 and 4 of the study demonstrated a statistical increase in high grade tumors as compared to placebo. There were 12 tumors with a Gleason score of 8 to 10 in the dutasteride group, as compared with only 1 in the placebo group (P=0.003).28 The authors speculate that this was caused by more frequent early detection of low grade tumors in the placebo group that might have progressed if left untreated. As expected, dutasteride also demonstrated improved outcomes with BPH. Prostate volume, acute urinary retention and BPH related surgery, and urinary tract infections were all significantly reduced in the dutasteride group.

The REDUCE trial, like the PCPT, demonstrated a significant effect on low grade cancers (Gleason 5, 6), but did not appear to alter the prevalence of high grade cancers. Overall it was well tolerated but did have significant effects on libido, erectile dysfunction, and semen volume.

#### **3.3 Controversy: high grade tumor risk and generalizability**

The effects of 5ARI's on low grade cancer is consistent across both the PCPT and the REDUCE trials. However, each trial demonstrates a trend towards high grade disease. Initially, the authors from the PCPT concluded that the risk of finasteride on high grade tumors was uncertain. They recommended finasteride not be used as a chemopreventative agent.29

Secondary analyses based on the PCPT data were then performed to better understand the conclusions drawn from the data. These authors concluded that biases due to prostate specific antigen (PSA), digital rectal exam (DRE), and prostate volume detection were

detected on biopsy 2 or 4 years after treatment. Biopsies performed out of the protocol were considered protocol independent biopsies. Other important endpoints included Gleason score, tumor volume, percent of positive biopsy cores, presence of HGPIN, and presence of

The dutasteride arm represented an absolute risk reduction of 5% and a relative risk reduction of 23% (857 in the placebo arm versus 659 in the dutasteride arm, P<.001). This benefit was across all subgroups including age, family history, PSA level, prostate volume, or body mass index. The odds ratio for prostate cancer, detected on biopsy, with dutasteride as compared with placebo was 0.60 for all tumors (P<.001) and 0.62 for Gleason scores of 7

Low grade tumors (Gleason 5, 6) were statistically higher in the placebo group (617 in placebo vs. 437 in dutasteride, P<0.001). The evidence of premalignant lesions was also decreased. HGPIN had a relative risk reduction of 39% (p<0.001) and small acinar

For high grade tumors, REDUCE demonstrated no significant overall increase in Gleason score 7 to 10, high grade, cancers. This was different from the PCPT, which alarmingly showed an increase in high grade cancer. Overall, there were 220 tumors with a Gleason score of 7 to 10 among 3299 men in the dutasteride group and 233 among 3407 men in the

For the Gleason grade 8-10 tumors, there was no overall statistical difference looking at biopsies from all four years: 19 in placebo vs. 29 in dutasteride (P = .15). However, focusing on the second round of biopsies during years 3 and 4 of the study demonstrated a statistical increase in high grade tumors as compared to placebo. There were 12 tumors with a Gleason score of 8 to 10 in the dutasteride group, as compared with only 1 in the placebo group (P=0.003).28 The authors speculate that this was caused by more frequent early detection of low grade tumors in the placebo group that might have progressed if left untreated. As expected, dutasteride also demonstrated improved outcomes with BPH. Prostate volume, acute urinary retention and BPH related surgery, and urinary tract infections were all

The REDUCE trial, like the PCPT, demonstrated a significant effect on low grade cancers (Gleason 5, 6), but did not appear to alter the prevalence of high grade cancers. Overall it was well tolerated but did have significant effects on libido, erectile dysfunction, and semen

The effects of 5ARI's on low grade cancer is consistent across both the PCPT and the REDUCE trials. However, each trial demonstrates a trend towards high grade disease. Initially, the authors from the PCPT concluded that the risk of finasteride on high grade tumors was uncertain. They recommended finasteride not be used as a chemopreventative

Secondary analyses based on the PCPT data were then performed to better understand the conclusions drawn from the data. These authors concluded that biases due to prostate specific antigen (PSA), digital rectal exam (DRE), and prostate volume detection were

small acinar proliferation. BPH endpoints were also evaluated.

proliferation (ASAP) had a relative risk reduction of 21% (p=0.04).

to 10 (P<.001).

placebo group (P=0.81).

volume.

agent.29

significantly reduced in the dutasteride group.

