Prostate Cancer: Emerging Concepts: Part I

Glossary

Androgen deprivation therapy: The elimination, by either medical or surgical means, of male androgens, used to treat prostate cancer.

Clinical and pathologic staging of cancer (cT1, cT2a, cT2b, cT3, pT1, pT2, pT3): A description of prostate cancer tumor size and degree of extension using clinical criteria, such as results of digital rectal examination. The letter “c” denotes staging using clinical criteria; “p” denotes pathologic staging, usually determined by the pathologist examining the radical prostatectomy specimen. T1, tumor not palpable; T2a, tumor limited to one lobe; T2b, tumor limited to both lobes; T3, tumor with regional extension.

Gleason pattern score: A pathologic grading system to describe the pathologic characteristics of prostate cancer on biopsy and surgical specimens. In general, pattern scores of 2 to 4 represent well-differentiated cancer; scores of 5 to 7 indicate moderately differentiated cancer; and scores of 8 to 10 suggest poorly differentiated cancer.

Stage T1c cancer: Nonpalpable prostate cancer that is diagnosed by examination of biopsy specimens. The biopsy is prompted by an abnormal prostate-specific antigen blood test result.

In 1993, Annals published a review article describing clinical aspects of prostate cancer [1]. An accompanying article [2] reported risk factors for prostate cancer, and subsequent articles discussed in detail the status of prostate-specific antigen (PSA) screening [3, 4]. In 1993, the annual incidence of prostate cancer was estimated to be 165 000 newly diagnosed cases. In 1996, an estimated 318 000 new cases will be diagnosed in the United States alone [5]. Thus, largely on the basis of the widespread use of PSA-based screening, many more patients are receiving a diagnosis and are being offered treatment at earlier stages of disease.

In August 1994, the Food and Drug Administration approved the use of PSA testing in association with digital rectal examination for early detection of prostate cancer. This approval has created a new spectrum of management decisions for a broad range of physicians, including many internists [6]. In the first part of this review, we assess these challenges; screening for the early detection of prostate cancer is covered in a separate articles by the Clinical Evaluation Task Force of the American College of Physicians (American College of Physicians. Clinical practice guidelines on the early detection of prostate cancer. Unpublished report).

New Concepts in Interpretation of Prostate-specific Antigen Values

Most of the initial screening studies that assessed an abnormal PSA value used 4.0 ng/mL as the upper limit of normal. Several investigations have challenged this normal value and have suggested refinements in the interpretation of PSA test results. Emerging concepts that may help interpret the significance of elevated PSA levels include density, velocity, age-specific reference ranges, and free values compared with bound values.

Prostate-specific Antigen Density

The PSA density refers to a numerical ratio determined by dividing the PSA serum value by the volume of the prostate gland, as determined by transrectal ultrasonography. Because prostate cancer tissue produces more PSA per gram of tissue than does normal prostate tissue or benign prostatic hyperplasia, elevated PSA densities (usually more than 0.15) may more strongly indicate the presence of prostate cancer. One study [7] generated a probability curve that predicts the likelihood of prostate cancer on the basis of the PSA density. Another study [8] indicated that in the presence of a persistently elevated PSA value and negative biopsy results, a PSA density greater than 0.15 could be used to predict that examination of a subsequent biopsy specimen would have a greater than 80% likelihood of detecting prostate cancer. Inherent difficulties surround PSA density, including errors in measuring the prostate volume with transrectal ultrasonography and a variation as great as 15% in repeated examinations in the same patient [9]. Thus, despite the suggestion that PSA density can differentiate between large groups of patients with benign prostatic hyperplasia and patients with prostate cancer, the sensitivity in extrapolation of these data to the individual patient varies considerably. Additional data are needed.

Prostate-specific Antigen Velocity (Rate of Change of the Antigen with Time)

Prostate-specific antigen velocity refers to the rate of change in the PSA value over time. Patients with an elevated PSA value that continues to increase with time are more likely to have prostate cancer than are patients with stable PSA elevations. Thus, patients whose PSA value changes by more than 0.8 ng/mL per year are more likely to have prostate cancer. This rate of change in the PSA value may help identify patients who should have biopsy even if the results of the digital rectal examination and transrectal ultrasonographic examination are normal. These data are largely based on longitudinal studies done in Baltimore [10], in which men had serum samples stored for decades and were then followed before various disorders, including benign prostatic hyperplasia and prostate cancer, began to develop. In this study, the rate of change in the PSA value with time could be distinguished among cohorts of patients who never developed cancer, who had benign prostatic hyperplasia, and who had both localized and metastatic disease. Subsequent studies have confirmed these velocity changes, indicating that levels that change by more than 0.75 ng/mL per year should be regarded with a high degree of suspicion. Although additional data are needed, the rate of change of the PSA value with time may be the most notable indicator of a malignant or premalignant change in the histologic characteristics of the prostate. Careful follow-up of the rate of change may be an early indication of low-volume (that is, curable) prostate cancer or may provide an opportunity for the implementation of preventive strategies.

