From the Department of Urology, The University of Texas Southwestern Medical Center at Dallas.

Presented at the continuing medical education symposium, “Urology: Frontiers 2000,” held on March 10, 2000, at Baylor University Medical Center.

Corresponding author: John D. McConnell, MD, Department of Urology, The University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Boulevard, Dallas, Texas 75390.

BIOLOGY OF PROSTATE CANCER

The tumorigenesis of prostate cancer parallels that of virtually all other tumors. In this process, predisposing genetic risk factors combine with a variety of environmental factors, which leads to the initiation of the tumor and to its progression. As with most cancers, dysphasia eventually progresses to invasive carcinoma, metastatic spread, and androgen-independent cancer.

Many models have been created to explain the development of prostate cancer. The classic paradigm at a molecular genetic level is similar to that described by Vogelstein for colon cancer. According to this model, as normal prostatic cells move through the histologic transition to the development of locally invasive disease, cells experience chromosomal loss in a variety of areas: areas that express important cell adhesion molecules and areas that encode enzymes that restrict the unregulated growth of cells. In addition, there may be overexpression of other signaling molecules that leads to cell growth.

Another series of events causes those cells, which have developed into localized disease, to transcend that local environment. For this to occur, cells must develop motility. They do this largely by producing enzymes that degrade the basement membrane to allow cells to gain access into the lymphatic and vascular system. These cells maintain their ability to respond to androgens throughout all of these stages, but at some point between this step and the development of androgen-independent cancer, they acquire growth regulatory pathways that are not androgen-independent.

While theoretically and at a macro level this stepwise progression is well outlined, we still do not know the essential events that occur in the development of invasive prostate cancer. We do know the disease has a genetic basis. The familial risk of prostate cancer is actually higher than that for colon or breast cancer. Even so, only about 10% of all cases of prostate cancer can be explained on the basis of inherited risk.

Although androgens are essential for the regulation of normal prostatic function and the secretion of prostatic fluids, and although the withdrawal of androgens leads to programmed cell death, we do not fully understand their role in the development of prostate cancer. Clearly, androgens do not cause prostate cancer. But, on the other hand, prostate cancer does not occur in their absence. Androgens activate a series of enzymes that prevent programmed cell death, and increasing evidence points to defects in the apoptotic pathway as responsible for the development of androgen-independent prostate cancer. We do know that androgens do not fuel the growth of cancerous cells; a variety of growth factors and other signaling processes inside cancer cells themselves leads to their proliferation.

Research continues to provide new information about the biology of prostate cancer. But so far, we have not been able to take advantage of that knowledge to design specific therapies for the management of prostate cancer. Perhaps the 3 most exciting areas for potential treatment modalities are 1) the development of drugs to address what has gone wrong in the wiring of prostate cancer cells (cell signaling); 2) the repair of problems related to adhesion of individual cancer cells; and 3) the development of targeted therapy to interfere with the normal interaction between bone and prostate cells that promotes a favorable environment in which cancer cells grow.

PREVENTION OF PROSTATE CANCER

Patients ask frequently what they can do to prevent prostate cancer. There are 5 basic categories of preventive measures (Table), and these have variable levels of scientific support.

The first comes from 2 studies that conclude that vitamin E lowers the risk for prostate cancer in certain high-risk populations. One of these studies involved smokers and used a very low dose of vitamin E (<100 units a day). We must extrapolate to apply those results to the general population. Nevertheless, vitamin E cannot be ignored as an option.

Second, the preventive value of selenium and lycopenes is based on epidemiologic observations. The benefit of selenium has been demonstrated only in regions where the diet is deficient in selenium; thus, we do not know if the diet in North Texas would be considered selenium deficient. Epidemiological studies also suggest that populations consuming higher amounts of lycopenes have a lower risk of prostate cancer. But other explanations may account for this finding--there may be other foods in the diets of people who eat lots of tomatoes, for example. Therefore, although these observations are interesting, they require further verification.

Third, the preventive benefit of soy protein has been extrapolated from 2 lines of evidence. One study shows that Asian men living in the USA are more likely to develop prostate cancer than Asian men living in China or Japan. Analysis suggests that the difference in these cancer rates is most likely attributable to soy protein levels in the diet. Soy protein has also been shown to slow progression of breast cancer and other malignancies.

Fourth, the drug finasteride had been shown to reduce the total level of androgen in the prostate by 50%. Randomized clinical trials to test its potential prostate cancer prevention role are under way, and results should be reported in about 3 years. Until then, finasteride should not be considered a potential preventive agent, regardless of its ability to lower prostate-specific antigen (PSA) levels.

