Health And Disease Status The Blsa--the Prostate Gland

Aging

Investigators

Linked publications & trials

Abstract

The Baltimore Longitudinal Study of Aging (BLSA) prostate aging and disease study has both retrospective and prospective arms involving repeated assessments of anatomical, physiological, hormonal, and behavioral aspects of age-associated changes in prostate size. The retrospective arm of the study examines the sex steroid and PSA levels from frozen sera stored during the three most recent visits and visits closest to 10, 15, 20 and 25 years before study initiation. The prospective arm involves male BLSA participants and has continued for more than 10 years. Data collection has been on hold since December 2002, and has just been restarted. Our previous work suggests that prostate cancer develops over a period of at least 10 years in most men, and that PSA can stratify men at risk as long as 20 to 30 years prior to diagnosis. Over the past year, several studies were completed and published, and we have focused our work on examining the risk for death from prostate cancer. Two studies examined risk factors for prostate cancer, one examined psa and prostate symptomatology, and one on the association of psa and all-cause mortality, and most recently examining psa relationship to death from prostate cancer. 1. Insulin, glucose and anthropometry: A major interest was to identify and understand the risks associated with prostate cancer. Risk differences in rates have been found based on geographical localization. Western lifestyles increase the risk of clinically relevant prostate cancer as compared to Asia. One factor implicated in this difference is related to dietary differences and resulting differences in obesity. We examined the relationship of insulin, glucose, and anthropometry with the subsequent risk of prostate cancer in men from the BLSA. Fasting insulin and glucose levels were unrelated to prostate cancer risk in our overall analysis. No risk was noted for body mass index or waist circumference, but the relative risk of prostate cancer for the second through fourth quartiles of the waist/hip ratio compared with the lowest quartile was 2.10, 1.96, and 2.06, though these ratios were not significant. The results of this study do not support positive associations of markers of insulin and glucose metabolism and obesity with prostate cancer. 2. NSAIDS: Laboratory and epidemiologic studies suggest that aspirin and non-aspirin non-steroidal anti-inflammatory drugs (NSAIDs) reduce the risk of cancer, possibly via inhibition of the cyclooxygenase enzymes. We evaluated the association of aspirin and non-aspirin NSAIDs with subsequent prostate cancer in BLSA men. The study suggested that men, particularly younger men, who had ever used aspirin or non-aspirin NSAIDs may have a modestly lower risk of prostate cancer. 3. Relationship of psa to lower urinary tract symptoms: Benign prostatic hyperplasia (BPH) is a common disease of aging men that can cause bothersome lower urinary tract symptoms (LUTS) affecting quality of life. There is growing interest in the identification of men who are more likely to have progression of BPH, in part because available medical therapy with alpha blockers and 5-#\ reductase inhibitors are now available for prevention of disease progression. PSA has been shown to be a reasonable surrogate for prostate volume with areas under the ROC curve ranging from 0.76 to 0.78 when PSA is used to predict prostate volumes cutoffs of 30, 40, and 50ml. Since PSA, as a surrogate of future prostate growth, is predictive of disease progression, and there is available medical therapy that can prevent prostate growth and disease progression, we evaluated the relationship between a baseline PSA and subsequent symptom progression over 3 decades among men who were not selected for the presence of LUTS in the Baltimore Longitudinal Study of Aging. We divided subjects into age groups of less than 50 years, and 50-69.9 years at the time of their first PSA evaluation. There was no statistically significant difference in the symptom score distribution across PSA percentiles for men age less than 50 years (p=0.87) or age 50-69.9 years (p=0.59). These data suggest that PSA level is not a useful predictor of the development of lower urinary tract symptoms in unselected asymptomatic men. 4. PSA and mortality: Because PSA is a marker of two diseases, prostate cancer and benign prostatic hyperplasia, that are closely related to aging, we wondered if PSA levels might serve as a proxy for the aging process. Our hypothesis was that men in whom prostate disease is present might be biologically more advanced in age than those without prostate disease. If so, all-cause mortality, as a surrogate for aging, should be directly related to PSA. We demonstrated a statistically significant direct relationship between PSA and all-cause mortality, cancer mortality, and the development of cancer. For each ng/mL increase in PSA, there was a 3% to 5% increase in the risk of an event. There was no statistically significant relationship between PSA and cardiovascular mortality, and this finding was not due to the lack of events in the cardiovascular group since most deaths were from a cardiovascular event. The inclusion of body mass index, testosterone, and free testosterone index did not alter the observations. One possible explanation for our findings is that the development of prostate disease with age reflects a loss of growth-control mechanisms within the prostate that also occurs simultaneously in other tissues within the body. If so, men with prostate disease (and elevated PSA levels) would be at greater risk of non-prostatic neoplasia. Although we do not know why men with the highest PSA levels are at an increased risk of development and death from cancer other than prostate, the relationships between PSA and cancer mortality and between PSA and cancer development suggests that the events that lead to PSA elevations with age may be related to the development of cancer in non-prostate tissues. 5. PSA and PSA velocity on death from prostate cancer. A substantial number of men who have PSA levels below 4.0ng/ml harbor prostate cancer, but no single threshold value can be used do determine whether a cancer will be aggressive and lead to death. We evaluated the rate of change in PSA (PSAV) as a method for identifying life threatening prostate cancer at a time when cure would be likely. Subjects were divided into those without prostate cancer (n=856), those with prostate cancer who were alive or died of another cause (n=106), and those who died of prostate cancer (n=20). At 10-15 years prior to diagnosis, men who died of prostate cancer had exponential changes in PSA that were absent among those with non lethal disease and those without prostate cancer. When average PSA levels were below 4.0ng/ml (10-15 years prior to diagnosis), cancer specific survival was 96 percent (90-99) at 25 years for those with a PSAV of 0.2ng/ml per year or less as compared to 70 percent (42-86) for men with a PSAV more than 0.2ng/ml per year (p=.02). The relative risk of prostate cancer death was 5.2 (1.5-18.6) for those subjects with a PSAV more than 0.2ng/ml per year compared to those whose PSAV was 0.2ng/ml per year or less (p=.01). At a time when PSA levels indicate the presence of curable disease, a rise in PSA of more than 0.2ng/ml per year indicates a relatively high risk of death from prostate cancer over then next 3 decades. This information could be useful for identifying those men who will benefit from active treatment. At this time, we are continuing to examine factors that contribute to prostate growth and the development of cancer, and strategies for early diagnosis of prostate disease.