Despite having had surgery or radiation, about 25% to 40% of prostate cancer patients will experience a rising prostate-specific antigen (PSA) level, often a sign of recurrent disease. In these cases, doctors generally turn to hormone therapy to stanch the body’s production of testosterone, the male hormone (or androgen) that fuels the growth of prostate cancer. The drugs prevent the natural secretion of LHRH, short for luteinizing hormone–releasing hormone, which is the brain’s chemical signal to start testosterone production.
Hormone therapy for prostate cancer, also called androgen deprivation therapy or chemical castration, causes testosterone levels to drop by up to 95%. Testosterone levels can also be reduced by surgically removing the testicles. But given the irreversible nature and potential psychological impact of this procedure, it is not usually the treatment of choice.
Hormone therapy may slow disease progression for more than a decade in some patients; in others, it keeps cancer in check for only a few months. But eventually, prostate cancer cells begin to resist the treatment, and PSA levels rise anew. Depending on who you talk to, this condition is called androgen-independent, hormone-refractory, or castration-resistant prostate cancer.
Until relatively recently, men with androgen-independent prostate cancer* didn’t have many treatment options. They could try a second-line hormone therapy, start chemotherapy, or enroll in a clinical trial of an experimental drug. But now, thanks to some of those drug trials, as well as ongoing laboratory research, physicians have a better biological understanding of prostate cancer — and they may soon have more effective drugs to offer patients.
For an overview of the latest advances, our editors recently spoke with Steven P. Balk, M.D., Ph.D., an associate professor of medicine at Harvard Medical School and a medical oncologist at Beth Israel Deaconess Medical Center in Boston. Dr. Balk studies prostate cancer biology, and his research has helped show that structures called androgen receptors, located inside cells, play a central role in prostate cancer’s development and progression.
For starters, give us a little background on androgens and androgen receptors. That might help readers better understand what some of the emerging therapies will look like.
Virtually all prostate cancers start out dependent on androgens, such as testosterone, for their growth. Castration, through surgery or with medication, decreases androgen levels — testosterone goes down dramatically — making it an effective treatment. But because the adrenal glands also produce androgens and aren’t affected by these treatments, testosterone doesn’t drop to zero.
Early studies showed that suppressing these residual androgens might be of some value, and that led to the combined androgen blockade. [A combined androgen blockade uses two drugs. One acts on the brain to block androgen production, and the other stops androgen activity in prostate cells by targeting androgen receptors.] But later studies showed that the survival advantage associated with a combined androgen blockade is very small; some people would consider it insignificant. I think that’s why a lot of practitioners, though not all, abandoned the idea of further targeting androgen receptors in patients who relapsed after castration. They probably thought that the androgen receptor wasn’t playing a major role.
But we’ve since learned that in recurrent, or so-called castration-resistant, tumors, androgen receptors express at very high levels, meaning that there are more androgen receptors than normal. (The gene is actually amplified in about one-third of patients with these tumors.) Also, patients who were treated with certain androgen receptor antagonists developed mutations that rendered them insensitive to the drugs. Importantly, patients’ relapses almost always coincided with or were preceded by an increase in PSA, which is regulated by androgens. So, we had evidence suggesting that the androgen receptor was still active in patients with recurrent tumors, but we didn’t understand why it was still active — and even if we did, we didn’t have the tools to shut it off.
So the thinking has changed?
In the last four or five years, yes. Studies have shown that during a relapse, the level of androgens throughout the body, or systemic androgens, doesn’t go up very much. In fact, it is about the same as it was when the patient was “castrated.” But data indicate that androgen levels in the tumors are higher, back up to where they were prior to castration, so it seems that tumors step up production of androgens on their own in order to compensate. Had we seen a rise in androgen levels systemically, we would have investigated it. But because systemic levels of androgens didn’t rise, I think people decided that the relapse wasn’t related to androgens. Now we realize that the tumors themselves are adapting in a way that allows them to accumulate the androgens that are available or make their own androgens. Hence, the tumor has the ability to make the substance it needs to fuel its own growth.
Could a drug prevent the tumor from making androgens?
