Beyond the Abstract - Usefulness and predictive value of PSA density, adjusted by transition zone volume, in men with PSA levels between 2 and 4 ng/ml, by Prof. Abdelatif Janane, MD

BERKELEY, CA (UroToday.com) - The cut-off level of prostate-specific antigen (PSA) at 4.0 ng/ml has been and continues to be the most important and widely used value in the screening, detection, and monitoring of prostate cancer (PCa).

However, Partin et al. demonstrated an unexpectedly high percentage of cases with extracapsular disease even if PSA levels were between 2 and 4.0 ng/ml.^[1] Several studies showed relatively high detection rates of prostate cancer in men with PSA levels between 2.0 and 4.0 ng/ml, which remained curable with high probability.^[2-6] A previous study also demonstrated that 65% of cases were multi-focal and 34% of cases already had a high malignant potential in prostate cancer cases detected at PSA levels of 4.0 ng/ml or lower.^[7]

Owing to the limited sensitivity and specificity of PSA, various approaches such as percent free PSA, PSA velocity, and PSA density have been used to improve the diagnostic validity of an elevated PSA level.^[3,8] On the other hand, the percentage of men with PSA levels greater than 2.0 ng/ml was high at about 32% in PSA-based screening for prostate cancer.^[9] The percentage of men with PSA levels greater than 2.0 ng/ml was higher than that of subjects who died from prostate cancer. Therefore, prostate biopsy should not be performed for all men with PSA levels between 2.0 and 4.0 ng/ml. Catalona et al. demon¬strated that the use of the free/ total PSA ratio (fit PSA) could avoid unnecessary biopsies for subjects with PSA levels between 2.0 and 4.0 ng/ml. However, further studies would be necessary to improve the diagnostic accuracy of prostate cancer in the PSA ranges between 2.0 and 4.0 ng/ml. In the present study, we investigated the usefulness of prostate volume-related PSA parameters, PSA density (PSAD), and PSAD adjusted by transition zone volume (PSADTZ) to address this issue.

Between June 2003 and March 2010, 839 men underwent serum PSA measurements, digital rectal examination (DRE), and transrectal ultrasonography (TRUS) in mass screening for prostate cancer in the out-patient clinic of a military University Hospital. A total of 134 men (16.3%) had PSA levels between 2.0 and 4.0 ng/ml. Of these men, 69 men who underwent prostate biopsy for their abnormal findings on DRE/TRUS, or for their exclusion of prostate cancer before radical cystectomy for bladder cancer, or transurethral resection of the prostate for clinical diagnosis of benign prostatic hyperplasia, were enrolled in this study. All patients were biopsied for the first time. There was no case with a present or past history of detecting prostate cancer or treatment for benign prostatic hyperplasia (BPH). Of these 134 patients, 34 men (25,1%) had abnormal findings on DRE or TRUS, and 100 (74,9 %) did not have any suspicious findings for cancer at both examinations (Table 1). Their ages ranged from 48 to 76 years old. At the actual screening, serum PSA levels were measured using a Tosoh II PSA assay. All PSA values were converted to the values of the Tandem¬ R PSA assay using the following formula; PSA (Tandem-R) = 0.963 x PSA(Tosoh) - 0.123, PSA(Tandem-R) = 3.303 x PSA (Markit) - 0.585 (data not shown). Systematic sextant needle biopsies and two additional transition zone biopsies were performed for all participants. An additional TRUS-guided biopsy was also performed at the hypocycloid region and an additional DRE-guided biopsy was performed at the nodular region. Total prostate volume was calculated by the 3-axis (length x width x height x 0.52) method and transition zone volume was calculated by 2-axis (width x width x height x 0.52) method. All prostate volumes and transition zone volumes were calculated by one urologist (M.O). PSAD and PSADTZ were calculated to divide serum PSA levels by individual total prostate volume and transition zone volume, respectively. All TRUS procedures were performed using the Bruel and Kjaer Model with a biplaner 5 la 7.5 MHz endorectal probe. All biopsies were performed using an 18-gauge core biopsy needle with a spring-loaded biopsy instrument (Bard Urological, Covington, GA, USA). The 1997 VICC TNM classification of prostate carcinoma was used. The Gleason score of all cases with prostate cancer were reviewed again by one urologic pathologist for this study.

We subdivided PSA levels as 2.0 to 3.0 ng/ml and 3.1 to 4.0 ng/ml. Then, we investigated the usefulness of PSAD and PSADTZ as biopsy indications using receiver operating characteristic (ROC) curves. Diagnostic efficiency was calculated using the following formula: diagnostic efficiency (%) = sensitivity (%) x specificity (%)/ 100. Differences were considered significant when p was <0.05 using Mann-Whitney U-test or the x²-test.