**3.3 Controversy: high grade tumor risk and generalizability** 

responsible for a trend towards increased detection of higher grade tumors found in the "for cause" biopsies for the finasteride arm of the PCPT trial. Consequently, the "end of study" biopsies showed no increased risk of high grade tumor.30 The increased sensitivity of digital rectal exam performed while a patient is on finasteride has also been shown to increase sensitivity for detecting nodules. Detection increased from 16% to 21% (P=.015).31

Finally, PSA sensitivity for detecting prostate cancer increased with the use of 5-ARIs. Both finasteride and dutasteride reduce the PSA laboratory value by approximately half. Consequently, PSA performs better (is more sensitive) in detecting high grade tumors.32 For the PCPT, for cause biopsies due to a higher PSA level artificially selected for more biopsies to be taken from the treatment group. Prostate volume decreases with finasteride use thus allowed for a larger percentage of the gland to be sampled due to the ratio of the needle to the prostate. The PCPT did not stratify on prostate volume size and it is unclear if this theoretical bias altered the trend towards high grade tumors.

A reevaluation of the PCPT data adjusted for prostate sampling density and baseline PSA levels found that the finasteride group actually had a small net reduction in high grade cancers and a 53% decrease in low grade tumors.26 A subsequent analysis of the PCPT data using end point prostatectomy instead of prostate biopsy demonstrated a 16% reduction in high grade tumor. At the time of prostatectomy, significant upstaging of cancer was detected in the placebo group but not in the treatment group. In fact, the misclassification of true high-grade disease (to low-grade disease on biopsy) was significantly lower for finasteride (34.6%) than for placebo (52.6%).33 Finally, another study looking at 500 patients who underwent prostatectomy demonstrated a high-grade cancer rates of 8.2% (placebo) versus 6.0% (finasteride), a 27% risk reduction (RR, 0.73; 95% CI, 0.56-0.96; P = 0.02) with finasteride.34 The authors concluded that it finasteride probably does not cause high grade cancers, and may in fact improve the detection of them.

The REDUCE trial proved that there was no overall difference for high grade tumors between dutasteride and placebo during the entire study. However subgroup analyses demonstrated a trend towards Gleason 8-10 tumors in the dutasteride group study (19 in placebo vs. 29 in dutasteride, P = .15). In addition the statistically significant increase of high grade tumors in years 3 and 4 was alarming (1 in placebo vs. 12 in dutasteride, P = .003). The authors of the study concluded that the increased number of high grade tumors was due to the increased detection bias in the dutasteride arm as well as removal of low grade cancer in years one and two in the placebo arm. Post hoc analyses similar to the ones performed on the PCPT are pending.

#### **3.4 Current political environment of 5-ARI**

In December of 2010, the Federal Drug Administration's Oncology Drugs Advisory Committee (ODAC) reviewed 5-ARI's indication for the prevention of prostate cancer in men at increased risk for prostate cancer. GlaxoSmithKline submitted an application to add prostate cancer prevention as an indication for dutasteride, and Merck submitted an application to alter the labeling for finasteride to reflect a more favorable safety profile with regard to preventing prostate cancer.

ODAC analyzed the results from the PCPT and REDUCE trials. They concluded that the majority of cancers prevented were low risk, thus providing no evidence of prostate cancer

The Changing Landscape of Prostate

continue to make interpretation of the data difficult.43

**4. Discussion** 

prevented will benefit.45

from dutasteride prevention.47

2002;94:981-90.

**5. References** 

Cancer Chemoprevention: Current Strategies and Future Directions 437

Currently the only proven chemopreventative agents for prostate cancer are finasteride and dutasteride with up to a ~24% reduction in prostate cancer (24% PCPT, 23% REDUCE). However, as discussed earlier, the perceived risk of high grade cancers and the unclear long term benefit for low grade cancers has caused the ruling against the use of 5-ARI's for chemoprevention. Reanalysis of the Prostate Cancer Prevention Trial (PCPT) does not support this ruling and suggests that high-grade cancer is not associated with finasteride therapy and that prostatectomy is the only definitive diagnosis for the evaluation for prostate cancer.42 In addition, the effects of PSA, DRE and prostate volume bias will

It is important to note that as many of 30% of clinically insignificant cancers on first biopsy are then upstaged on second biopsy to become significant cancers. Consequently, the long term benefits of preventing low grade disease are uncertain.44 Also, by reducing the incidence of low grade disease, the potential for decreasing the "burden of treatment" is uncertain. But in the US environment where greater than 90% of men diagnosed with prostate cancer seek treatment, it is clear that individuals who have low grade cancer

