ml

Influence of Prostate Volume and Percent Free Prostate Specific Antigen on Prostate Cancer Detection in Men With a Total Prostate Specific Antigen of 2.6 to 10.0 ng/ml J. James Bruno, II,* Noel A. Armenakas and John A. Fracchia From the Department of Surgery, Division of Urology, Lenox Hill Hospital, New York, New York

Purpose: Percent free prostate specific antigen and prostate specific antigen density have been independently shown to increase the specificity of prostate cancer screening in men with prostate specific antigen levels between 4.1 and 10.0 ng/ml. Recent data suggest the total prostate specific antigen cutoff for performing a biopsy should be 2.6 ng/ml. We assessed the influence of percent free prostate specific antigen and prostate volume on cancer detection in men with a prostate specific antigen between 2.6 and 10.0 ng/ml. Materials and Methods: From 1991 to 2005 all transrectal ultrasound guided prostate biopsies (5,587) for abnormal digital rectal examination and/or increased age specific prostate specific antigen were evaluated. A total of 1,072 patients with a prostate specific antigen between 2.6 and 10.0 ng/ml and any percent free prostate specific antigen were included in study. The cancer detection rate was calculated for each percent free prostate specific antigen/volume stratum. Results: Prostate cancer was detected in 296 patients (27.6%). The mean age and prostate specific antigen of the patients with benign pathology and prostate cancer were similar. Mean percent free prostate specific antigen was 17.5% and 14.1% (p ⬎0.05), and the mean volume was 62.0 and 46.0 cc (p ⫽ 0.001), respectively. The strongest risk factors for a positive biopsy were percent free prostate specific antigen (odds ratio 0.004, p ⬍0.001), volume (OR 0.977, p ⬍0.001) and digital rectal examination (OR 1.765, p ⫽ 0.007), but not total prostate specific antigen (p ⫽ 0.303). When stratified by volume and percent free prostate specific antigen, distinct risk groups were identified. The probability of detecting cancer inversely correlated with prostate volume and percent free prostate specific antigen. Conclusions: In men with prostate specific antigen levels between 2.6 and 10.0 ng/ml, the probability of detecting cancer was inversely proportional to prostate volume and percent free prostate specific antigen. This table may assist in predicting patient risk for harboring prostate cancer. Key Words: prostate, prostatic neoplasms, prostatic hyperplasia, prostate-specific antigen

he use of serum prostate specific antigen as a screening modality for prostate cancer has become a widely used marker. The accuracy of PSA in detecting clinically significant cancer is limited due to its biological elevation in benign and malignant processes. As such, refinements in PSA interpretation have been used to enhance its precision. PSA density and %fPSA are such refinements which have independently demonstrated an improvement in the specificity of prostate cancer screening.1–3 In addition, these 2 parameters have been analyzed together as covariables to further augment prostate cancer detection while decreasing unnecessary biopsies in men with serum PSA between 4.1 and 10.0 ng/ml.4,5 Recently studies have defined the prevalence of prostate cancer in men with a PSA less than 4.0 ng/ml and recommended a new cutoff of 2.6 ng/ml that would detect more organ confined tumors.6,7 The proposed lowering of total

T

Submitted for publication August 28, 2006. * Correspondence: Department of Surgery, Division of Urology, Lenox Hill Hospital, 100 East 77th St., New York, New York 10021 (telephone: 212-434-2691; FAX: 212-434-3495; e-mail: jamie_bruno@ hotmail.com).

See Editorial on page 1607.

0022-5347/07/1775-1741/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION

PSA has the potential to significantly increase the number of unnecessary biopsies performed. Others have examined the usefulness of using %fPSA measurements in this cohort to further enhance specificity.8 This protocol has also been used in conjunction with age stratification in men with a total PSA between 2 and 4 ng/ml.9 As an extension of these studies our goal was to determine the impact of %fPSA and prostate volume on refining prostate cancer detection in men with a PSA between 2.6 and 10.0 ng/ml.

