Changes in testosterone levels within individuals and risk of prostate cancer in hypogonadal men
The relationship between testosterone and prostate cancer has been an area of discussion among for decades. In 1941 it was reported that prostate cancer growth was driven by testosterone and that castration resulted in regression of prostate cancer.1 However, the 1941 study only provided evidence on the role of testosterone in the progression/regression of prostate cancer, not the development (initiation) of prostate cancer. Whether testosterone plays any role – good or bad - in the actual development of prostate cancer, as opposed to its further growth, is still unknown.
Here we present the results of a study which examined whether fluctuations in testosterone levels during adult life are associated with development of prostate cancer later in life.2 Specifically, the study investigated whether men with a testosterone drop earlier in life are at higher risk for prostate cancer than those without such a drop, and whether further declines or large variations in testosterone levels increase the risk of prostate cancer.2
What is known about testosterone and prostate cancer
Even though most studies show either no association between testosterone levels and prostate cancer or a reduced risk of prostate cancer in men with higher testosterone levels, there are a few studies showing an increased risk of prostate cancer.3 However, much of these inconsistencies appear to be due to differences in study designs, definitions and methodologies.
What all these studies have in common is that they were guided by a static paradigm, which investigated the relationship between a single testosterone measurement and later prostate cancer diagnosis. This approach cannot provide information about the relationship between changes of testosterone levels over time and prostate cancer risk. Furthermore, these studies did not analyze the role of testosterone in the development of prostate cancer in the context of individual variation of testosterone level.
To address shortcomings in previous studies, a new model has been developed, which suggests that the relationship between testosterone and prostate cancer development is dependent on:
- Magnitude of the age-related decline in testosterone.
- Individual testosterone level threshold to maintain normal function of prostate gland.
The new model emphasizes that the level of testosterone measured at a single time point is not indicative of prostate cancer risk. Instead, it is the magnitude of the age-related decline in testosterone that is a key factor. The risk of prostate cancer increases when testosterone levels fall below a threshold where prostatic cells reach the limit of their compensatory capabilities to function properly during conditions of low androgen levels, which triggers the prostate carcinogenesis process, as illustrated in Figure1.
The dynamic model may help explain conflicting findings in previous studies that only measured testosterone at a single time point. Figure 2 illustrates possible scenarios.
For example, men with prostate cancer who had higher levels of testosterone than others when they were young can have testosterone levels that are higher at the time of cancer diagnosis, although they may already have experienced significant declines in testosterone. If such men are included in research, high testosterone will be interpreted as a risk factor for prostate cancer when compared with men without prostate cancer (controls) who had lower peak testosterone levels at young age and who experienced a smaller decline in testosterone (Figure 2 A).
In contrast, men with prostate cancer who had lower peak testosterone when they were young, will have lower testosterone levels after experiencing a significant decline in testosterone. In this scenario, low testosterone will be interpreted as a risk factor for prostate cancer when compared with controls who had higher peak testosterone levels at young age and who experienced a smaller decline in testosterone (Figure 2 B).
If all individuals from Scenario A and Scenario B were analyzed together, it is possible that no association between testosterone levels and prostate cancer is seen (Figure 2C).
What this study adds
To investigate whether intra-individual variations of testosterone levels contribute to risk of prostate cancer development, 376 untreated men with hypogonadism (testosterone 12.1 nmol/L or less), 45 to 74 years old, were recruited from a urology office in Germany. Age at study entry served as a surrogate for age at the first detection of testosterone below 12.1 nmol/L. Testosterone levels and PSA were monitored semiannually or annually during the entire follow-up period of 2 to 11 years.
Changes in testosterone levels in individual men were measured with 3 indicators:
- Intra-individual variation in testosterone levels over time.
- Maximum testosterone decline relative to the individual average testosterone level prior to the decrease.
- The average largest drop of testosterone in a short time period in an individual.