Finally, randomized clinical trials are also beginning for selective cyclooxygenase-2 inhibitors. This is an exciting approach that has a sound biologic basis for the prevention of prostate cancer. There is also interest in exploring nonsteroidal anti-inflammatory drugs. These drugs have been shown to reduce the risk of colon cancer by about 30%.

Other issues need to be considered when evaluating efforts to prevent prostate cancer. The first relates to study design. Most men who have prostate cancer diagnosed during the course of various trials have their cancer at the time they entered the trial. Therefore, any so-called preventive agent may be slowing the progression rather than actually preventing the cancer. A 20-year study is necessary to conclusively resolve this issue. Second, extrapolation can be dangerous. For example, beta-carotene, which has been recommended as a preventive agent for certain cancers, has been shown in some studies to increase the risk of poorly differentiated lung cancer. Patients must understand that not everything natural is necessarily good. Third, as the studies with selenium demonstrate, some preventive benefits may be limited to certain subgroups of patients. Finally, some agents may carry unanticipated risks.

EARLY DETECTION

The issue of early detection has been one of the most contentious in prostate cancer research. We do not have, and probably never will have, a definitive trial to show that it reduces mortality. However, this year a study in Tyrol, Austria, demonstrated a decline in prostate cancer mortality in men screened by PSA testing. Clearly, our goals in employing early detection strategies would be to find cancers when they are confined to the prostate in men who have at least a 10- or 15-year life expectancy and to be able to institute some form of curative therapy with minimum morbidity.

Our tools for early detection are PSA levels and the digital rectal examination (DRE). These tools have been refined considerably during the past decade, primarily in the area of specificity. The combined sensitivity of PSA and DRE for the detection of prostate cancer is as good as that of any screening modality in medicine--as good as the Pap smear and better than mammography. The problem has been on the specificity side, where a significant number of men with abnormal tests, i.e., elevated PSAs, do not have cancer but instead have benign enlargement.

One of the major advances in the past 5 years has been the introduction of the free PSA assay. This allows us to determine the amount of circulating PSA that is complexed with other proteins vs circulating PSA that is free. Routine use of the free PSA has decreased the number of prostate biopsies.

Another major advance developed during the 1990s was an understanding that decisions about the risk of prostate cancer cannot be made in a snapshot of time based on a single PSA value. Men known to have developed prostate cancer will almost always have had their cancer preceded by an exponential increase in PSA some years before a clinical diagnosis. The controversy with PSA “velocity” lies with the determination of a valid cut point, i.e., what is a normal PSA change over time? For instance, in a 1-year period, the rate of allowable PSA change in a 55-year old man differs from the rate of change that might cause concern in the case of a 75-year-old man. Therefore, if asked what a normal PSA is, I would have to consider many variables. The art of interpreting PSA changes remains a major part of urological decision making.

Although interpreting PSA values is difficult, some criticisms levied against PSA and DRE are inappropriate. We are not diagnosing occult tumors. The rate of small, incidental tumor detection with these paradigms is low, perhaps 2% to 5%. And, while the rate of false-positive errors is a real issue with these measures, a similar rate occurs with mammograms.

Regarding other means of early detection, although specificity will improve, I do not expect anything more than minor advances in the blood test. With ligand-specific positron emission tomography scanning, we may be 5 to 10 years away from clinical use. This tool will probably remain too expensive for use with screening, but it may help solve certain diagnostic quandaries. Biopsy strategies may undergo some fine-tuning. Urologists may reach a consensus regarding how many biopsies are needed or when biopsies should be taken in patients who have previously had negative biopsies but continue to have elevated PSAs.

STAGING

Pathologic diagnosis is the only technique that allows us to identify capsular penetration. Computed tomography and transrectal magnetic resonance imaging have not been particularly helpful. A modification of Murphy's Law for urologists might be that, when you absolutely need a test to sort out a particular question, that test will fail you. Unfortunately, a significant probability (20% to 30%) of extracapsular extension and about half that rate of positive surgical margins exist among men with moderately differentiated cancers.

In the past decade, we have learned that we can make prudent, cost-effective decisions regarding imaging. The average patient--a man with T1c disease, PSA <10, and an absence of poorly differentiated cancer--does not need a routine bone scan. Several large studies among similar groups of patients show that computer-assisted tomography and magnetic resonance imaging are of minimal value in low-risk patients.

Additional tests, such as positron emission tomography, may be helpful in the future. In my opinion, the ProstaScint scan has great potential but has sensitivity problems. One exciting area is so-called molecular diagnostics, i.e., using the molecular characteristics of cells identified on biopsy, in addition to state and grade, to predict the risk of nodal spread and/or extracapsular disease.