Prostate-specific Antigen Reference Ranges Specific to Patient Age

In the past, one normal range of values of PSA was used, regardless of the patient's age. Recent data have shown that the upper-limit normal PSA value of a man who is in the fourth decade of life and has no known prostate disorders is less than the “normal” value of a man in his sixth or seventh decade [11, 12]. Benign prostatic hyperplasia causes an increased prostatic volume, which in turn leads to an increased serum PSA level even without a malignant change. Thus, refining the previously accepted normal values of 0 to 4 ng/mL for all men may improve the specificity, sensitivity, and positive predictive value of the PSA test in diagnosing prostate cancer. Table 1 shows one recent recommendation for age-specific reference ranges. The ultimate value of this approach is that it decreases the number of biopsies in older patients with PSA values greater than 4 ng/mL and that it may increase the rate of cancer detection in younger men with PSA values less than 4 ng/mL. Preliminary data suggest that such goals can be reached when these new reference ranges are used.

Serum Free and Bound Prostate-specific Antigen Levels

In a more recent refinement of the test for serum PSA values, the relative percentage of free PSA and PSA that binds to serum proteins (bound PSA) is determined. Several reports [13, 14] indicate that men with a higher ratio of bound to free PSA are more likely to have prostate cancer seen on examination of biopsy specimens regardless of the total serum PSA level. Although no confirmatory studies have been done, the serum bound to free ratio may be an important factor in determining which patients should have prostate biopsy regardless of the total PSA value. A recent study [15] established more definitive reference ranges.

Correlation of Prostate-specific Antigen Value, Clinical Stage, Gleason Pattern Score, and Prediction of the Pathologic Extent of Disease

One of the major clinical challenges in the management of clinically localized prostate cancer (for example, cT1, cT2a, cT2b, and certain cT3 Figure 1 is to provide appropriate, curative treatment of cancer that appears to be confined to the prostate gland. Although this seems straightforward, nearly half of patients who are originally considered to have organ-confined prostate cancer develop cancer that spreads beyond the prostate gland when radical prostatectomy is done (Table 2). In pathologic terms, this may be cancer that has spread into the regional lymph nodes or the seminal vesicles, cancer that has perforated or penetrated the prostate capsule, or cancer that has reached the “inked margin.” In this process, the pathologist, after receiving the prostatectomy specimen from the operating room, inks the outside surface of the gland and marks the left and right sides of the gland. If cancer is present at this margin, it is presumed that the in situ portion of the resected margin contains cancer that has exceeded the margins of surgical resection.

Prostate cancer that is not confined to an organ is associated with an alarmingly high incidence of recurrence. Table 3 shows the clinical and total recurrence rates after radical prostatectomy according to the pathologic state of the resected gland (the total recurrence rate is the clinical recurrence rate plus the recurrence rate for patients with detectable serum PSA values). The rates shown in Table 3 raise the question of whether radical surgery was appropriate, because these patients are often not cured by surgery alone.

By using the combination of the clinical stage of the cancer (as determined by digital rectal examination), other routine staging procedures (not including endorectal magnetic resonance imaging), serum PSA value, and histologic grade of the prostate biopsy specimen, probability tables can be generated to predict which patients will probably have organ-confined disease or positive lymph node involvement (Table 4) [24-26]. This information may be helpful in defining the most appropriate treatment choice. For example, a patient with a T3a lesion, a PSA value less than 20 ng/mL, and a Gleason score of 8 may not be an optimal candidate for radical prostatectomy because the probability of cancerous involvement of the lymph nodes is almost 50%. Instead, this patient may be considered for an initial staging lymph node dissection. If the results are negative, radical prostatectomy or radiation therapy may be considered. Other options could include the use of hormone therapy if either of the lymph nodes are positive or the use of neoadjuvant hormone therapy before more definitive localized treatments.

In contrast, a patient with a low Gleason pattern score (2 to 4), a low PSA value (< 10 ng/mL), and a T1 or T2 cancer that is nonpalpable or only minimally apparent on digital rectal examination could be offered a curative procedure in the form of radical prostatectomy or radiation therapy treatments because the likelihood of organ-confined disease is so high. When these probability tables are used, a greater influence in the management of patients with prostate cancer is assumed.

