Testosterone therapy and its influence on growth and malignancy of the prostate gland

Joel M. Kaufman, MD, Associate Clinical Professor, University Of Colorado, USA

I’m an urologist in Denver, Colorado. I specialise in erectile dysfunction and I’m going to talk today about testosterone therapy from a clinical point of view as it relates to the prostate gland. Earlier today Dr. Morales gave some recommendations about testosterone treatment, and I want to say that I was pleased at what he had to say about the prostate, and agree completely with those recommendations. I hope to provide some of the scientific basis for the recommendations that Dr. Morales gave earlier.

I’m going to talk primarily from a clinical point of view, but just by way of background we recognize that hypogonadism is increasingly a condition of men over the age of 50. This is the age where we treat men for prostate conditions. The recent availability of Transdermal Testosterone Replacement has enhanced our ability to treat these men, but there are concerns about risks and safety of testosterone replacement in older men, especially the effects on the prostate.

Prostate cell growth, we know that DHT, we’ve heard earlier, is the primary androgen stimulator Glandular epithelial cells are androgen dependent through many growth factors. Androgens are required for the growth, maintenance and functional activity of prostate cells, and we know that men castrated before puberty or those with androgen insensitivity syndromes, do not develop BPH or prostate cancer.

I’m going to talk about testosterone and prostate safety. I’m going to talk about BPH and review the effects on symptoms, flow, size and PSA. I’m going to then focus on testosterone and prostate cancer. I’m going to give a few epidemiological studies, historical review of the problem, some current series and some recommendations.

Before I go on to my presentation I want to mention three presentations that were given at this excellent meeting. Mark Feneley and Malcolm Carruthers from London on Thursday reviewed 1500 hypogonadal men with a mean age of 54 years. Twelve had an abnormal PSA or abnormal DRE at screening and were found to have cancer. And of the rest, only seven in 2100 man years of follow up were diagnosed with prostate cancer based on a change in PSA, or rectal examination. And all seven of these had clinically localized disease. I think these investigators should be congratulated for the excellence of their screening, the excellence of their follow up, and for demonstrating the safety of this treatment.

Dr. Bain and Fine from Canada, I believe this paper is being presented later today so I won’t focus on this too much. They reviewed 48 hypogonadal men with abnormal PSA or rectal exam, and they found before treatment that nine of 18 either had cancer or atypia, seven cancer and 15 of 30 of these men, including eight with cancer, were diagnosed after they had been started on testosterone replacement therapy. Again, showing the importance of screening men before starting them on testosterone therapy, and the importance of careful follow up.

The series that I am most familiar with is the AndroGel series from Dr. Wang and Swerdloff. I was one of the participants in this trial and I would like to review some of these data since they do speak very well about prostate issues. I’m going to show a few slides having to do with BPH. In the handout, which is a review article in the Aging Male Journal, there’s a much more extensive list of references regarding these issues. In this trial, 227 men, with a mean age of 51, who are hypogonadal based on a total testosterone of less than 300, were randomised either to the testosterone gel or testosterone patch for six months, and then the gel for a 36 month follow up. It’s important to recognize that exclusion criteria in the study were a PSA greater than four, an abnormal rectal exam, or an abnormal urinary flow rate. So men with BPH were excluded from this trial.

This slide shows the International Prostate Symptoms scores in the gel groups five and ten gram, and the patch group. What you can see is that at base line these men have very little in the way of symptom scores. For non-urologists the symptom score of zero is the lowest you can get and 35 is the highest. And what you can see at day 90, there is essentially no change, in other words with all three of these treatments, at the end of 90 days there is no change in any prostate symptom scores.

This important slide shows what happens to PSA over 42 months. I’m going to talk a little bit more about this but what you can see is there is initially a very small increase, and then over the next 36 months very, very minimal changes, and these certainly are well within the normal range. We’ve recently reviewed the 42-month data and have shown that at that time period there’s still very little effect on PSA.

These show the actual data at baseline. Most of these men were below one and at final visit, very, very small changes. Now these are statistically significant but clinically totally insignificant. We looked at PSA changes in men younger and older than age 50, because obviously we were concerned that perhaps over the age of 50 this might be more of an issue.

This slide summarizes the men under the age of 50 and what you can see is that at baseline the PSA values are very low, and at final evaluation very small increases, statistically significant, but clinically not significant.

