Guidelines for Radiographic Studies in Prostate Cancer Questioned

We read with interest the recent practice guidelines on prostate cancerpublished by the National Comprehensive Cancer Network (NCCN) in a supplementto the November 1996 issue of ONCOLOGY (pp 265-288). The establishmentof such guidelines is an important step in moving clinical practice inthe direction of "evidence-based medicine." Due to increasingconcerns regarding containment of medical costs and a desire to providethe best care to patients, it is becoming increasingly important to baseclinical decision-making on rigorous analyses of available data.

Although we agree with much of the management algorithm for prostatecancer outlined by the NCCN Prostate Cancer Practice Guidelines Panel,we would like to comment further on the use of abdominal/pelvic CT or MRIin prostate cancer staging. We agree with the panel's statement that "computedtomography and magnetic resonance imaging of the abdomen and pelvis haveproven to be of little value in the detection of pelvic lymph node metastasisand/or seminal vesicle invasion." The authors also state that suchstudies should be used in patients with T3 or T4 tumors. They donot cite any literature to support the latter statement. It is our viewthat T-stage alone may not be the best criterion for selecting prostatecancer patients for staging abdominal/pelvic CT/MRI.

Two Recent Studies

We recently completed two studies that examined the ability of serumprostate-specific antigen (PSA) to predict the results of staging abdominal/pelvicCT/MRI.[1,2] In our most recent analysis,[2] 425 newly diagnosed prostatecancer patients had a mean serum PSA of 22.1 ng/mL. Approximately 28% ofthe group had T3 or T4 tumors on digital rectal examination. A total of14 patients (3.6%) presented with a positive abdominal/pelvic CT/MRI (12with adenopathy, 1 with a renal cell tumor, and 1 with an adrenal metastasis).Of these 14 patients, 11 (79%) had a serum PSA of 30 mg/mLor more (range,30.0 to 234 ng/mL). Not all patients had T3 or T4 tumors. One-quarter toone-third of patients had disease confined to the prostate on examination.In fact, of three patients with positive scans and serum PSAs less than20 ng/mL, two had T2 tumors. Overall, fewer than 1.0% of patients witha serum PSA less than 20 ng/mL had a positive scan.

At present, very few data are available on which to base guidelinesfor the use of staging abdominal/pelvic imaging in patients with newlydiagnosed prostate cancer. Although our data are quite limited, we feelthat serum PSA, possibly in conjunction with other parameters, may proveto be a better guide for selecting patients for radiographic staging thanis T-stage alone.

A large percentage of patients with newly diagnosed prostate cancercontinue to undergo staging abdominal/pelvic imaging.[3] Such studies arenot only time consuming but also costly (more than $1,000 per scan). Westrongly believe that much further investigation is needed to more clearlydefine the appropriate indications for these studies.

We understand that the practice guidelines outlined by the authors arein a constant state of revision. Nonetheless, we would not consider theuse of these radiographic staging studies in this clinical context as "uncontestedand generally accepted by all authorities," ie, a level 1 recommendation.Rather, given the lack of hard supportive data, we would characterize thisquestion as one requiring further analysis for firm recommendations (ie,a level 3 recommendation).

Michael Huncharek, MD, MPH
Assistant Medical Director,Metaworks Inc, and Adjunct Assistant Professor,Boston University School of Public Health

Joshua Muscat, MPH
Senior Epidemiologist, Division of Epidemiology, American Health Foundation,New York

The NCCN Panel Chairman Responds

The NCCN Prostate Cancer Practice Guidelines Panel agrees that emergingdata may better define how to select patients for further staging studiesbased on defined risk for identifiable metastasis. Nomograms may becomenecessary for this, rather than just clinical stage, PSA, or tumor grade.In a retrospective study published in Urologic Oncology, Konety et al concludethat CT scanning has minimal to no utility in detecting extraprostaticdisease in patients with clinically localized prostate cancer.[1]

Laurence H. Baker, DO
Chairman, NCCN Prostate Cancer Practice Guidelines Panel, Professor ofMedicine, Director for Clinical Research, University of Michigan ComprehensiveCancer Center, Ann Arbor

References:

1. Huncharek M, Muscat J: Serum prostate specific antigen as a predictorof radiographic staging studies in newly diagnosed prostate cancer. CancerInvest 13(1):31-35, 1995.

2. Huncharek M, Muscat J: Serum prostate specific antigen as a predictorof staging abdominal/pelvic computed tomography in newly diagnosed prostatecancer. Abdom Imaging 21:364-367, 1996.

3. Mettlin C, Jones GW, Murphy GP: Trends in prostate cancer care inthe United States, 1974-1990: Observations from the patient care evaluationstudies of the American College of Surgeons Commission on Cancer. CA CancerJ Clin 43:83-91, 1993.

1. Konety BR, Naraghi R, Wooding W, et al: Evaluation of computerizedtomography for staging of clinically localized adenocarcinoma of the prostate.Urol Oncol 2:14-19, 1996