The low probability of lymph node involvement or bone scan positivity in patients with a low clinical stage, low PSA value, and a favorable Gleason biopsy score has led some researchers to question the utility of doing either a staging lymph node dissection or bone scan in the initial evaluation [27]. However, as with all other predictive indices, the variability is high. In a study at Memorial Sloan-Kettering Cancer Center, 226 patients had ultrasonography-guided transrectal needle biopsy of the prostate followed by radical prostatectomy. Of these, 30% had a disparity in the Gleason score of 2 or more between the biopsy and the prostatectomy specimen. In 5%, the Gleason score of the biopsy specimen was higher than that of the prostatectomy specimen; in 25%, the Gleason score of the biopsy specimen underestimated the grade of the lesion.

In addition, because degenerative joint disease is so common in men in the seventh and eighth decades of life, a baseline bone scan is a valuable reference point for future follow-up. However, the likelihood of an abnormal bone scan with an initial PSA value less than 10 ng/mL is extremely low.

For many patients with prostate cancer, the arguments about active treatment versus no intervention (watchful waiting) should be considered. Support for watchful waiting emanated from studies done in Sweden [28], in which patients with clinically localized prostate cancer were followed rather than actively treated. Because the survival rate of this “watched” group seemed to be similar to the survival rate of an age-matched control sample, proponents of a nonintervention policy argued that watchful waiting is appropriate. However, that study contained many methodologic flaws that make the results difficult to use as the basis for recommending management decisions to patients with curable cancers [29]. Other investigators have studied quality-of-life issues and additional years of survival in treated compared with untreated (observed) controls and have suggested that no active intervention has a benefit in several broad populations of men with prostate cancer [30-32]. These studies also had substantial flaws in their design and in the interpretation of their results.

Under certain circumstances, a policy of watchful waiting may be the appropriate treatment choice. Elderly men, or those with substantial comorbid disease that may severely limit their life expectancy, may be considered for observation if they have a normal to minimally abnormal digital rectal examination, a low Gleason score (2, 3, or 4) on assessment of biopsy specimens, and a slowly increasing PSA value (< 1 ng/mL per year). However, data are not yet available with which to make such recommendations for younger patients who have clinically organ-confined cancers and are otherwise healthy. We believe that such patients should be treated with curative intent rather than observed; the latter practice potentially allows the cancer to grow, metastasize, and become incurable. A recent study [33] suggested that watchful waiting may be appropriate in men older than 65 years of age who have a low-volume, low-grade lesion; many men older than 65 years of age will die of other causes (primarily cardiovascular) before a slowly growing prostate cancer shortens their life. However, it is difficult to extrapolate these data to an individual person.

The advent of PSA screening, which enables the detection of smaller and perhaps clinically insignificant tumors, has added to the dilemma by deciding which patients should have aggressive or conservative treatment. The familiar quotation by the late Dr. Willet F. Whitmore summarizes the problem succinctly: “Is cure necessary in those for whom it is possible? Is cure possible in those for whom it is necessary?” [34].

Long-term Management of Patients Treated with Clinically Organ-confined Prostate Cancer

Interpretation of Radical Prostatectomy Pathologic Findings

Pathologic interpretation of the radical prostatectomy specimen is important for prognosis and in directing decisions related to postoperative management. Usually, after the radical prostatectomy specimen has been removed, the pathologist receives tissue that contains the prostate gland, periprostatic tissue, and seminal vesicles. The specimen is then marked with ink, indicating the resected margin of the tissue. Often, different colors of ink are used to distinguish the right from the left sides of the gland. When the gland is sectioned, the pathologist then has a visual reference point for determining the anatomical boundaries of the cancer (the prostate capsule and inked margins serve as boundaries). Organ-confined disease indicates that the cancer is limited to the prostate gland and does not extend into or through the prostate capsule. Cancer that has spread through the capsule but is still within and not contiguous with the outermost inked margin is considered specimen-confined. The pathologist should also comment on the periprostatic involvement of the surrounding soft tissues and the presence or absence of the involvement of the seminal vesicles or lymph nodes. In a recent analysis [35] of patients in whom cancer extended to the inked margin, researchers concluded that only specimens in which the cancer-containing margin was irregular constituted a true “positive” margin. In specimens with a smooth surface, even if malignant cells were found on the inked margin, the hypothesis was that all of the tumor was shelled out, leaving a negative in situ interface. Whether this interpretation will correlate with long-term follow-up results is unknown. In the interim, however, the varying pathologic interpretations of what constitutes a positive margin remains confusing for clinicians faced with the dilemma [36-38].

As would be expected, pathologically organ-confined cancer in most series has had a favorable prognosis, as measured by the time to PSA failure (defined as an increase in PSA levels in a patient who had an undetectable PSA level after radical prostatectomy), the time to progressive disease, and survival. A worse prognosis accompanies more extensive localized cancerous growth, and the likelihood of both local and systemic relapse increases substantially in cancers that have penetrated the capsule, show a positive inked margin, or show cancer at the urethral margin. The prognosis is worst if the seminal vesicles or lymph nodes are involved (Table 5 and Table 6).