MATERIALS AND METHODS From December 1991 through June 2005 we evaluated our institutional data set of 5,587 patients who had undergone transrectal ultrasound guided prostate biopsy. The indications for prostatic biopsy were abnormal DRE and/or increased age specific PSA. Patients with a total PSA between 2.6 and 10.0 ng/ml were included into the study. All patients without a %fPSA measurement obtained were excluded yielding 1,072 patients comprising the study group (19%). A total of 988 patients in the study had a total PSA greater than 4.0 ng/ml whereas 84 patients (7.8%) had a total PSA between 2.6 and 4.0 ng/ml. %fPSA—as calculated by the ratio of free PSA-to-total PSA, multiplied by 100 —was

1741

Vol. 177, 1741-1744, May 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.01.067

1742

PROSTATE VOLUME AND PROSTATE CANCER DETECTION

TABLE 1. Clinical characteristics of the 1,072 study patients

No. pts (%) Mean age (⫾SD) Mean ng/ml total PSA (⫾SD) Mean %fPSA (⫾SD) Mean cc prostate vol (⫾SD) No. pos DRE % Positive DRE (⫾SD)

Benign Pathology

Ca

776 (72.4) 64.6 (8.5) 6.28 (1.75) 17.5 (7.7) 62.0 (30.0) 75 9.7 (0.3)

296 (27.6) 66.3 (9.2) 6.38 (1.76) 14.1 (6.5) 46.0 (26.3) 50 16.9 (0.4)

p Value 0.1 0.4 0.2 0.001 0.001

categorized as low (%fPSA less than 15%), intermediate (%fPSA 15% to 24%) and high (%fPSA 25% or greater). All prostate volumes were calculated by TRUS using axial, transverse and anteroposterior dimension measurements (␲/6 ⫻ d1 ⫻ d2 ⫻ d3). Prostate volume was stratified into quartiles of less than 35.0, 35.0 to 49.9, 50.0 to 64.9 and 65.0 cc or greater. Systematic 6 to 12 core biopsies were then performed under TRUS guidance using a standard 18 gauge needle biopsy gun. Histological evaluation divided each patient prostate specimen into either a noncancerous (benign) or cancerous pathological category. The cancer detection rate— defined as percentage of patient with histologically confirmed prostate cancer on biopsy divided by total patients for each subgroup—was determined for each %fPSA/volume stratum. Student’s t test was used to statistically compare continuous variables. The variables of total PSA, %fPSA, DRE and prostate volume were further analyzed using univariate and multivariate logistic regression models to predict the presence of cancer. All differences with a p value less than 0.05 were considered statistically significant. All statistical analyses were performed using commercially available SPSS® 10.0 software. RESULTS Of the 1,072 men included in the study, prostate cancer was detected in 27.6% of the patients and benign pathology in 72.4% (table 1). The mean age, mean total serum PSA and mean %fPSA of the patients in each category were not statistically different. However, the mean prostate volume of the benign group was significantly larger (62.0 cc) than the cancer group (46.0 cc) (p ⫽ 0.001). In addition, an abnormal DRE was noted in 16.9% of the men in the prostate cancer group compared to 9.7% in the benign group (p ⫽ 0.001). On univariate logistic regression all variables were significantly predictive of the presence of cancer. However, on the multivariate model total PSA was not a significant independent risk factor (p ⫽ 0.303). The multivariate analysis demonstrated the strongest risk factors for a positive biopsy were %fPSA (odds ratio 0.004, p ⬍0.001), prostate volume (OR 0.977, p ⬍0.001) and DRE (OR 1.765, p ⫽ 0.007).