The overall detection rate of prostate cancer was high at 23.8% (32/134). The age and clinical findings in all participants and the clinical stage and pathological findings in cases with prostate cancer are shown in Table 1. The detection rate was 23.1% and 25.0% in men with PSA levels of 2.0 to 3.0 ng/ml and 3.1 to 4.0 ng/ml, respectively. The clinical stage was T1c in 24/32 cases (75.0%) and T2a in 8 cases (25.0%). The distribution of the clinical stage was not significantly different between cases with PSA levels between 2.1 and 3.0 ng/ml and those with PSA levels between 3.1 and 4.0ng/ml. Primary Gleason grade 4 or 5, on prostate specimen, was obtained from 26.7% of cases.

Prostate volume, PSAD, and PSADTZ were compared between cases with and without prostate cancer. The mean PSA and PSAD were not significantly different between cases with and without prostate cancer. Alternatively, the total prostate volume, transition zone volume, and PSADTZ were significantly different between these two groups.

Table 2 shows the comparison in the diagnostic reliability between PSAD and PSADTZ. The area under the ROC curve (AUC-ROC) for PSADTZ was significantly higher than that for PSAD in the PSA range of 2.0 to 4.0 ng/ml (p <0.05, /-test). The diagnostic efficiency was highest at the PSAD cut-off of 0.08 and 0.12 ng/ml/cc in men with PSA levels between 2.1 and 3.0 ng/ml and between 3.1 and 4.0 ng/ml, respectively. The diagnostic efficiency of PSADTZ was relatively high in comparison with that of PSAD in the PSA range of 2.1 to 4.0 ng/ml. The diagnostic efficiency was highest at the PSADTZ cut-off of 0.23 and 0.28 ng/ml/cc in men with PSA levels of 2.0 to 3.0 ng/ml and 3.1 to 4.0 ng/ml, respectively. In men with PSA levels of 2.0 to 3.0 ng/ml, the PSAD cut-off of 0.07 ng/ml/cc resulted in 100% sensitively and 47.1% (16/34) specifically. The PSADTZ cut-off of 0.23 ng/ml/cc resulted in 100% sensitivity and a relatively high specificity of 73.5% (27/34). In men with PSA levels between 3.1 and 4.0 ng/ml, the PSAD cut-off of 0.09 ng/ml/cc resulted in 87.5% (7/8) sensitivity and 30.0% (6/20) specificity. The PSADTZ cut-off of 0.28 ng/ml/cc resulted in 87.5% (7/8) sensitivity and a high specificity of 75.0% (15/20).

If the sensitivity was fixed at 100% and 93%, the number of unnecessary biopsies would decrease to 78% (42/54) and 35% (19/54), respectively, in combination with PSATZO cut-offs in men with PSA levels between 2.0 and 4.0 ng/ml. The introduction of PSA measurement into clinical practice has increased the chance of diagnosis of localized PCa. The present dilemma is deciding whether to perform prostate biopsy immediately for all men with PSA levels between 2.0 and 4.0 ng/ml. The prediction of positive biopsy and clinical stage are still far from optimal and require the combination of different PSA isoforms and variables. This multi-parameter approach is essential given that PSA value taken singularly is not sufficiently accurate due to the interference of age and frequently coexisting conditions, such as benign prostate hyperplasia and histological prostatic inflammation in asymptomatic patients.^[1,2] The density parameters, i.e., PSA Density (PSAD) and PSA density of the transition zone (PSADTZ), are ratios obtained dividing PSA level by total prostate and transition prostate volume, the influence related to the nonmalignant portion of the gland, which is believed to account for most of the physiological PSA leakage into the serum, ought to be reduced.^[1,5] Both parameters were first introduced to improve the diagnostic accuracy of PSA, in cancer detection and aggressiveness, with conflicting results. The clinical use of these parameters for both applications has been severely criticized because of the inherent limits of TRUS. In fact, the commonly used ellipsoid formula is not considered to be the optimal method to calculate prostate volume, and the results obtained are not easily reproducible, since volume estimation is operator and experience dependent. Furthermore, most reports have used the receiver operating characteristic curves, while few have examined how PSA and the density variables may additionally relate to positivity of the biopsy and pathological features of the cancer.^[2,6] In the present study, we prospectively compared PSA and the density parameters with the cancer rate and the pathological tumour characteristics by means of logistic regression analysis, in order to assess which is the most accurate predictor and whether the formula used to calculate PSAD and PSADTZ truly reflects the amount of PSA produced by the tumour. Our results reflect that both PSA and the density parameters were significantly related to biopsy positivity, to pT, to pN, and definitive Gleason score on the prostate specimen: an increased production of PSA per gram by higher stage and higher grade, possibly due to progressively greater tumour volume, as reported by several authors.^[3,9]

