Uncertainty in Prostate Cancer Care: Title and subTitle BreakThe Physician's Role in Clearing the Confusion

Early intervention for prostate cancer can theoretically cure a potentially disabling and deadly disease. However, evidence suggests that this approach may not improve survival and quality of life and may result in adverse effects. Therefore, men with clinically localized prostate cancer face difficult decisions regarding the management of their disease. To guide them in choosing between treatment options, patients seek information and recommendations from physicians.

The article by Fowler and colleagues¹ in this issue of THE JOURNAL suggests that urologists and radiation oncologists in the United States more commonly recommend interventions that they provide rather than accurately disseminating information about the uncertainty regarding the risks and benefits of early detection and treatment of prostate cancer. Based on a survey designed to assess clinicians' beliefs and practices regarding prostate cancer management, nearly all respondents recommended prostate-specific antigen (PSA) testing for men aged 50 to 74 years. In men with a life expectancy of less than 10 years (ie, older than age 75 years), routine PSA testing was still recommended by 43% of radiation oncologists and 16% of urologists. For men with moderately differentiated, clinically localized cancer and greater than a 10-year life expectancy, 93% of urologists recommended radical prostatectomy whereas 72% of radiation oncologists recommended radiation. Furthermore, 82% of radiation oncologists thought that radical prostatectomy was overused, while 13% and 26% responded that external beam radiation and brachytherapy, respectively, were overused. Conversely, 34% of urologists thought prostatectomy was overused. Virtually none of the physicians recommended observation, except for a small group of patients with low-grade tumors and a life expectancy of less than 10 years.

Recommendations for PSA testing and early intervention suggest they are proven strategies. However, global variations in the management and outcomes of prostate cancer, and the observed disagreements among US physicians on the use of interventions, highlight the lack of conclusive evidence that screening and treatment improve survival or quality of life. In Sweden, physicians rarely screen for prostate cancer or use radical therapies.² The most common treatment is observation with delayed androgen deprivation for palliation of symptomatic disease (watchful waiting). Prostate cancer mortality rates are higher than in the United States, suggesting this approach may not be optimal. However, mortality rates for prostate cancer decreased in Sweden from 1993 to 1996, even though rates of intervention with curative intent declined 5-fold.² Additionally, the only randomized controlled trial (RCT) comparing surgery with watchful waiting demonstrated no survival difference after 23 years.³ Even though this study was conducted prior to PSA testing and was too small to conclusively conclude that surgery is ineffective, it remains the most valid assessment of these 2 options. In the United Kingdom, mortality rates are similar to the United States even though PSA screening is not routinely performed. Radiotherapy is the primary treatment for early disease despite the fact that the only RCT comparing radiation with surgery demonstrated that disease recurrence was higher in men treated with radiation.⁴

In the United States, PSA testing is widespread and radical prostatectomy is the most common intervention for clinically localized prostate cancer.⁵ A 2.5% decline in annual prostate cancer deaths has occurred since routine PSA screening began in the early 1990s.⁵ However, geographic areas and health plans within the United States having the highest detection and treatment rates do not have the lowest rates of prostate cancer death. Furthermore, while the rate at which men are diagnosed with prostate cancer has increased by about 80 per 100,000 since the early 1990s, prostate cancer death rates have only decreased by 4 per 100,000 and remain greater than in the 1970s and 1980s.⁶ The small decline began too early after widespread PSA testing and early intervention to be due to these factors.

The preferred therapy for localized prostate cancer currently is not known and can only be determined by conducting RCTs. Selective interpretation of uncontrolled reports (ie, regarding surgery, antiandrogen therapy, external beam radiotherapy, brachytherapy, cryotherapy, or watchful waiting) leads to advocating a particular option by a specialist or country and dismissal of other options. This apparent "salesmanship of medical care" is not unique to prostate cancer and in fact may be well intended. However, results from previous RCTs are reminders of the limitations of observational research and the importance of conducting prostate cancer trials. For instance, screening for lung cancer with chest radiographs was widely promoted before RCT results demonstrated that screening did not improve survival.⁷ Likewise, bone marrow transplantation was advocated for curative treatment of advanced breast cancer until results of RCTs demonstrated that this approach was ineffective.⁸