When stratified by prostate volume quartiles and the 3 %fPSA categories, distinct risk groups were identified (table 2). The probability of detecting cancer inversely correlated with prostate volume and %fPSA. Significantly this association was noted for each %fPSA strata with a cancer risk increase from 17.6% to 54.8% (low), 10.1% to 42.2% (intermediate) and 7.4% to 42.8% (high) as volume decreased. Therefore, as prostate volume increased the likelihood of detecting cancer steadily decreased. This general trend was also demonstrated as %fPSA increased. DISCUSSION The overall aim in prostate cancer screening is to identify curable disease while minimizing unnecessary transrectal biopsies. DRE suffers not only from its failure to accurately discriminate between benign and malignant pathology but also in its inability to identify nonpalpable but curable disease. The advent of PSA has ushered in a new era in screening— providing a simple blood test to enhance the detection of prostate cancer compared to DRE alone. Despite this development, the use of PSA has been shown to lack precision due to its elevation in benign and malignant processes. Therefore, numerous studies have attempted to use PSA in conjunction with other parameters to refine its sensitivity. In addition to initially determining a PSA range of greater than 4.0 ng/ml as having a predictive value in diagnosing prostate cancer10 other modalities have been analyzed including age specific PSA,11 PSA density1 and PSA velocity.12,13 In 1992 Benson et al introduced the concept of PSA density as a method of improving PSA sensitivity.1 In evaluating 127 men, this analysis proposed that although a cohort of men may have a similar PSA, by stratifying this value with prostate volume relative levels were obtained which correlated to benign and malignant processes. This stratification, in turn, could then be tailored to each individual in the decision to perform a biopsy. The usefulness of PSA density has been questioned, but it appears to significantly decrease the unnecessary biopsy rate without significantly sacrificing prostate cancer detection. In 1995 %fPSA was reported by Catalona et al as another modality in refining prostate cancer screening.2 Their analysis of 113 patients demonstrated that by using a %fPSA cutoff of 23.4% or less, 90% of cancers would be detected and nearly a third of unnecessary biopsies would be avoided. The same group supplemented this data in 1998 with a study of 773 men. They found that a %fPSA cutoff of 25% or less would detect 95% of cancers while avoiding 20% of unnecessary biopsies.3 Despite the success of these improvements, the majority of data fails to incorporate total PSA levels less than 4.0 ng/ml. Contemporary evidence indicates that a PSA cutoff of 2.6 ng/ml would detect more organ confined tumors at an

TABLE 2. Percentage of positive biopsies stratified by prostate volume and %fPSA Prostate Vol (cc) Less than 35.0 35.0–49.9 50.0–64.9 65.0 or Greater All vol

%fPSA Less Than 15% (low)

%fPSA 15%–24% (intermediate)

%fPSA 25% or Greater (high)

