Note from the National Guideline Clearinghouse: This guideline references a drug(s) for which important revised regulatory and/or warning information has been released.

ACR Appropriateness Criteria®

Clinical Condition: Pretreatment Staging Prostate Cancer

Variant 1: T1-2 and GS <6 and PSA <10 and <50% biopsy cores positive.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 2: T1-2 and GS < 6 and PSA <10 and >50% biopsy cores positive.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 3: T1-2 and GS <6 and PSA 10 to <20 and <50% biopsy cores positive.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 4: T1-2 and GS <6 and PSA 10 to <20 and >50% biopsy cores positive.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 5: T1-2 and GS = 7 and PSA <20 and <50% biopsy cores positive.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 6: (T1–2 and GS <6 and PSA >20) or T1-2 and GS = 8–10 and PSA <20 and <50% biopsy cores positive.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 7: T1-2 and GS >7 and PSA >20 or >50% biopsy cores positive.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 8: Clinical T3, seminal vesicle or bladder neck invasion.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Summary of Literature Review

Prostate cancer is the most common noncutaneous malignancy of men in the United States and is the second leading cause of cancer death in American men. The American Cancer Society recommends that men over the age of 50 have an annual digital rectal examination (DRE) and a serum prostate-specific antigen (PSA) test, and that men with a family history of prostate cancer or who are of African-American descent begin annual screening at age 45.

If either the DRE or PSA test suggests neoplasm, a transrectal ultrasound-guided needle biopsy of the prostate gland is usually performed. Alternatively, prostate cancer may be found in the tissue obtained during a transurethral resection of the prostate (TURP), although this procedure is becoming less common. Pretreatment staging is important, because clinically localized disease (stage T1 or T2) is generally amenable to local therapy, while more advanced disease may require multimodal therapy (e.g., hormonal ablation and radiation therapy). The TNM system encompasses the status of the primary tumor (T), the lymph nodes (N), and any metastasis (M) (See Appendix 1 in the original guideline document).

Digital Rectal Exam

The DRE, is considered insensitive for detecting extracapsular tumor extension. At least 40% of patients with cancers judged to be clinically confined (T1 or T2) by DRE are found to have extraprostatic extension at surgery. Thus, DRE alone has proven unsatisfactory for determining stage.

Prostate-Specific Antigen

Serum PSA is used as a biomarker, not only in identifying men with prostatic cancer but also in predicting pathologic stage, especially when combined with a patient's Gleason score, and for monitoring treatment response. In general, the higher the PSA, the more advanced the disease; moreover, the likelihood of having organ-confined disease is inversely proportional to the level of the PSA. Despite its utility, it is clear that as many as 15% of men with a normal PSA will have prostate cancer on one or more biopsy specimens. Recent data also suggest that the correlation with extent of disease is poor for men with relatively low PSA levels (e.g., <9 ng/ml).

The initial PSA value correlates with the likelihood of being free of biochemical evidence of persistent disease and surviving prostate cancer. PSA measurements are evaluated alone or by comparison with a prior measurement [PSA velocity and PSA doubling time (PSADT)], or in the context of the patient's gland volume (PSA density). There are also age-specific PSA levels available. The density and age specificity help to separate the elevations in PSA due to benign prostatic hyperplasia (BPH) from those due to cancer; however, these methods provide guidance only on the likelihood of cancer versus benign disease. The capability of PSA level alone to accurately predict final pathologic stage in an individual has a prohibitively high false positive rate. Recently, the bound and free components of PSA have been measured; the proportion of free PSA (ie, not bound to plasma proteins) was lower in patients with cancer than in those with BPH. For instance, free PSA values <15% were associated with more aggressive tumors, whereas free PSA values >25% generally had low-risk tumors.

Prostate Acid Phosphatase

With the introduction of PSA in the 1980s, prostate acid phosphatase (PAP) fell into disfavor because PSA performed significantly better in terms of screening and monitoring response to treatment. However, recent radical prostatectomy, external-beam radiation therapy, and brachytherapy series have demonstrated that PAP is a statistically significant predictor for biochemical progression-free survival and/or cause-specific survival in patients with intermediate- and high-risk prostate cancer. PAP appears to be particularly valuable in predicting distant failure in higher-risk patients for whom high levels of local control are achieved with aggressive local treatment. If PAP is to be introduced as a standard component of the initial diagnostic workup of prostate cancer, additional clinical studies are necessary to corroborate the currently published data.

Gleason Score

The Gleason scoring system has been shown to correlate well with the extent of disease and prognosis. It is the single best predictor of the biological activity, and therefore the stage, of the tumor. The scoring ranges from 2 (well differentiated, minimally aggressive) to 10 (anaplastic, highly malignant). The probability of seminal vesicle and lymph node involvement increases with the Gleason score, and some investigators have found a combination of the Gleason score and serum PSA level to give the greatest prognostic information.