There are drugs that seem to do this. One drug, abiraterone, dramatically decreases androgen levels [see “Abiraterone’s action,” below]. With castration, a man’s testosterone level goes from, let’s say, 400 ng/dL down to 30 ng/dL. That’s a big drop, but testosterone is still being produced. Abiraterone blocks a key step in androgen synthesis, so when you give it to a patient, his testosterone will drop to 1 ng/dL or less. In clinical trials, we’ve seen about 60% of relapsed patients [patients whose prostate cancer has progressed despite the use of other therapies] respond to abiraterone, meaning that their PSA has declined by more than 50%.
Attard G, Reid AH, Yap TA, et al. Phase I Clinical Trial of a Selective Inhibitor of CYP17, Abiraterone Acetate, Confirms that Castration-Resistant Prostate Cancer Commonly Remains Hormone Driven. Journal of Clinical Oncology 2008;26:4563–71. PMID: 18645193.
Why not use the drug ketoconazole (Nizoral), which is already approved?
Some of the men participating in clinical trials of abiraterone have already tried ketoconazole. It targets the same enzyme that abiraterone does, CYP17, but it doesn’t do it nearly as well. In various studies, we’ve found that testosterone levels drop by maybe 50% — a man’s testosterone might go from 30 ng/dL to 15 ng/dL. You’ll see a decrease, but the drop in androgen levels isn’t nearly as dramatic as with abiraterone.
How do these drugs work?
To make androgens, the body turns cholesterol into the hormones pregnenolone and progesterone, which are progestins, through a series of chemical steps. Pregnenolone and progesterone are turned into DHEA. DHEA is turned into androstenedione, and androstenedione is turned into testosterone. To get to DHEA, however, you need an enzyme called CYP17. Without CYP17, the process of converting progestins to androgens is derailed; testosterone can’t be made.
Abiraterone inhibits CYP17, which is also called 17,20-lyase. Ketoconazole inhibits a variety of enzymes, not just CYP17. It will decrease testosterone, but only moderately. Patients who no longer respond to hormone therapy will often respond to ketoconazole, at least for a while. But because abiraterone specifically targets CYP17, it does a better job at reducing testosterone than ketoconazole.
Would you start a patient on ketoconazole first and then try abiraterone when it stops working? Or would you just start by prescribing abiraterone?
Abiraterone isn’t an approved drug yet, so we can only use ketoconazole. But if abiraterone gets approved, I think doctors will abandon ketoconazole and just use abiraterone.
Can patients get abiraterone as part of a clinical trial?
Most of the phase III trials of abiraterone for the treatment of castration-resistant prostate cancer have all the patients they need and have closed. One new phase III trial is recruiting patients in a few states. Patients can learn more about it by logging on to www.clinicaltrials.gov and searching for abiraterone and prostate cancer.
The big question is whether patients who take abiraterone will live longer. Presumably, the FDA will want to see a survival advantage, and intuitively, you’d think that if a drug is reducing the size of the tumor that you’ll see a survival advantage. But whether that survival advantage will be large enough to see in the study I just mentioned, I’m not sure. [Editor’s note: Based on a clinical trial showing that patients taking abiraterone lived about four months longer than those in a control group, the FDA approved abiraterone (Zytiga) on April 28, 2011. ]
Does the drug actually shrink the tumors or does it just stop their activity?
Many patients don’t have what we call “measurable disease.” They don’t have a mass that you can measure to see if gets bigger or smaller. They have disease primarily in bone marrow. You can see it on a bone scan, but you’re not looking directly at tumor activity. The only thing you can measure in these patients is PSA. A falling PSA generally correlates with diminished disease activity.
Does abiraterone cause side effects? Is its toxicity profile significantly better than that of other medications?
It’s had remarkably little toxicity. I’ve heard through the grapevine that a few of the men taking it have had some abnormalities in liver function, but over all, it seems to be extremely well tolerated. From that perspective, it looks like a pretty good drug.
Tell us more about the androgen receptor. How does it work?