When we looked at the PSA changes in men over the age of 50, again the PSA numbers were a little bit higher, but still well within the normal range, statistically significant increases over the course of the study, but clinically insignificant due to the very low numbers.

Now, this is just summarizing the study again in the T-gel group. Very small changes in PSA on average, even in the 100 mg group. The bottom line shows that in this study, five patients during the course of the study were found to have a PSA greater than four. Four of these were in the 50 mg group, one was in the 100 mg group, and one patient, during the course of this study, was diagnosed with prostate cancer when his PSA went over four.

When we looked at patients after 42 months, 150 men were followed and overall four men developed a PSA above five, three cases of prostate cancer were detected after 42 months of therapy in 150 men.

I’m just going to talk about a few individual series in which various things like prostate volume and PSA were examined. In the first series, this is testosterone replacement in younger men with Klinefelter Syndrome. In this study the PSA and prostate volumes were looked at. There were 11 patients; you can see that they’re in the younger age group. They received intramuscular testosterone of 250 mg. every three weeks for up to 183 weeks, and they were evaluated with PSA and transabdominal prostate ultrasounds.

What this slide shows is that at the beginning of the treatment prostate volumes were small at around 10 ccs. Over the course of the study prostate volume increased. Testosterone also increased a lot, and there was no change in the PSA in this group of patients. So, in the Klinefelter Syndrome series, some increase in prostate volume, no increase in PSA.

In this study parenteral testosterone was administered to 48 hypogonadal men in the older age group, with a mean age of 65.9. This was reported in the Journal of Urology in 1997. The authors describe the PSA velocity was normal. They did a total of 11 prostate biopsies for abnormal rectal examinations, and all of the biopsies were benign.

A very important study that’s been quoted many times, from Urology 1997, looked at PSA and volume changes during the course of testosterone therapy. In this study 29 hypogonadal men, who are on intramuscular testosterone therapy were evaluated. At baseline the prostate size was 17 grams or 17 ccs. Testosterone was then withdrawn for eight weeks, and at the end of the eight weeks what you can see is the size of the prostate had diminished. They were then retreated with a testosterone patch. The prostate volume increased, the maximum size increase occurred by month three, and PSA results were parallel to this.

This shows in graphic form what happened, again at baseline already on testosterone therapy. This is the size of the prostate. When testosterone was withdrawn the prostates got smaller. When testosterone was re-administered, prostate volumes increased with a peak at about three months.

In a large series from an impotence clinic in Cleveland reported by Alan Seftel’s group in the Journal of Andrology about a year ago, they found 54 hypogonadal men as part of their evaluation for impotence. All men had a serum testosterone level below 300. To be included in this retrospective trial, the PSA had to be less than four and you can see the mean is about 1.8. This tended to be an older group of patients. The mean follow up was 30 months; some patients were followed up to 82 months.

What they found was that patients receiving intramuscular testosterone had a very nice improvement in their serum testosterone levels, nearly a 1000, and on average the PSA went up from 1.8 to 2.8. During the course of the study, six out of 54 men or 11% required prostate ultrasound when their PSA went above four, and one of these individuals was diagnosed with prostate cancer. His pre-treatment PSA was 3.7 and when it went up to 5.9 his clinically localized prostate cancer was identified and treated. They had 19 men who they followed for more than 36 months and in none of those individuals did the PSA exceed four.

And this is the table from that study looking at men who were followed for a minimum of 36 months - they had 19 patients. The mean follow up was about five years. An older group of patients, testosterone levels were very nice. Pre-treatment PSA of 1.0 up to 1.6 and no patients required prostate biopsy, or diagnosed with prostate cancer. This was one of the largest retrospective series of treating older men with testosterone over a long period of time.

Now, in concluding my remarks about testosterone and BPH, I should say that for virtually all studies, men with BPH have been excluded from the trials. Secondly, none of these studies have been more than four years in duration. However, with those caveats, treating hypogonadal men with testosterone has had little effect on BPH symptoms, PSA prostate size, or uroflow. The handout I’ve left out there gives more of the references. One of the things that I think would be very interesting would be to do a trial of actually treating men who have BPH with testosterone. The way I would design the trial would be to take men who have mild BPH, or perhaps moderate symptoms, treat them with or without an alpha blocker, and testosterone or not, or just treat them with a testosterone or placebo. It is my prediction that men who are made eugonadal, who do not have severe prostate disease, are not going to be adversely affected by making them eugonadal.