By the end of the study, a total of 26 new prostate cancer cases had been diagnosed (by biopsy). Results showed that the later (i.e. older age) the testosterone dropped below 12.1 nmol/L in a man, the lower his lifetime risk of prostate cancer. Further declines or dynamic variations of testosterone were also associated with increased risk of prostate cancer.
Men with a high intra-individual variation in testosterone levels over time had a nearly 5-fold increased risk for prostate cancer compared to men with a low variation. Men with a testosterone drop of 20% or greater from their individual average testosterone level during follow-up had over 8-fold higher risk for prostate cancer than men with a testosterone drop of less than 20% of their average. Men with the largest drop of testosterone in a short time period had a 2.7-fold higher risk of prostate cancer.
It was concluded that unstable testosterone levels, significant drops of testosterone levels, or decline in testosterone levels at a younger age increase the risk of prostate cancer. By collecting data from (semi)annual testosterone measurements for up to 11 years, this study provides insights regarding the role of testosterone in prostate cancer development, and suggest strategies for individualized testosterone therapy and prostate cancer prevention.
This is the first study showing evidence for an association between changes in testosterone levels and prostate cancer development. This knowledge may have substantial clinical impact on prostate cancer prevention and provides a critical message for doctors to start regular monitoring of testosterone levels, even in younger men.
Most previous studies on the link between testosterone levels and prostate cancer measured testosterone levels at a single time point or simply averaged testosterone levels during the studied time periods. This may explain previous inconsistent findings about the effects of testosterone on the risk of prostate cancer. In contrast, the study reported here conducted serial measurements of testosterone levels for up to 11 years, which allowed for detection of patterns of testosterone levels and their association with incident prostate cancer development.
Different role of testosterone in the initiation (onset) and progression of prostate cancer
Previous studies and discussions about testosterone and prostate cancer failed to distinguish between prostate cancer development vs. prostate cancer progression. These are two distinct processes that may be differentially impacted by testosterone.
Prostate cancer initiation is a process where normal prostate cells first adjust themselves to progressively declining testosterone levels. As testosterone levels fall below a critical threshold where normal prostate cells are not able to make additional adjustments without mutations, some of the normal prostate cells may turn into cancer cells.4 If testosterone therapy is started before reaching this threshold, prostate cancer development could be prevented.
Implications for testosterone therapy
The dynamic model suggests that testosterone therapy should be started before testosterone levels drop below the threshold beyond which normal prostate cells turn into cancer cells.
Given the large inter-individual variation in testosterone levels, androgen receptor sensitivity and thresholds, regular monitoring of testosterone levels is necessary to find out an individual man’s healthy baseline and to detect deviations. Monitoring of testosterone levels in the general population may need to start before age 30 since the incidence of prostate cancer in autopsy studies has been reported to be as high as 17% in men younger than 30 years old.5 Knowing a man’s young healthy baseline level of testosterone will also serve as an individualized reference therapeutic target during testosterone treatment.
Racial disparities of prostate cancer
Black men are approximately twice as likely to develop prostate cancer compared to white men.6 The increased risk of prostate cancer for blacks could be due to more rapid reductions in testosterone levels with age when compared to white men.7 During young adulthood, testosterone levels are higher in blacks than in whites, but the difference diminishes with age and completely disappears after 60 years of age.7-9 Thus, black men have a steeper decline in testosterone levels with age, which may explain, their higher risk of prostate cancer.
Comparison with previous studies
The present study confirms the results from a previous study, which found that faster age-related reductions in testosterone levels, rather than absolute testosterone levels captured either as a one-time measure or 5-year average, are significantly associated with the risk of prostate cancer.10 In this study, compared to men with a relatively stable testosterone, those with an annual testosterone reduction of more than 1 nmol/L (30 ng/dL) had 5-fold increased risk of prostate cancer.10 Both studies suggest that preventing the age-related testosterone decline with testosterone therapy may be an effective strategy for prostate cancer prevention.2,10