All %fPSA

54.8 33.8 26.8 17.6 37.8

42.2 28.4 19.5 10.1 20.7

42.8 30.0 23.8 7.4 17.6

51.3 31.3 22.5 11.1 27.6

PROSTATE VOLUME AND PROSTATE CANCER DETECTION acceptable cost.6-8 With these diagnostic recommendations maturing, the objective of our study was to evaluate prostate volume in conjunction with %fPSA as a means to augment the armamentarium of prostate cancer screening. In our cohort of 1,072 patients we found that there was no difference in age or total serum PSA between the benign and cancerous groups. Although the 2 groups had no statistical difference in their relative %fPSA levels there was a trend for the benign group to have a higher level, as expected from the literature. In terms of prostate size the benign category of patients showed a much larger gland on average than the group with cancer. The choice to use prostate size alone as a variable, rather than PSA density, was made to avoid confounding this parameter with total PSA. Finally, the DRE was included in this analysis to determine its relative contribution to overall prostate cancer detection in relation to the other variables. Again, as expected, the patients with cancer were nearly twice as likely to have an abnormal DRE as those with benign pathology. Using logistic regressions models, the total PSA was not a significant risk factor for prostate cancer. Indeed the strongest predictors were DRE, %fPSA and prostate volume. Although the %fPSA means were not different between the 2 groups on t test, their relative contribution as a negative risk factor (as %fPSA increases) on multivariate analysis was significant. Prostate volume too was shown to be a negative risk factor, albeit a modest one. Furthering our analysis, with total PSA statistically removed as a significant risk factor and DRE clinically removed due to its independent nature in prostate cancer screening, the variables of %fPSA and prostate volume were stratified into risk categories. The results are similar to the scheme proposed by Haese et al, who in 2004 stratified risk by age and %fPSA for those men with a total PSA between 2 and 4 ng/ml.9 In our study prostate volume and %fPSA proved to be independent negative risk factors which inversely correlated with the probability of detecting cancer. The inverse association was noted within each %fPSA strata with a cancer risk increase from 17.6% to 54.8% (low), 10.1% to 42.2% (intermediate) and 7.4% to 42.8% (high) as volume decreased. This stratified table has significant clinical implications. For instance, a patient with a %fPSA in the intermediate range (15% to 24%) and a large prostate (65.0 cc or greater) only has a 10% chance of harboring prostate cancer compared to a similar patient with a small prostate (less than 35.0 cc) yielding a greater than 40% chance of having cancer. Using these 2 parameters specific risk strata can be determined for each individual patient and used in counseling when determining the need for prostate biopsy. Due to the multiple variables used in this table, a simple sensitivity and specificity analysis was not possible. Our goal was not to assess the individual sensitivity, specificity, positive/negative predictive values and hazard ratio for PSA, %fPSA or volume in relation to pathological outcome. Rather, we evaluated the combined effect of these variables on prostate cancer detection which precluded this type of analysis. Despite these results and the usefulness of this table, this study has several limitations. Including the DRE into this analysis appears counterintuitive. Those patients with a positive DRE would naturally undergo counseling and likely a biopsy regardless of their PSA, %fPSA or prostate volume. Indeed, the majority of studies in this genre typically

1743

exclude those men with a positive DRE. Nevertheless, we initially included the DRE as a variable to evaluate its relative impact on risk and attempted to minimize its effect based on multivariate logistic regression. In addition, the calculation of prostate volume logically implies that TRUS was performed. The invasiveness of this procedure would limit its use as exclusively an imaging modality for calculating volume without obtaining tissue for a pathological diagnosis. This limitation may be averted by using suprapubic ultrasonography which has been shown to have a correlation of nearly 90% with TRUS measurements.14 We also did not solely investigate those patients with PSA levels between 2.6 and 4.0 ng/ml. The small number of patients (84) with PSA less than 4.0 ng/ml would not have allowed for meaningful conclusions. However, since total PSA proved not to be a significant variable in this analysis, and the clinical characteristics (age, %fPSA, volume, DRE) were not different between men with PSA greater than or less than 4.0 ng/ml, the results can be used throughout this entire PSA range. Finally, we recognize that refining the present variables with subset analysis (age strata, year of biopsy, number of cores per biopsy session) and evaluating additional variable (ethnicity, body mass index) would further enhance the strength of the pre-biopsy cancer prediction table. Overall, the combined application of prostate volume and %fPSA yields a modality that can enhance prostate cancer screening compared to total PSA and DRE alone. This further refinement may reduce the number of unnecessary TRUS biopsies performed without sacrificing prostate cancer detection. In the future these variables (PSA, %fPSA, prostate volume) may be used in conjunction with others (age, race, PSA velocity, etc) to create a comprehensive algorithm or nomogram that may provide the most accurate means of determining prostate cancer risk. CONCLUSIONS In patients with a total PSA between 2.6 and 10.0 ng/ml, the probability of detecting prostate cancer is inversely proportional to prostate volume and %fPSA. The stratified table derived from these 2 parameters may assist physicians in predicting individual patient risk of having prostate cancer and, consequently, the need for biopsy.