The recommendations reported by Fowler et al also do not accurately reflect the conclusions of the American Urological Association and the American College of Physicians Guidelines Panels, which indicate that radical prostatectomy, radiation therapy, and surveillance are all acceptable options because data do not provide clear-cut evidence for the superiority of any 1 treatment.^{9 - 10} Furthermore, these recommendations may not only impede the conduct of RCTs that are required to solve unanswered questions, but also promote "questionable methods of cancer care."¹¹ Proponents suggest that marketplace demand and testimonials from satisfied patients are proof that their remedies work. In patients with early-stage prostate cancer, long-term survival is generally excellent with any treatment option, including watchful waiting. By the manner in which recommendations are made, clinicians reenforce the appropriateness of the intervention even if the therapy is unnecessary, ineffective, or produces adverse effects. Therefore, few patients report treatment-related regret. However, "gathering empirical data from clinical trials in cancer patients using valid, rigorous methods is the only means currently available for determining whether a treatment is likely to be of value. . . ."¹¹

Moreover, the dissemination of unsubstantiated theories, as proven medical care, increases practices likely to be harmful.^{12 - 13} For example, because most prostate cancers do not cause mortality or serious morbidity, early intervention is not necessary in the vast majority of men. However, while watchful waiting appears to provide comparable survival, and avoids the harmful side effects that can occur with surgery or radiation therapy, it is rarely recommended in the United States. For the minority of men with a form of prostate cancer likely to cause disability or death (ie, with high Gleason scores and high PSA levels), radical prostatectomy, external beam radiation, and brachytherapy appear not to be effective.¹⁴ In contrast to physicians' recommendations for routine PSA testing reported by Fowler et al, a thorough review and balanced synthesis of evidence from the American College of Physicians⁹ determined that routine PSA screening is unlikely to be cost-effective, unless unproven assumptions are used that favor screening regarding cancer-specific mortality rates, diagnostic and treatment costs, complications, and effectiveness. Screening men 75 years and older, as recommended by some radiation oncologists and urologists, is likely to result in net harm.⁹ Recommendations by urologists and radiation oncologists for early androgen suppression in men with high PSA levels and Gleason scores subject these patients to adverse effects from long-term hormonal therapy without proven benefits. This also suggests that these physicians believe radical prostatectomy and radiotherapy are ineffective in men with more aggressive prostate cancer that is likely to require curative treatment.

How can patients with prostate cancer receive the most appropriate care? For many patients, participation in an RCT can be a preferred option.^{9 - 10} Concern is often raised that participating in RCTs will result in poor health care because it is comparable with choosing therapy by flipping a coin. Unfortunately, selection of prostate cancer treatment outside a RCT is currently based on physician recommendations, patient testimonials, or news media reports that rely on anecdotal comments or evidence from methodologically flawed studies.

Participation in an RCT ensures that patients receive quality care, results are independently evaluated for safety and efficacy, and answers to important questions are obtained. Accrual in RCTs is likely to increase by enhancing public awareness of the importance and benefits of clinical trials as the treatment of choice for cancer and by making participation socially, medically, and financially acceptable and preferred as the best current choice for patients and physicians.¹⁵ (Description and recruitment for RCTs are available at http://clinicaltrials.gov or http://cancernet.nci.nih.gov/pdg.html.) For instance, the Prostate cancer Intervention Versus Observation Trial (PIVOT), the Prostate, Lung, Colorectal and Ovarian Screening Trial, and a trial comparing brachytherapy with surgery provide opportunities to answer major questions about prostate cancer. PIVOT may determine whether radical prostatectomy or watchful waiting provides superior length and quality of life in men with localized prostate cancer.¹⁶ Nearly 700 men have enrolled, demonstrating that recruitment is feasible, although more patients are needed. Baseline characteristics indicate that participants are representative of men who would be treated for localized prostate cancer. Among eligible men who decline enrollment, treatment selection is evenly distributed between surgery, radiation, and watchful waiting, indicating that men choose a wide range of options when presented with valid information about risks and benefits.