THERAPY FOR PROSTATE CANCER

Watchful waiting

Many patients misunderstand the nature of watchful waiting. Watchful waiting does not mean that we wait to intervene until just before malignant cells escape the prostate. We do not have the tools for that. Instead, watchful waiting consists of following patients until they have metastatic progression and then putting them on hormonal therapy.

In a classic study on watchful waiting involving 220 patients with an average age of 74 years, the progression rate to metastatic disease was only 13%, and the overall disease-specific death rate was 10%. This study has been widely criticized because of the large number of patients with pathologic T1 disease and the poor follow-up of many of the patients. In subsequent studies, the progression rates in patients who had clinically significant cancer was 30% over 10 years. Of course, the progression rates in men with poorly differentiated cancer are well over 50%.

We are still trying to determine the ideal candidates for watchful waiting. Katan and colleagues, using a mathematical model, showed the distribution of additional life expectancy from time of diagnosis in men who are on watchful waiting. There was an average of a little under 11 years of additional life vs an additional 13 years of quality-adjusted life in men who have undergone radical prostatectomy. One has to be very careful in looking at these models because as assumptions and variables are changed, the results can be very different. When more modern projections of progression rates are inserted, the watchful waiting curve shifts to the left, with a mean of about 9 years. We need to have better ways to predict patients' natural life span, because it's an important variable in this equation.

Molecular diagnostics may have great value in helping us predict who will have disease progression. We also need better tools to allow detection of localized progression before progression occurs.

Radical prostatectomy

The best long-term study of radical prostatectomy shows an average 90% 15-year disease-specific survival in patients with clinically localized disease. With surgery, there is little decay in these rates over time: if a patient gets to about 7 years postoperatively with a PSA-free state, the recurrence rate afterwards is <2%. Conversely, the rate of recurrence with radiation therapy is about 1% to 2% per year indefinitely. Comparison of these rates indicates that surgery is most beneficial for younger patients.

Additionally, the length of hospital stay associated with radical prostatectomy has decreased. It now averages 2.5 days, and it may decrease to <2 days in the future. Overall perioperative morbidity has also been reduced. Progress is still needed, however, in 2 areas: incontinence and erectile dysfunction. Despite significant improvements, rates for these disorders are still far from ideal.

To improve incontinence rates, efforts need to focus on the striated sphincter and the bladder neck. As a general principle, the less dissection around the striated sphincter complex, the better. Even so, the best technique to accomplish this is debated. Some physicians have altered their technique and compared their results with historical data, but randomized trials comparing the techniques are required.

The bladder neck is the subject of a new controversy. Some surgeons believe that it is necessary to completely mobilize the bladder, taking the peritoneum off the anterior dome of the bladder to allow bladder mobility so that there is absolutely no tension on this anastomosis, and to spare the bladder neck. In my opinion, the bladder neck is not an essential part of this operation. I recently compared my incontinence results of wide bladder neck excision with the results of a colleague who always spares the bladder neck; they were about the same.

Other steps that may minimize incontinence are more precise ways to handle the dorsal venous complex and greater appreciation of the tremendous variability in the apical anatomy, which is also important in nerve sparing.

Nerve sparing to minimize erectile dysfunction is also controversial. Although, on one hand, we know where the cavernosal nerves reside and what we can do to preserve them, outcomes of nerve-sparing operations are highly variable. When I teach radical prostatectomy to residents, I worry more about them knowing how to do a non-nerve-sparing procedure. It takes specific effort to get wide margins on the posterior-lateral surface of the prostate. CaverMap may have some role for training programs and for the occasional surgeon, but I do not think it has been a major adjunct, at least with the present design. Until outcome data for this operation are reported from a nationwide trial or registry, controversies will continue.

Another significant question for this decade is whether lymphadenectomy needs to be continued in low-risk patients. In the past 10 years, I have operated on 2 men who had positive lymph nodes, and I think that experience is typical. Lymphadenectomy adds about $500 and some measurable additional morbidity to the cost of a radical prostatectomy.

Recently, laparoscopic radical prostatectomy has been introduced. A study of 120 cases reported an operating time of approximately 4 hours and a surgical conversion rate of only 6%. Seventy-two percent of patients were completely continent at 6 months, and half of the small number of men who had erections prior to surgery were potent postoperatively. Currently, laparoscopic procedures are quite feasible but extremely difficult to do.

Radiation therapy

Radiation therapy effectively treats prostate cancer. As urologists, we have been far too negative about this over the years. Still, while the initial cure rate for radiation therapy is, stage by stage, only slightly lower than that of surgery at 10 years, radiation therapy has a higher ongoing risk of recurrence.