Abbreviations and Acronyms DRE %fPSA PSA TRUS

⫽ ⫽ ⫽ ⫽

digital rectal examination percent free prostate specific antigen prostate specific antigen transrectal ultrasound

REFERENCES 1.

2.

3.

Benson MC, Whang IS, Pantuck A, Ring K, Kaplan SA, Olsson CA et al: Prostate specific antigen density: a means of distinguishing benign prostatic hypertrophy and prostate cancer. J Urol 1992; 147: 815. Catalona WJ, Smith DS, Wolfert RL, Wang TJ, Rittenhouse HG, Ratliff TL et al: Evaluation of percentage of free serum prostate-specific antigen to improve specificity of prostate cancer screening. JAMA 1995; 274: 1214. Catalona WJ, Partin AW, Slawin KM, Brawer MK, Flanigan RC, Patel A et al: Use of the percentage of free prostate-

1744

PROSTATE VOLUME AND PROSTATE CANCER DETECTION

specific antigen to enhance differentiation of prostate cancer from benign prostatic disease: a prospective multicenter clinical trial. JAMA 1998; 279: 1542. 4. Morote J, Raventos CX, Lorente JA, Lopez-Pacios MA, Encabo G, de Torres I et al: Comparison of percent free prostate specific antigen and prostate specific antigen density as methods to enhance prostate specific antigen specificity in early prostate cancer detection in men with normal rectal examination and prostate specific antigen between 4.1 and 10 ng./ml. J Urol 1997; 158: 502. 5. Catalona WJ, Beiser JA and Smith DS: Serum free prostate specific antigen and prostate specific antigen density measurements for predicting cancer in men with prior negative prostatic biopsies. J Urol 1997; 158: 2162. 6. Krumholtz JS, Carvalhal GF, Ramos CG, Smith DS, Thorson P, Yan Y et al: Prostate-specific antigen cutoff of 2.6 ng/mL for prostate cancer screening is associated with favorable pathologic tumor features. Urology 2002; 60: 469. 7. Thompson IM, Pauler DK, Goodman PJ, Tangen CM, Lucia MS, Parnes HL et al: Prevalence of prostate cancer among men with a prostate-specific antigen level ⬍ or ⫽4.0 ng per milliliter. N Engl J Med 2004; 350: 2239. 8. Catalona WJ, Smith DS and Ornstein DK: Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/ml and benign prostate examination: enhancement

9.

10.

11.

12.

13.

14.

of specificity with free PSA measurements. JAMA 1997; 277: 1452. Haese A, Chun K-HF, Graefen M, Huland E and Huland H: Age- and percent free PSA stratified risk analysis for prostate cancer detection in men with a total serum PSA 2– 4 ng/ml and benign digital rectal examination. J Urol, suppl., 2004; 171: 332, abstract 1260. Cooner WH, Mosley BR, Rutherford CL Jr, Beard JH, Pond HS, Terry WJ et al: Prostate cancer detection in a clinical urological practice by ultrasonography, digital rectal examination and prostate specific antigen. J Urol 1990; 143: 1146. Oesterling JE, Jacobsen SJ, Chute CG, Guess HA, Girman CJ, Panser LA et al: Serum prostate-specific antigen in a community-based population of healthy men. Establishment of age-specific reference ranges. JAMA 1993; 270: 860. Carter HB, Pearson JD, Waclawiw Z, Metter EJ, Chan DW, Guess HA et al: Prostate-specific antigen variability in men without prostate cancer: effect of sampling interval on prostate-specific antigen velocity. Urology 1995; 45: 591. Smith DS and Catalona WJ: Rate of change in serum prostate specific antigen levels as a method for prostate cancer detection. J Urol 1994; 152: 1163. Isom-Batz G, Basillote J, Pareek G, Rivera AC, Armenakas NA and Fracchia JA: Size matters! Simple, reliable assessment of prostate volume. J Urol, suppl., 2003; 169: 461, abstract 1729.