In the absence of any androgens, the receptor is inactive or very nearly inactive. In order for it to become active, it has to bind to androgens — either testosterone or dihydrotestosterone (DHT). At that point, it’s as if the right key were put into a lock. It can move from the cytoplasm into the nucleus of the cell, bind to DNA, and turn on specific genes. It probably regulates hundreds of genes; some, like the PSA gene, we know about. But we don’t know or fully understand the entire spectrum of genes that it regulates.
In many cases of prostate cancer, we have discovered that an androgen-regulated gene will fuse with a known oncogenic transcription factor, a chemical that can interact with a cell’s DNA, let certain genes be turned “on,” and allow the cell to grow. Androgen receptors stimulate expression of these abnormal gene fusions, which can by themselves stimulate tumor growth. Therefore, it appears that one function of androgens is to increase these fusion genes. Decreasing these fusion genes directly (although no currently available drugs do this) or indirectly by inhibiting the androgen receptors may be effective treatments.
The androgen receptor itself has several different areas, or domains, each of which has a different task. One domain stimulates transcription, the replication of DNA during cell division, allowing cancer to grow unchecked. Another domain is the region that binds to DNA by recognizing specific gene sequences. There’s a small region that seems to be involved with the receptor’s movement into the nucleus and eventual degradation. The last domain is where testosterone and DHT bind. When the androgens latch on to the receptor, the receptor changes shape, which stimulates its movement and binding to DNA. The change in shape somehow recruits other proteins to join in the misdeeds of the cancer cell.
You mentioned that current drugs haven’t given men a lasting PSA response. They may work for several months, but then the PSA starts rising again. Why is that?
That’s what happens with androgen receptor inhibitors, or anti-androgens. Bicalutamide (Casodex) is probably the best of the currently available anti-androgens, but men still have minimal responses to it. With bicalutamide, the androgen receptor can still bind to DNA, but not as tightly as it can with testosterone or DHT. Moreover, when bound to bicalutamide, the androgen receptor does not effectively recruit other proteins that are needed to stimulate gene expression. It’s not entirely clear what happens when the tumors relapse after responding to bicalutamide, but it appears that bicalutamide no longer prevents the recruitment of these proteins, so the androgen receptor becomes active.
A recent study of a drug called MDV3100 noted that it could impair the androgen receptor. [See “Blocking the androgen receptor,” below.] How does that work?
Let me back up and explain how the drug came to be. Scientists started by “castrating” mice with prostate cancer. They then waited until the mice relapsed, and they looked at the tumors before and after relapse. They noticed a consistent increase in androgen receptors.
That led them to hypothesize that maybe an increase in androgen receptors alone is enough to cause castration resistance. So they engineered cells to produce more androgen receptors. When they attempted to “castrate” mice with tumors containing those cells, they found that the animals were relatively resistant to castration; their tumors kept growing. In addition, the high levels of androgen expression also seemed to make conventional androgen receptor antagonists like bicalutamide ineffective.
Collaborating with colleagues at other institutions, the researchers began screening drugs that might block the androgen receptors in cells that were overexpressing them, which was a perfectly logical thing to do. They tested nearly 200 compounds and found two that were effective at inhibiting androgen receptors. One was the Medivation drug MDV3100, which bound to the androgen receptor five to eight times more readily than bicalutamide in laboratory experiments. Experiments also showed that MDV3100 helped prevent the androgen receptor from binding to DNA, and that limits the cancer’s ability to grow. Bicalutamide induces a change in the androgen receptor, too, but the androgen receptor can still bind to and affect the DNA.
Blocking the androgen receptor
Tran C, Ouk S, Clegg NJ, et al. Development of a Second-Generation Anti-Androgen for Treatment of Advanced Prostate Cancer. Science 2009;324:787–90. PMID: 19359544.
Has MDV3100 been tested in humans?
Yes. After seeing the results in mice, the researchers launched a phase I/phase II clinical trial to assess the drug’s safety and efficacy. In 30 men with prostate cancer that had progressed despite the use of first-line medications like bicalutamide, 22 saw their PSA drop for at least three months. Thirteen of the 22 saw their PSA drop by more than half. Follow-up with those patients continues, and researchers are now testing MDV3100 at higher doses in 110 more patients. So even though the data are preliminary, the drug looks promising.