Now in the year 2002 at the ISSAM Congress, the following position statement was made; “Androgen administration is absolutely contraindicated in men suspected of having prostate cancer, carcinoma of the prostate”

An even stronger statement came out of the International Consultation on Prostate Cancer the 2002 meeting, sponsored by the World Health Organization and International Union Against Cancer. The quote from that meeting is, “A history of prostate cancer is an absolute contraindication to testosterone supplementation. Men with prostate cancer cannot be given testosterone supplementation under any circumstance.” These were the conditions under which I began treating men with prostate cancer with testosterone. I’ve been called brave by my colleagues.

This is an epidemiological study from Slater and Oliver in 2000. They looked at 25 studies in which men with and without prostate cancer were evaluated for their serum testosterones. In the 25 studies, in 15 of the studies the mean testosterone level was the same in men with and without prostate cancer. In four of the studies, the testosterone levels were higher in the cancer patients than in the controls. In six the levels of testosterone were lower than in controls. In other words, there is no consistent finding of testosterone changes in men who have prostate cancer.

In a different study, this was a Meta-analysis from Shaneyfelt, in the Journal of Clinical Oncology in the year 2000, they only considered prospective cohort and case controlled studies, and as a result only three studies were included. What they determined in these three studies was that men who had high testosterones were more likely to develop prostate cancer.

There were two papers that came out in the Journal of Urology last year that have been widely quoted, and I’d like to bring them to your attention. These papers looked at testosterone levels in men who were diagnosed with prostate cancer. What they found was that low serum testosterone seemed to be associated with worse pathological staging.

In the first paper, from Massengill et al, Journal of Urology 2003, this was a US trial of 879 men who had radical prostatectomies and had a preoperative testosterone drawn. What they found was that patients with non-organ confined disease, T3 and T4, had lower mean testosterone levels than men with organ-confined disease. But I would bring to your attention that the differences are very small. The ones with lower testosterone the mean was 340 versus 356. I would contend that this article is being over-quoted and these differences are so small as to not be that clinically meaningful. Further, there was no correlation in terms of PSA recurrence, whether you had a low or a normal testosterone.

In a similar article in the Journal of Urology from Vienna, 39 men were looked at after their radical prostatectomy in terms of histologic parameters that might have indicated a worse prognosis. Of the 39 men, 16 had a testosterone level below 300, and 23 had a testosterone level of above 300. What they found was that certain parameters such as microvessel density, and androgen receptor density, were correlated with the testosterone level. That is, men with the lower testosterone levels were more likely to have these adverse findings. But the most impressive thing is, the Gleason scores. In the low testosterone group, the Gleason score was 7.4, and in the normal testosterone group it was 6.0. So there may be some adverse influence if you have a low serum testosterone on the pathologic findings if you develop prostate cancer.

Where does the contraindication come from? That you cannot give testosterone to men with prostate cancer? Everybody knows that you can’t do that. It took me quite a while to actually figure out where it came from, and I believe it came from Dr. Willard Whitmore, a great Urologist at the Memorial Sloan Kettering Hospital in New York City. He reported with Jay Fowler in 1982, 52 cases that they had accumulated from 1949 to 1967 of men with bony metastatic prostate cancer. Thirty-four of these men had previously been treated with either oestrogen or orchiectomy and were in symptomatic relapse. Fourteen had been previously treated and were in symptomatic remission and four were symptomatic and untreated. But all had bony metastatic disease. The rationale of the study was that they were going to give these men testosterone to activate the cancer and then treat them with systemic chemotherapy. They gave patients a large amount of testosterone. They received a mean of 34 mg daily for 42 days on average, prior to getting their chemotherapy.

The evaluation, and this is the 1940s, 50s and 60s, was subjective and objective with the use of x-rays and acid phosphatase. There was no PSA back then. What they found was that 87%, 45 out of 52, had an unfavourable response, including 67% who developed an increase in bone pain, 38% an increase in acid phosphatase, and eight out of the 52, or 15% had serious morbidity or mortality.

In a table in that article they reviewed 138 patients from previous trials, in which testosterone had been administered to patients with prostate cancer, and what they found was that 128 out of 138 had an unfavourable response. I believe that it was based on this article in 1982, that the contraindication developed that you cannot give testosterone to men with prostate cancer.