Nomograms and Risk Group Stratification

The work by one group of investigators has led to the development of nomograms that predict the probability of extracapsular extension (ECE), seminal vesicle involvement (SV+), and lymph node involvement (LN+). This work was subsequently validated by others and led to attempts to correlate nomograms with prognosis. Most nomograms use combinations of clinically available prognostic factors such as PSA level, grade, and clinical T stage to estimate the risk. Estimates of the probability of LN positivity derived from such nomograms have subsequently been shown to be of use in determining the utility of staging studies and in guiding therapy.

Clinicians have widely adopted a simplified approach to predicting outcome based on the same pretreatment parameters used in the nomograms. Using such an approach, patients with similar risk of biochemical recurrence can be divided into risk groups that, with additional follow-up, have been correlated with mortality:

Alternative risk stratification schemes have also been described, but despite their differences they support the notion that Gleason score, clinical T stage, and PSA can be used to predict survival and direct therapy. More recently the number of positive biopsies (e.g., >5) and the percentage of each core that is positive for biopsy (e.g., >50%) have been associated with increased risk of recurrent disease.

Summary of Nonimaging Methods of Staging

While digital rectal examination, PSA test, or Gleason score individually predict stage, they are less accurate than when they are combined into nomograms that provide estimates of risk. Patients can be stratified by their risk for extraprostatic, nodal, and disseminated disease.

Imaging potentially improves these general estimates of risk by specifically identifying lesions with anatomic abnormalities. However, interpretation of imaging findings should be made in the context of the nonimaging findings. Due in part to the limitations of clinical staging, efforts have been made to use imaging modalities to better predict the extent of disease and outcome.

Imaging Methods

Ultrasound

Gray-scale ultrasound (US) has not proven satisfactory for local staging of prostate cancer. The ability of transrectal US to predict extracapsular extension varies widely from 37%–83% in different settings and populations; however, it is generally acknowledged that US is of limited value due to limitations of its spatial resolution. The addition of color Doppler and power Doppler improves the detection of prostate cancer by identifying increased vascularity but has not yet been shown to improve staging accuracy. Failure to identify a neurovascular bundle near the site of a tumor is suggestive of extracapsular extension, but there is not yet consensus that its use is mandatory for staging. Contrast-enhanced US has the potential to substantially improve the staging of prostate cancer but has not yet been tested in a multi-institutional trial. Similarly 3D US is under investigation to improve the delineation of the cancer and prostate capsule.

Magnetic Resonance Imaging

Endorectal coil magnetic resonance imaging (erMRI) provides the highest spatial resolution among the imaging modalities currently available. Three major techniques that have been used to stage prostate cancer with erMRI: T2-weighted MRI, MR spectroscopic imaging (MRSI), and dynamic contrast-enhanced MRI (DCE-MRI). It is generally accepted that an endorectal coil is required to achieve sufficient signal-to-noise ratios to allow small field-of-view (12–16 cm) imaging which, in turn, allows images to be acquired with high resolution (~0.5 mm). Additionally, 3-Tesla (3T) erMRI may be beneficial by providing higher signal, thus further improving spatial (or temporal, in the case of DCE-MRI) resolution. One group of researchers have shown that 3T erMRI imaging is accurate for staging of prostate cancer, that there is moderate to substantial interobserver agreement, and that minimal capsular invasion could be detected. However, there are insufficient data in the literature to support the routine use of 3T erMRI.

T2-Weighted MRI

Over 15 years of clinical experience exists with T2- weighted erMRI. Improvements in coil design (dual endorectal coil and torso coil arrays), pulse sequences, and image correction have led to improvements in the performance of T2-weighted imaging, but some inherent limitations remain. Low-signal lesions on T2-weighted imaging can be due to cancer or can be caused by benign processes such as prostatitis. Endorectal coil MRI remains limited in its ability to identify microscopic or early macroscopic capsular penetration due to restrictions on spatial resolution and motion artifacts. Moreover, individual radiologist expertise is an important determinant of staging accuracy. In one study, one reader achieved an accuracy of 91%, while the other had an accuracy of only 56%.

Early studies from the 1990s reported accuracies from 51%-82% in distinguishing T2 and T3 disease. More recently, erMRI has been shown to improve the prediction of neurovascular bundle invasion prior to radical prostatectomy. One study demonstrated that the differences between "expert" readers and less experienced readers could be reduced by incorporating other clinical data (e.g., PSA value, tumor grade) and using strict imaging criteria. Another study showed that using dynamic contrast-enhanced MRI rather than T2-weighted images can improve staging performance by less experienced readers, when compared to more experienced readers. Endorectal MRI has also been shown to be accurate in demonstrating seminal vesicle invasion. The combination of a tumor at the base of the prostate that extends beyond the capsule combined with low signal in the seminal vesicles that have lost normal architecture is highly predictive of seminal vesicle invasion.