Interestingly, in the study that was published, they argue that drugs like bicalutamide are actually agonists — they stimulate tumor activity — in advanced, castration-resistant prostate cancer.
Initially or over a period of time?
Initially. They say the reason that the drug doesn’t work is because it’s an agonist in advanced prostate cancer, and they point to the anti-androgen withdrawal response as evidence. [See “The anti-androgen withdrawal response,” below.] But that withdrawal response doesn’t occur in every patient. Plus, if you give bicalutamide to a patient who has castration-resistant prostate cancer but hasn’t been on the drug, you generally don’t see a big jump in PSA. So the drug is not functioning like testosterone, but it’s not clear why the drug doesn’t work in these patients.
The anti-androgen withdrawal response
Patients who use a combined androgen blockade (an LHRH agonist and an anti-androgen) for an extended period may experience anti-androgen withdrawal syndrome. This paradoxical effect occurs when androgen receptors in the cancer cells mutate and use the anti-androgen, which is designed to block growth, to promote growth instead. When this happens, 25% to 30% of those who stop taking the anti-androgen and continue using the LHRH agonist alone experience shrinking tumors and dropping PSA levels. This response to withdrawal from the anti-androgen typically lasts from six to eight months, but in some patients, it can last as long as two years.
What about mutations or changes in the androgen receptor during treatment?
A number of years ago, my colleagues and I wondered about mutations in the recurrent tumors because we wanted to find an explanation for why the androgen receptor might be active. When you initially castrate, the androgen level drops, and the androgen receptors are inactivated. When you look at the recurring tumors a few years later, the androgen receptors are now active, even though systemic, circulating androgen levels haven’t changed. We found that mutations in the androgen receptor provided a partial explanation for these relapses in some patients. We found these mutations in about 30% of patients who had been taking the anti-androgen flutamide (Eulexin), but rarely found the mutations in patients with relapsed tumors who hadn’t ever taken flutamide or in patients who’d been on it for a very short time. Importantly, flutamide actually stimulates these mutant androgen receptors, and many patients with mutations improved temporarily when the drug was stopped. Therefore, it’s clear that mutations can enhance androgen receptor activity in some, though not most, patients.
Are there any tests that can be done to see if a patient’s cancer is becoming androgen-independent, or as you call it, castration-resistant?
No. We just see the PSA start to rise. All patients develop androgen resistance at some point, so you can tell the patient when you start the castration therapy that he will develop castration resistance. There’s no test that tells you how long that will take, but it usually happens within two to three years.
Are you studying any agents that have novel mechanisms of action?
I’ve been very interested in developing novel, more effective androgen receptor antagonists. I’ve been collaborating with a colleague who does computer-based drug screening. If you know the structure of a protein, you can program the computer to comb through virtual libraries of compounds to see which ones might be able to bind with the protein. We’ve done this, and we’ve come up with a panel of compounds that we think might be good androgen receptor antagonists. But this work is still very preliminary.
We’re also trying to come up with novel drug combinations. Abiraterone appears to be useful, but tumors still recur. Is there something that we could combine with abiraterone to get a more lasting response?
We’ve covered some complex scientific concepts. What’s the main message readers should take away from this interview?
Some form of castration can be effective when prostate cancer returns, but invariably the patient will relapse.
In the last few years, I think there’s been a sea change in the way people think about these castration-resistant tumors. In some ways, these tumors are like the tumors before castration, because the androgen receptors have been turned on again. If you can figure out how to turn the receptors off again, you may be able to get treatment responses that are similar to the original response you got with castration.
But you can’t turn them off with castration again. That’s the problem.
That’s right. You can only remove the testes once, or block the testicular function once, so the trick is to target the mechanism that activates the androgen receptor. We have two tools in development that look like they can do that safely and effectively: abiraterone dramatically decreases androgen levels, and MDV3100 blocks the androgen receptor.
Will these drugs cure patients? No. Will they prolong survival? I think so. I think we will have secondary hormone therapies that, while not curative, will be effective for some period of time. Researchers must now focus on how to build on these advances to develop therapies that will be effective for much longer periods.
Originally published September 2009; last reviewed May 2, 2011.