If participation in an RCT is not feasible, how can patients evaluate the vast amount of conflicting data and recommendations about prostate cancer management? As shown by Fowler et al¹ relying solely on recommendations of urologists and radiation oncologists is unlikely to provide information necessary for informed decision making. Physicians often have insufficient time available during the office visit to clearly explain the condition and the treatment choices (including watchful waiting). Physicians frequently underestimate patients' desire for and ability to cope with information, and may not be prepared for discussions of uncertainty. Both primary care physicians and specialists appear to lack accurate current knowledge and may disregard the evidence that reveals the unproven efficacy and known risks of treatment options.^{9 - 10} As a result, while a brief discussion of treatment options may be provided, patients are frequently confronted by competing, confusing recommendations for the specialists' therapy.

Primary care physicians can help clarify the confusion about prostate cancer management by being knowledgeable about, and providing patients with, information derived from evidence-based sources.^{9 - 10 ,17} For example, a report assessing androgen suppression for advanced prostate cancer demonstrated the feasibility of conducting prostate cancer RCTs and the usefulness of these reports in evidenced-based health care.¹⁸ Most other informational sources, whether directed at patients or clinicians, omit relevant data, fail to give a balanced or understandable view of risks and benefits, ignore uncertainty, and do not promote a participative approach to decision making. Unfortunately, all current prostate cancer decision aids rely on data from uncontrolled studies. Reliable estimates of risks and benefits require valid results from RCTs.

Primary care physicians have the opportunity to serve as their patient's independent health care coordinator and consultant. They can help them weigh the potential risks and benefits of treatment options, acknowledge treatment uncertainty rather than advocating a particular treatment option, facilitate the subsequent shared decision-making process, and incorporate patient preferences into the treatment plan. Shared treatment decision aids tailored to clinical characteristics and including elements of uncertainty as well as patient preferences and values can also be integrated into practice. These decision aids improve knowledge, reduce decisional conflict, and stimulate patients to be more active in decision making without increasing anxiety. For example, after viewing a treatment decision aid videotape, men with localized prostate cancer reporting their understanding of treatment as good to excellent increased from 44% to 95%.¹⁹

In the absence of definitive evidence that any 1 treatment is superior to another, including watchful waiting, primary care physicians can inform men about the commonly accepted options. An unbiased discussion of the estimates of the benefits and harms should be provided. For example, men with a prostate cancer with a low-to-moderate Gleason score (ie, from 2 to 6) can be informed that their risk for prostate cancer–related morbidity and mortality in the next 15 to 20 years appears small. Patients may choose watchful waiting to avoid complications related to surgery, radiation, or androgen suppression that can substantially decrease quality of life. Young men or those with a higher Gleason score (eg, ≥7) or PSA level are at greater risk for prostate cancer progression, disability, and death. These men may be more willing to accept the risk of treatment-related complications in an attempt to achieve cure or delay disease progression. Whether surgery, external beam radiation, brachytherapy, cryotherapy, or androgen suppression is a preferable early intervention strategy is not known. Selection depends on the patient's personal weighting of the results of the RCT comparing surgery vs radiation therapy,⁴ operative risk and desire for single or multiple outpatient radiation procedures, and avoidance of a surgical intervention vs a chance to remove the prostate gland and all the cancer. Additionally, primary care physicians should encourage their patients to evaluate how the reported frequency and severity of adverse effects of each intervention, including, but not limited to, urinary and fecal incontinence and erectile dysfunction, would affect their quality of life.^{13 - 14}

Among the 180,000 men in the United States diagnosed with prostate cancer each year, 80% have clinically localized disease.⁵ There is still no reliable information to determine the most appropriate treatment for these men including whether any intervention improves survival and quality of life compared with watchful waiting. The study by Fowler et al¹ indicates that recommendations for prostate cancer detection and treatment may reflect promotion of unproven yet, in most cases, sincerely held beliefs rather than dissemination of valid evidence. The opportunity exists for physicians to help patients resolve the confusion in prostate cancer care. Physicians also should encourage their patients to enroll in RCTs. However, until these RCTs are completed, physicians can serve as their patients' advocate by providing an unbiased presentation of the known risks and unproven benefits of all the treatment options and incorporating patient preferences into the treatment selection.