More recently, similar strategies to include erMRI in a neural network have resulted in overall accuracies of 88% to 91% depending on the exact implementation. These results are superior to conventional results with Partin's tables. In this study, Gleason score was the most influential predictive factor, followed by erMRI results and then PSA levels. Several studies have documented that erMRI is most successful in men with intermediate-risk prostate cancer based on Partin's tables. In these men, erMRI staging was highly predictive of PSA recurrence. In a study involving 344 patients, one group of investigators demonstrated that erMRI added statistically meaningful staging data regarding extracapsular extension. Endorectal MRI has also proven helpful in directing 3D conformal radiotherapy and improving outcomes.

MR Spectroscopy

One group of investigators have demonstrated that prostate cancers have a characteristic loss of the citrate peak and gain in the choline/creatine peak on MR spectroscopic imaging. Moreover, the ratio of choline to citrate is related to the Gleason score, suggesting that MRSI may provide information about tumor aggressiveness. Improvements in diagnostic accuracy and staging have been reported. However, MRSI is technically demanding and time consuming. It has not been proven in multi-institutional trials, although a clinical trial under the auspices of the American College of Radiology Imaging Network (ACRIN®) is currently underway. Thus, MRSI cannot yet be considered a routine diagnostic tool.

Dynamic Contrast-Enhanced MRI

Prostate cancers, like many tumors, demonstrate angiogenesis that can be detected on dynamic contrast-enhanced MRI (DCE-MRI). DCE-MRI demonstrates earlier and more intense enhancement in sites of tumor. One group of investigators found minimal improvements in diagnostic accuracy over conventional T2-weighted scans using DCE-MRI. Another group showed that tumors could be distinguished from noncancerous prostate with high reliability, although the study did not specifically address staging. It has also been shown that DCE-MRI can improve staging performance when used in conjunction with T2-weighted images for less experienced readers when compared to more experienced readers. However, this method still suffers from a lack of a uniformly accepted analytic method and has not been tested in multi-institutional trials. Thus, it is still of unproven benefit.

Nodal Staging with MRI

MRI has been shown to be at least equivalent to computed tomography (CT) for detecting abnormal lymph nodes in men with prostate cancer. Neither MRI nor CT scans are as accurate as laparoscopic node dissection. Unfortunately, metastatic lymph nodes in prostate cancer are often small, so that conventional size criteria underestimate the extent of nodal disease. Thus, low sensitivities are observed, even in high-risk patients. Ultrasmall particles of iron oxide (USPIO) have been shown to dramatically improve sensitivity of MRI for nodal metastasis; however, the iron-based contrast agent ferumoxytran (trade name Combidex) is not yet approved by the FDA. The role of MRI for nodal staging will need to be reassessed if the FDA approves Combidex.

Computed Tomography

CT of the abdomen and pelvis is occasionally used to preoperatively stage prostate cancer, but its staging accuracy is usually considered poor. CT scans have suffered from poor sensitivity in detecting capsular penetration, seminal vesicle involvement, and lymph node extension and should be reserved for use in patients with a high probability of lymph node involvement. Overall accuracy in staging was reported as 65% by one group of investigators and as 67% by another group. For loco-regional staging, such as extracapsular penetration, the accuracy has been reported as low as 24%. Even with refined techniques in performing CT (3 mm slice thickness and 5 mm table increments with both IV and oral contrast), it has been concluded that CT is of little value in staging the local extent of prostatic carcinoma. However, one study reports 93.7% accuracy for CT in detecting positive lymph nodes, which increases to 96.5% if CT-guided fine-needle aspiration biopsy is added. This degree of accuracy was only achieved by using a threshold of 6 mm or larger as pathologic. Thus, CT of the abdomen and pelvis is of limited value in local staging and nodal staging and should be reserved for intermediate- and high-risk patients.

ProstaScint (Indium Capromab)

The reliability and usefulness of ProstaScint scan based on indium-111 radiolabeled capromab pendetide (a first-generation monoclonal antibody against prostate-specific membrane antigen [PSMA]) as a method to help initial staging in prostate cancer remain unproven at this time. Initial studies suggested that this technology may improve the detection of metastatic lymph nodes when applied to patients estimated to have a risk of lymph node involvement of >20%. Studies are needed with a sufficient number of patients with histopathological correlation to document sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. One group of investigators conducted histopathological correlation in lymph nodes after ProstaScint scan in 31 patients (43 samples). The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy value were 94%, 42%, 53%, 92%, and 65%, respectively. Its limitations appear to be due to the intracellular binding site of the antibody as well as nonprostatic expression of PSMA. Routine ProstaScint scanning as an initial staging procedure is not justified based on evidence at this time. However, many studies show its utility in postoperative failure settings, especially to guide radio therapeutic decisions. New methods to suppress normal uptake as well as coregistration and fusion with CT or MRI seem to improve its utility in defining target volumes in radio therapeutic settings.

