Journal Logo

AACR Annual Meeting

Prostate Cancer

Four Diseases in One?

Tuma, Rabiya S. PhD

doi: 10.1097/01.COT.0000303124.95203.d8
  • Free

Washington, DC—If etiologic factors that underlie prostate cancer are to be identified and their role in prostate cancer understood, it may be necessary to divide prostate cancer two ways, aggressive versus less-aggressive and young versus old age at onset, said Walter Willett, MD, DrPH, during a “Meet-the-Expert” session here at the American Association for Cancer Research Annual Meeting.

Walter C. Willett, MD, DrPH: “The difference between aggressive and non-aggressive, I think, is well appreciated in the field, and most people, if they were looking at etiological factors, would now consider that you need to divide cases that way. But I am suggesting that dividing them by age may also be important.”

“The difference between aggressive and non-aggressive, I think, is well appreciated in the field,” said Dr. Willett, who is Chairman of the Department of Nutrition in the Faculty of Public Health and Professor of Medicine at Harvard School of Public Health. “And most people, if they were looking at etiological factors, at this point in time would consider that you need to divide cases that way. But I am suggesting that dividing them by age may also be important.”

In collecting and analyzing data on risk factors for prostate cancer, Dr. Willett and colleagues at Harvard have found several factors that appear to be more important in one or another of these subsets of prostate cancer.

For example, researchers studying a community of Seventh Day Adventists in California found a 2.4-fold increased relative risk of dying from prostate cancer in men who drank three or more glasses of milk per day compared with those who did not. Yet when the same researchers looked at prostate cancer incidence and dairy consumption, they found no association.

Dr. Willett and Edward Giovannucci, MD, ScD, Professor of Nutrition and Epidemiology at Harvard School of Public Health, have recently revisited the issue using prospective data from the Health Professionals Follow-Up Study. They found the same pattern—an increased risk for aggressive cancer but not for overall incidence. (Dr. Willett noted that it is not clear whether dairy products are the problem or whether it is calcium intake that comes with them.)

The researchers have also found that some risk factors seem to impart the vast majority of their risk toward young-onset or late-onset cancers.

Body Mass Index

For example, a higher body mass index (BMI) appears to be protective for younger ages, conferring a relative risk of 0.5 compared with leaner men, when BMI at age 21 is considered.

“In fact this is one of the strongest risk factors that we see for prostate cancer,” Dr. Willett said. It is not clear why this would be the case, but it resembles the pattern seen for premenopausal breast cancer.

However, in men over 60 the relationship flips, showing an increased risk of prostate cancer with higher BMI, although the association is not as strong as in the young men.

In these BMI analyses, the researchers didn't separate aggressive and non-aggressive disease. “If you have to divide by aggressive/non-aggressive and age, then you need a lot more data, because you are slicing it into smaller and smaller pieces,” said Dr. Willett.

Exercise, Lycopene

But BMI isn't the only risk factor that differs by age. Men less than 65 do not appear to derive a risk reduction benefit from vigorous exercise, while men over 65 do. (Again, this is similar to what has been found in pre- and post-menopausal breast cancer.)

Also, there is an inverse relationship between risk reduction and increasing lycopene intake in the population overall, but when the data are stratified by age, the older men derive substantially more benefit with a relative risk of 0.67 vs younger men with a relative risk of 0.89.

Although this framework is “a work in progress,” said Dr. Willett, he and Dr. Giovannucci can find some logic behind it. As men age, their levels of anabolic hormones, including testosterone and insulin-like growth factor (IGF-1) decrease, the relative ratio of estrogen increases, and adiposity increases, which is a key trigger for inflammatory factors.

“Therefore, we have hypothesized that the dominant etiologic pictures may differ by age,” Dr. Willett said. “In the younger age group, factors that promote cell cycling may predominant and hormones promoting cell turnover. Whereas at older ages, the inflammatory processes may dominate.”

If that is the case, then things that reduce inflammatory processes should correlate with reduced risk for older cases, which is the case. Vigorous physical activity reduces insulin resistance and should improve an inflammatory state. Also, antioxidants, like lycopene, and anti-inflammatory factors, like NSAIDs or omega-3 fatty acids, would help.

If the risk factors do vary between younger- and older-onset patients, then current study designs may limit how much is learned about disease in younger men, Dr. Willett said.

“Our studies tend to put heavy weight on the older cases just because we are following a cohort over time, and are not replacing cases with younger people. We are just watching people get older, so we see a disproportionate representation of older cases. That is something that really motivated our thinking about the desirability of looking at younger cases separately”


An example where this might help sort out risks is in the role testosterone plays in prostate cancer. Castration dramatically drops disease risk and hormone therapy has an influence on disease progression, which suggest that the hormone plays a role in disease. But researchers have yet to find a clear relationship. If, however, younger cases, in which testosterone is high, were separated out from older cases, scientists might have a better chance of finding an association.

Of course, just how to do this four-way division is not entirely clear. Given that most prostate cancers are diagnosed at a relatively early stage, it is not obvious how to determine which cancers should be labeled as aggressive.

Gleason scores help but are not ideal. One possibility would be to follow cases over time and see who goes on to develop aggressive disease and then reclassify them as appropriate in databases. However, in the traditional case-control study design cases are not followed beyond diagnosis, which currently precludes that approach.

Age as Stand-in for Some Biological Factor

In terms of splitting patients into young and old cases, the researchers don't think it is age per se that is important, but rather that age is a stand-in for some biological factor. They hypothesize that factor is the shifting “hormonal milieu.”

But what is the appropriate age cut-off for young and old in data analyses? In breast cancer studies, researchers use menopause as the breaking point, although it is clear that the hormonal changes that underlie it occur slowly over years. Similarly hormonal changes in men occur over years, leaving any particular cut-off age a bit arbitrary.

“I would emphasize that this is a hypothetical framework and certainly needs more examination to be more confident that some of these interactions will with age stand up to the test of time,” said Dr. Willett. “It does appear that risk factors may vary by age and age of exposure. Examination of age as an effect modifier and also the variables for which age is a surrogate is worth future study.”

© 2006 Lippincott Williams & Wilkins, Inc.
    Home  Clinical Resource Center
    Current Issue       Search OT
    Archives Get OT Enews
    Blogs Email us!