Dr. Morgantaler, a well-known American urologist, routinely does prostate biopsies on every man before he starts them on testosterone therapy. This is his article in the Journal of the American Medical Association in 1986. What he described was that 14% of men, 11 out of 77 in this group, who had normal PSAs, normal rectal examinations, were found to have prostate cancer on routine biopsies prior to starting testosterone therapy. Fortunately, no one else has been able to duplicate those results, and most urologists feel that it’s not appropriate to do prostate biopsies prior to starting testosterone therapy in men who have a normal PSA, and normal rectal examination. And that was one of the recommendations that Dr. Morales made this morning, but this was a very startling paper that you should be aware of, that there are men that we’re starting on testosterone therapy who possibly could have prostate cancer.

Where do we stand now in the current era? Nowadays most men with prostate cancer are diagnosed with early stage. We can attribute that to widespread PSA testing. When I started in urology we hardly ever found anybody with early disease. Even an abnormal rectal examination is not necessarily early prostate cancer. With PSA testing we are now diagnosing prostate cancer most of the time at an early stage. We also now have aggressive therapy, primarily in the States radical prostatectomy, but also radiation therapy, and we have very, very high cure rates. We have cure rates in the range of 95% for early stage, organ confined, low PSA disease. Further, accurate pathologic staging and PSA follow up, do allow us with reasonable confidence identify patients who are probably cured. Lastly, I would make the claim that we have seen no evidence that being eugonadal worsens prostate cancer.

Dr. Morales in 2002 in the journal, European Urology, made the statement, and it was quoted again this morning, “After a prudent period (this remains an uncharted territory) without recurrence of their cancer, the pros and cons of Androgen Replacement Therapy should be carefully considered and the restriction may be lifted.”

I’m going to present my own series of seven patients. We’ve presented this at the Journal of the American Urological Association last spring. We have seven patients, one to 11 years after radical prostatectomy. The patients are four of them in Colorado, and three with my colleague Jim Graydon in Connecticut. All seven patients had Gleason scores of 6 or 7; actually only one of them had a Gleason 7. All of them had negative margins. In all cases, when the patients presented with their hypogonadism, their PSAs were undetectable. All were severely gonadal. Four of the men had serum testosterone levels less than 100. All have been treated effectively and safely with more than a one-year follow up.

And just to continue describing the series, all seven had preoperative PSAs in the range of 4.0 to 6.0. We know that that’s a curable patient most of the time. Four of these patients were on replacement therapy with testosterone at the time that their prostate cancer was diagnosed. They were desperate to resume therapy and were pushing their urologists to get back on their hypogonadal therapy with testosterone. Three of the seven patients were diagnosed later but were very symptomatic. All seven patients were carefully counseled regarding risk, and in all seven patients they have been kept in the low normal to normal range with careful testosterone monitoring.

Recently Dr. Morgantaler, who I just referred to, treated a number of men who were thought to be at high risk for prostate cancer; namely those with prostatic intraepithelial neoplasia, PIN. This was presented in the Journal of Urology just a few months ago. As part of his program of diagnose of biopsying every patient before starting him on testosterone, he biopsied 75 men who are hypogonadal, with a mean age of about 60, and what he found was that 55 were PIN negative and 20 were PIN positive. We know that men who are PIN positive have a fairly high risk of subsequently developing prostrate cancer. What he found at a year was that there was no difference in PSA changes between the two groups. He did a repeat prostate biopsy if there was a change in the rectal examination or the PSA, and he wound up only doing six biopsies. Two of them were in the PIN positive group and he did find one case of prostate cancer. But if you took 20 men who were PIN positive, and followed them for a year, for one of them to develop prostate cancer is really not unexpected. So, Dr. Morgantaler concluded that testosterone replacement was safe even in a high-risk group.

In summary, androgen replacement in hypogonadal men causes little change in BPH parameters, such as prostate size, symptoms, flow rates and PSA. Most men with prostate cancer are eugonadal, that is, they have normal testosterone levels. My own feeling is that bringing a hypogonadal man into the eugonadal, or normal range should not worsen prostate cancer. Men who are eugonadal should not be treated with androgens regardless of their prostate situation. But while controversial, hypogonadal men, cured of prostate cancer, can be treated with androgens.