Bone Scan

The radionuclide bone scan is a standard component of the evaluation for many patients diagnosed with prostate cancer. However, original work by one group of investigators has shown that in patients with low PSA level (<10 ng/mL) who have no pain, the yield of a staging bone scan is too low to warrant its routine use. In their experience, no patient with a PSA ≤ 10/ng/mL had a positive bone scan, and only one patient in 300 with a PSA level ≥20 ng/mL had a positive radionuclide bone scan. Such observations have been confirmed by more recent studies as well. These studies suggest that for patients with no skeletal symptoms and a serum PSA level of 10 ng/mL or less, a staging radionuclide bone scan is not necessary; however, this recommendation has to be modified under specific circumstances such as T3 or T4 disease or a high Gleason score.

The rate of positive bone scans depends on the PSA value and Gleason score. Patients with PSA ≤20 ng/mL and Gleason Score <8 have a 1% to 13% rate of positive bone scans. For this reason patients with a PSA ≥20 ng/mL (with any T stage or Gleason score), locally advanced disease (T3 or T4 with any PSA or Gleason score), or Gleason score of 8 or greater (with any PSA or T stage) should be considered for a radionuclide bone scan. Patients with skeletal symptoms or advanced stage disease should also be considered candidates for bone scans.

Positron Emission Tomography

The role of positron emission tomography (PET) in the staging workup of newly diagnosed and recurrent prostate cancers is still being evaluated. It has the potential to play an important role in detecting early metastatic spread and monitoring post-therapy response. The most commonly available PET tracer, FDG-PET, has proven disappointing in the initial staging of prostate cancer. In that study 23 of 24 primary prostate cancer lesions were not detected by FDG-PET. FDG-PET can play a role in the detection of local recurrence and/or distant metastases with increasing PSA after initial treatment failure. Several additional radiotracers have been extensively studied, including C11 or F18 choline and acetate, C11 methionine, F18 fluoride, and fluorodihydrotestosterone. PET scans using these radiotracers may help in the clinical decision-making process, especially in patients with high-risk primary disease, but these agents are not yet widely available. For instance, in the detection of nodal metastases, C11 choline or acetate PET appears to be promising.

New agents such as fluorodihydrotestosterone and gallium-68-labeled peptides are being studied, and these approaches using small chelator-coupled peptides can have advantages over other traditional agents. These tracers remain experimental. Thus, PET scanning has a limited role in the staging of prostate cancer.

Chest Radiography

There are no data in the literature documenting the yield of a chest x-ray. Therefore, it should be performed as part of the initial staging only with suspected metastatic disease (e.g., PSA >100 ng/mL) or in patients who are heavy smokers with clinically localized disease.

Summary

In summary, the guidelines for pretreatment staging of prostate cancer should be individualized based on consideration of the clinical parameters that are predictive of the likelihood of ECE, SV+ and LN+. These clinical parameters should include: the pretreatment PSA level and the rate of rise or doubling time, the Gleason score, the palpation T stage, the number of positive biopsies, and the percentage of the specimen involved.

The role of imaging in low-risk patients is controversial. In intermediate- and high-risk individuals, imaging may play a role in staging and thus in directing therapy. MRI using endorectal coil techniques appears to be the most accurate imaging test available for local staging of the prostate, providing both loco-regional and nodal evaluation. The accuracy of the technique appears related to the experience of the radiologists. MR spectroscopy and dynamic contrast-enhanced MRI may be useful adjuncts in the future but are, as yet, unproven in multiinstitutional trials.

In truly high-risk patients (clinical T3, very high PSA levels, and Gleason score ≥8), radionuclide bone scans and CT may be useful for detecting bony metastases and lymph nodes, respectively. ProstaScint scans may also play a role in detecting nodal metastases in selected high-risk patients, but the modest accuracy of this test has led most experts to consider its value dubious. PET scans with FDG are of limited value in initial staging but may be more useful in recurrent and metastatic disease.

Anticipated Exceptions

Personal Digital Assistant (PDA) Downloads

The following are available:

The National Guideline Clearinghouse™ (NGC) does not develop, produce, approve, or endorse the guidelines represented on this site.

All guidelines summarized by NGC and hosted on our site are produced under the auspices of medical specialty societies, relevant professional associations, public or private organizations, other government agencies, health care organizations or plans, and similar entities.