Prostate cancer prevention: proof is elusive

Prevention is so much better than cure because it saves the labour of being sick. Thomas Adams, 1618 Inferior doctors treat the full-blown disease; mediocre doctors treat the disease before evident; superior doctors prevent disease. Nai Ching, 1st Chinese Medical Text, 2600 BC. Enthusiasm for prevention is hundreds, even thousands of years old. In the field of prostate cancer, profound differences in the regional variation of prostate cancer around the world (highest in North Americans and Scandinavians, lowest in Asians), despite the similar incidence of histological occult prostate cancer and shifts in the incidence in mortality amongst immigrant populations moving from low to high prostate cancer regions, led to a firm belief that clinical disease was preventable. This belief was supported by the known long initiation phase for prostate cancer, providing an opportunity over decades for diet and micronutrient intake to influence the likelihood of disease progression. In addition, many epidemiological studies have pointed to the benefits of fruit and vegetable intake high in vitamin E, selenium, beta carotene, lycopene and other micronutrients, and a diet low in animal fat. Recently, however, several pivotal studies have taken the bloom off the rose of prevention. In particular, the SELECT study reported a 17% increased rate of prostate cancer in men on vitamin E, and an increase in diabetes mellitus in men on selenium 1. The study was resoundingly negative. In addition, both high intake of multivitamins, and high dairy and calcium intake, have been associated with an increased risk of fatal prostate cancer 2. Folic acid intake results in an increased incidence of prostate cancer. Despite the positive results of the Prostate Cancer Prevention Trial (PCPT) and Reduce trials, the 5α-reductase inhibitors were not approved for prevention by the US Food and Drug Administration because of concerns about an increased risk of high grade prostate cancer, despite the reduction in positive biopsies in men on the drug (mostly attributable to a decrease in low grade cancer). Further, studies of the association between dietary intake of fruit and vegetables and prostate cancer are inconsistent. For example, one large study in 130 544 men found no significant association between fruit and vegetable intake, including cruciferous vegetables, and prostate cancer 3. Another study showed dietary modification, reducing fat and increasing fruits, vegetables and fibre, had no impact on PSA concentration 4. Yet, despite the negative intervention studies, a lingering spark of hope exists that the many positive population, epidemiological and pre-clinical studies supporting dietary approaches to prevention will be vindicated. The MEAL study in the current issue of BJU Int is, therefore, a laudable and ambitious initiative 5. Remarkably, 478 men have been randomized to validated dietary counselling intervention vs no intervention. The authors report the initial demographics and eligibility data in the present paper. It is undoubtedly the first of many publications that will arise from this important trial. Will this study achieve its ambitious goal, which was to demonstrate that prostate cancer progression can be influenced by dietary modification? While the initiative is laudable, I suspect the hurdles are insurmountable given the sample size and conceptual basis for the study. The study is being performed in men on active surveillance, and the primary endpoint will be the risk of disease ‘progression’. The study references the Redeem study, which showed a 44% reduction in ‘disease progression’ with dutasteride compared with placebo 6. What we have learned since the Redeem study was initiated more than a decade ago was that the major limitation of conservative management in men diagnosed with low grade prostate cancer on systematic biopsy is not disease progression as it is usually defined (i.e. developing worse disease over time); it is grade misattribution, based on sampling and missing co-existent higher grade cancer in pathological analysis 7. Higher grade cancer is present in ~30% of men with Gleason 6 cancer on systematic biopsy. Finding this on subsequent systematic biopsy is largely a matter of luck, location of the cancer, and biopsy strategy and number. In contrast, true grade progression (from Gleason pattern 3 to pattern 4 or 5) is uncommon, estimated to occur in only 1–2% of patients per year 8. The adoption of MRI and targeted biopsy into the surveillance algorithm has reduced the misattribution problem. Thus, the true ‘event rate’ (exclusive of misattribution) is likely to be in the 15% range at 10 years. A study with the power to detect a 20% relative difference in these events, i.e. a 3% absolute difference, would require >1 000 patients, followed for 10 years. In the Redeem study, the reduction in ‘progression’ was entirely related to a decrease in the volume of low grade cancer. Indeed, the rate of upgrading was 13% in both arms in Redeem; therefore, the decrease in progression in that study probably reflected the cytoreduction effect of 5α-reductase inhibitors, and not a real biological effect on cancer progression. To be meaningful, prevention studies in men on surveillance should therefore identify, at the very least, a real reduction in grade progression, based on state-of-the-art evaluation at baseline with MRI and targeted biopsies as warranted, and long-term follow-up. A decrease in the rate of volume progression of Gleason 6, a major endpoint of the present study, is not meaningful. In the study as described, which does not explicitly incorporate MRI, an imbalance in the number of patients having off-protocol MRI and targeted biopsies between the two arms could significantly bias the outcome. A further problem with long-term studies of dietary intervention relates to the well-known methodological limitations in this area, namely, ensuring long-term compliance, recall bias of food intake and contamination of the control arm. Nonetheless, the authors deserve strong congratulations for pursuing this major initiative. We will follow the course of this study with interest. None declared.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.