Despite evidence from the PCPT and REDUCE trials, there does not appear to be a trend towards prescribing 5-ARIs for the prevention of prostate cancer. Approximately 64% of Urologists and 80% of primary care providers never prescribe finasteride for prostate cancer chemoprevention. Over half of Urologists reported concerns for inducing high grade tumors. In addition, half of the primary care physicians were not aware 5-ARIs could be used for chemoprevention.46 With the current ruling by ODAC, it is likely that physicians

Further analyses are required to pinpoint the subgroup population that will benefit most and the timing that is required to be effective. The PCPT revealed a high prevalence of prostate cancer, but no reliable markers were available to determine who would benefit most from biopsy or treatment. REDUCE prospectively collected samples to retrospectively analyze potential biomarkers but has not yet identified a subgroup that would yet benefit

In the current political environment, the adoption of 5-ARI medications is slim due to the uncertainty surrounding finasteride and dutasteride. Further analysis into subgroups of patients may reveal that 5-ARIs are clinically significant and economically available for at risk individuals. Maybe then, the benefits of 5-ARI medications will outweigh their uncertainty regarding high risk disease. In the meantime, investigations into new biomarkers, nutritional supplements and other pharmacologic agents may elicit a potential

[1] Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012;62:10-29. [2] Etzioni R, Penson DF, Legler JM, et al. Overdiagnosis due to prostate-specific antigen

screening: lessons from U.S. prostate cancer incidence trends. J Natl Cancer Inst

will trend away from prescribing 5-ARI medications for chemoprevention.

solution to the perplexing problem of primary prevention.

mortality reduction. Also, the controversy with the potential increase in high grade cancers could not be overlooked. Their conclusion was based on the fact that the risks of high-grade cancer were potentially real and could not be explained entirely by volume grade bias, increased sensitivity of PSA and DRE or removal of low-grade cancers. Finally, since the both REDUCE and the PCPT trials utilized end-of-study biopsies, the trial results were not yet generalizable to the US population. ODAC voted against the new indication for dutasteride (yes = 2, no = 14, abstain = 2) and against the new labeling for finasteride (yes = 0, no = 17, abstain = 1).

Due to ODAC's ruling, the FDA does not currently support the use of 5-ARIs for prostate cancer chemoprevention. After the decision made by ODAC, a publicly funded clinical trials cooperative group, SWOG (Southwest Oncology Group), released a statement disagreeing with ODAC's decision to not use 5-ARIs for the prostate cancer prevention.35

Clearly, disagreement remains after the publication of the two trials. GlaxoSmithKline has subsequently announced that it is withdrawing applications for similar approval in other countries.36 More long term data is needed to help experts agree on a consensus although we doubt that any further trials will be done with these agents to prevent prostate cancer. And likely because the other trials mentioned above have been negative, future trials for chemoprevention of prostate cancer will be rare.

### **3.5 Economic impact of prostate chemoprevention**

Current economic assessments for chemoprevention are based on varying assumptions that take into account data from the PCPT, treatment costs, and life expectancy. In addition, variations on the model include how the PCPT data is integrated into the model. For example, PCPT high grade disease can include projections based on the actual data set or based on a data set adjusted for biases.

To be effective, chemoprevention for cancer should cost roughly \$100,000 or less per life year (LY) saved due to low adoption rates when cost is greater than \$100,000/yr and high when cost is less than \$20,000/yr .37 Svatek *et al* used SEER data to estimate real world incidence and analyzed the cost-effectiveness of finasteride. Their group concluded that finasteride was too expensive at \$578,000 to \$1,107,000 USD per LY saved.38

Subsequent studies supported that chemoprevention was too expensive and estimated the expense per LY saved at up to \$1.6 million.39 However, an analysis on quality adjusted lifeyears (QALYs) based on PCPT prevalence rates showed a lower cost per LY due to higher PCPT prevalence rates. This analysis demonstrated \$122,000 per QALY saved.40 If finasteride is assumed to not increase the incidence of high grade cancers, then the analysis demonstrated \$112,000 per QALY saved. Thus more patients benefit from chemoprevention. These analyses were based on the current price of finasteride at \$66/month. If finasteride becomes less expensive, then the cost per QALY has the potential to drop significantly. Similarly, one study evaluated dutasteride based on the REDUCE study data and concluded that it too was not cost effective as a general chemopreventative agent.41

Currently chemoprevention with either finasteride or dutasteride is not cost effective. The identification of a high risk subgroup would potentially decrease the cost to benefit ratio thus making chemoprevention feasible from an economic standpoint. In addition, decreasing the cost of either 5-ARI has the potential to make chemoprevention feasible.