The use of PSA-based monitoring after radical prostatectomy may be the best indicator of disease activity. Prostate-specific antigen should be undetectable after radical prostatectomy, and a detectable value after surgery almost always implies the presence of cancer, either residual local or distant metastatic disease; usually, such cancer is undetectable radiographically or by physical examination for months to years after the initial increase in the PSA value. The PSA value should revert to undetectable levels within 1 month after radical prostatectomy.

In older prostatectomy series, surgeons commonly left a “button” of normal prostatic apical tissue to improve postoperative continence rates. This practice is no longer necessary because of improved surgical techniques. The modern radical prostatectomy permits the total removal of the prostate (and, thus, a nondetectable PSA value after surgery) with minimal morbidity from urinary incontinence.

Positive Margins and Elevated Prostate-specific Antigen Values after Prostatectomy

The management of patients who have either a positive margin or an elevated PSA value after radical prostatectomy remains controversial. A PSA value that becomes detectable within 1 year after radical prostatectomy (if it initially decreased to undetectable levels after surgery) generally implies systemic relapse. In contrast, measurable PSA values occurring more than 1 year after surgery are more likely to indicate localized recurrence. Some experts advocate blind biopsies of the urethral anastomosis in the presence of an elevated PSA value. Others recommend obtaining a biopsy specimen of the prostatic fossa only if a palpable abnormality is present; still others rarely obtain a biopsy specimen of the local area when the PSA value is increasing.

The two major treatments of an increasing PSA value after radical prostatectomy are localized radiation therapy and systemic treatment with hormonal therapy. However, neither therapy is universally recommended, and several investigators are considering experimental treatments. These treatments include inducers of cellular differentiation, such as retinoic acid or vitamin D analogues, and gene therapy using cytokine-transfected prostate cells [41].

Results associated with the use of radiation therapy vary [42, 43]. External beam radiation therapy may make PSA values undetectable and may maintain these low levels for 5 years in about 30% of patients. Although the follow-up studies have not yet covered long periods, patients who do not maintain undetectable PSA values appear to have relapse with physical or radiographic evidence of disease. Hormonal treatment is generally recommended at that time.

The elevations in the PSA value after radical prostatectomy often predates the documented presence of demonstrable cancer by months to years, and most patients remain asymptomatic. Thus the management and the exact timing for initiation of hormonal therapy are controversial. Considerations regarding treatment with either localized or systemic therapy should include such factors as the patient's age, desire to maintain sexual activity (if present after radical prostatectomy), and level of anxiety about living with an elevated PSA value. In addition to affecting sexual function, long-term hormonal ablation therapy may lead to substantial quality-of-life changes secondary to decreased muscle mass, increased fatigue, decreased stamina, breast enlargement, possible premature bone loss, hot flashes, and psychological changes.

Experimental studies done using agents such as finasteride and flutamide (which do not decrease serum testosterone levels but block the androgen receptor and inhibit the conversion of testosterone to dihydrotestosterone) show that these agents appear to minimize these secondary effects. In preliminary studies, these agents seem to maintain potency in most patients [44, 45].

Another way to minimize the effects of permanent androgen deprivation is by using intermittent androgen deprivation therapy treatment. In studies investigating this strategy, hormone therapy is administered cyclically and is usually stopped once the lowest PSA value is achieved. The therapy is stopped and begun again after the PSA value increases. Although preliminary clinical results with luteinizing hormone-releasing hormone analogue plus flutamide are encouraging, earlier studies done using diethylstilbestrol in patients with advanced prostate cancer showed minimal benefit [46]. In addition, animal studies using the Copenhagen rat model [47] showed no survival advantage in animals receiving intermittent therapy.

The management of PSA “failure” after radical prostatectomy is an important issue in the management of prostate cancer. Given the increase in the number of radical prostatectomies being done, the problem is likely to increase. It is predicted that nearly 40% to 45% of patients will develop recurrent disease, as judged by the presence of cancer that, on pathologic analysis of resected radical prostatectomy specimens, has either penetrated or perforated the prostate capsule, extended to the seminal vesicles, or involved the lymphatics and lymph nodes.

These pathologic findings are present even when patients with stage T1c disease are analyzed, because the likelihood of having cancer that is multifocal or moderately or poorly differentiated or that extends beyond the confines of the prostate gland is approximately 50% [48-50]. In patients with an abnormal digital rectal examination, the presence of more extensive disease is generally seen on pathologic examination, not on initial clinical estimation. Although not all patients who have disease that extends beyond the prostatic capsule necessarily have relapse, the recurrence rate, especially with margin-positive disease, can be appreciable (Table 3) (See also [16-2339, 40]).

Dr. Fair: Urology Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY 10021.