Three-dimensional treatment has significantly decreased the risk of rectal and bladder morbidity, and most patients now enjoy very few problems with radiation therapy. Brachytherapy has clearly been shown as equivalent to external beam radiation at 7 or 8 years. However, subgroups of patients (those with higher-stage, higher-grade disease) have significant failure rates with brachytherapy. Some conclude that brachytherapy should be combined with external beam radiation, but there is little data supporting such an approach.

Good results have been demonstrated in men with locally advanced prostate cancer when external beam radiation is combined with hormonal therapy. Induction of androgen withdrawal appears to reduce time to recurrence in certain subgroups. However, it is not necessary in T1c patients with moderately differentiated cancer who are being treated with external beam therapy.

Hormonal therapy

In 1994, the Health Care Financing Administration spent $1 billion to treat prostate cancer. Half that money went toward the use of luteinizing hormone-releasing hormone (LHRH). A tremendous amount of money has also been spent to define minute differences in treatment outcomes. Unfortunately, the efficacy of hormonal therapy has not changed since Huggins first described it >50 years ago.

On the positive side, patient acceptance is clearly higher for gonadotropin-releasing hormone as opposed to surgical castration, but this may relate to how these options are presented. The benefit of combination androgen blockade (CAB) is probably minuscule and may be of no benefit to the average patient. Perhaps the only hormonal therapy deserving attention in the next decade is the potential role of antiandrogen monotherapy (e.g., bicalutamide or flutamide), which is now enjoying fairly widespread use in Europe.

Hormonal therapy prior to surgery has been viewed as a way to decrease the high rate (15% to 30%) of positive margins. One study claims that such pretreatment reduced positive margins 30% to 50%. However, a follow-up study showed absolutely no difference in recurrence rate. The reported reduced positive margins may have been produced by optical illusion: pathologists may have been unable to find cancerous cells because of artifact induced by the therapy. Moreover, hormonal therapy makes nerve-sparing surgery considerably more difficult. For these reasons, this therapy should not be used before surgery.

Patients who do benefit from pretreatment with androgen deprivation therapy are those with higher-stage diseases who are undergoing radiation therapy. An empirical question is what would happen if these patients were treated with just androgen deprivation.

There is no difference in survival between patients treated with LHRH agonists or with orchiectomy or diethylstilbestrol. Nor is there a difference in survival at 2 years between patients treated with CAB (i.e., an LHRH analog plus bicalutamide or flutamide) vs those treated with monotherapy. At 5 years, CAB provides a slight advantage--a 3% to 9% improved survival in absolute terms. Past claims that CAB works best in men with minimal disease (defined as a small number of positive spots on bone scan) appears to not be the case. Moreover, the Southwest Oncology Group study, the only trial designed to answer this question, reports no advantage of CAB in men with minimal disease. CAB increases both expense and morbidity at a benefit of only 3% to 9% improvement in overall mortality rates. Economists define cost-effectiveness in terms of $100,000 in return for 1 year of high-quality life. According to this threshold, LHRH analogs would have to be 20% less effective than CAB for CAB to be cost-effective, and orchiectomy would have to demonstrate a 10% difference. But, since the differential response for CAB is between 3% and 9%, neither of these thresholds is achieved. Combination therapy is not cost-effective for the average patient.

If one decides to use an antiandrogen, there does not appear to be any difference in which is chosen. Although data are limited, side effects of CAB appear to be higher than those with LHRH monotherapy.

A large debate rages between proponents of early vs delayed therapy. As yet, only 2 trials address this issue. The Veterans Affairs Cooperative Trial years ago showed a 12% advantage to castrating patients earlier in the natural history of their metastatic disease, but that trial was flawed because some of the patients randomized to no therapy were allowed to progress to death and were never given hormonal therapy. A trial conducted in Great Britain shows a 4% benefit in survival by starting androgen deprivation earlier in the course of the disease, but this result was limited to a subgroup of patients with asymptomatic bone disease. Also, “early” in this trial was defined as a positive bone scan.

In summary, the proven indications for hormonal therapy are to relieve symptomatic disease, treat asymptomatic bone metastases, and pretreat patients before radiation therapy (but not before surgery). Hormonal therapy is not indicated as primary treatment for organ-confined disease, even in the elderly, or following decimal-point changes in the PSA after treatment of localized disease.

Chemotherapy

One of the major advances in the past 5 years has been the introduction of somewhat effective chemotherapy regimens, the most effective of which are based on the taxanes. In a group of patients with a mean life span, 10 years ago, of 6 months or less, a significant number of patients are enjoying a decrease in PSA and even objectively measured response rates with some of these chemotherapeutic combinations. This line of investigation is fruitful and will definitely improve the outcomes of patients with prostate cancer.