Concerning particularities of PCa with PSA levels lower than 4 ng/ml, Catalona et al demonstrated that cancer was detected in 22% (73 of 332 biopsies) of men with serum PSA levels between 2.0 and 4.0 ng/ml. Of these detected cancers, the great majority showed features of medically important tumors. Schroder et al. also asserted that the PSA cut-off level of 4.0 ng/ml would miss large numbers of cancers, and about half of these tumors might have aggressive characteristics and may still be organ confined [2, 9]. Furthermore, Ito et al. investigated the correlation between the clinical stage of prostate cancer and years elapsed until prostate cancer was diagnosed after the PSA levels increased above 4.0 ng/ml.^[10] They demonstrated that the risk of T3, NI or MI disease was significantly higher in prostate cancer cases detected with a screening interval of 2 or more years after the PSA levels increased above 4.0 ng/ml than in those detected within one year after PSA levels increased above 4.0 ng/ml.^[10,11] According to the Partin tables, cases with PSA levels between 2.0 and 4.0 ng/ ml, a clinical stage of T2aNOMO and Gleason score (Gleason grade: primary + secondary) in a biopsy specimen of 7 (4 + 3) are expected to show 33% with organ confined disease.^[4,7] Therefore, there may be candidates for prostate biopsy in men with PSA levels between 2.0 and 4.0 ng/ml.

On the other hand, the proportion of the clinical stage in the prostate cancer cases detected in the PSA range of 2.0 to 4.0 ng/ml in the present study was similar to that in prostate cancer cases detected within one year of screening after PSA levels increased above 4.0 ng/ml. Furthermore, the percentage of men with PSA levels greater than 2.0 ng/ml was high at 32% (3,360/ 10,523) in PSA-based screening for prostate cancer in Europe.^[8,12] The percentage of men with PSA levels greater than 2.0 ng/ml may be 8-fold higher than that of patients who died from prostate cancer, which is about 4%.^[13,14] Therefore, prostate biopsy should be performed for all men with PSA levels between 2.1 and 4.0 ng/ml.

At present, prostate biopsy should be recommended for men with PSA levels between 2.1 and 4.0 ng/ml and also with abnormal findings on DRE/ TRUS or highly suspicious findings on PSA-related parameters like fit PSA or PSADTZ. Catalona et al. demonstrated that using the fit PSA cut-off would have detected 90% of cancers, avoided 18% of benign biopsies, and yielded a positive predictive value of 24% in men who underwent biopsy.^[15] Therefore, free serum PSA measurements may be useful to decide whether prostate biopsy should be recommended. On the other hand, we proposed a new screening algorithm, which recommended an extensive use of TRUS for men with PSA levels between 2 and 4 ng/ml and without abnormal findings on digital rectal examination.^[16] This algorithm would improve the diagnostic sensitivity of clinically significant prostate cancer without increasing the cost of screening per prostate cancer diagnosis. Therefore, TRUS could be one of the most useful screening modalities for men with PSA levels between 2.0 and 4.0 ng/ml.

Several previous studies demonstrated the diagnostic significance of PSADTZ to enhance prostate cancer prediction within this intermediate PSA range. Zlotta et al. also demonstrated that the use of PSADTZ might be of additional value for identifying which patients with clinically localized prostate cancer and with PSA levels less than 10.0 ng/ml were at high risk for extracapsular disease.^[15] Recently, Djavan et al. investigated the usefulness of PSADTZ and PSAD in comparison with fit PSA and other PSA-related parameters in men with PSA levels of 2.5 to 4.0 ng/ml.^[16] They demonstrated that fit PSA and PSADTZ clearly out-performed all other PSA-related parameters using AUC-ROC analysis. With a 95% sensitivity for prostate cancer detection, the cut-off for fit PSA was 41%, and that for PSADTZ of 0.095 ng/ml/cc, unnecessary biopsies in 29.3% and 17.2% of cases in the PSA reflex ranges of 2.0 to 4.0 ng/ml, respectively, would be avoided. However, the performance of fit PSA and PSADTZ failed to retain their predictive power among patients with a prostate size smaller than 30 cc., in the present study, the diagnostic reliability of PSADTZ was maintained for men with a relatively low mean prostate volume of 32.3 cc. The number of unnecessary biopsies would decrease to 64.8% (35/54) with a sensitivity of 93.3%, if the PSADTZ cut-off of 0.23 ng/ml/cc was used for men with PSA levels between 2.1 and 4.0 ng/ml.

The innovative aspect of our study is that, surprisingly, albeit the density parameters are PSA ratios adjusted for total prostate and transitional volumes, they both maintained significant relations with most volume parameters. The PSAD and PSADTZ can be used as a strong diagnostic tool for differentiating the PCa from BHP because PCa releases more PSA per volume unit into the circulation than does BPH. However, the mathematical formula for PSAD and PSADTZ is not sufficiently effective in eliminating the contribution of PSA given by the non-malignant portion of the gland, even though the amount produced by cancer continues to exert its effect.

The present study suggests that a clinical use of PSADTZ cut-offs can reduce the number of unnecessary biopsies without missing most prostate cancer cases in the PSA range of 2.0 to 4.0 ng/ml. The discrimination between cases with intra and extra capsular disease may only be made by taking into account the percentage of positive biopsy cores. PSA is responsible for fewer false negatives in staging than the PSADTZ, possibly because the contribution to PSA given by the non-malignant portion of the prostate gland is not eliminated by the mathematical correction of the formula. The principal limit of our series was the number of cases without abnormal findings on both PSA, DRE and TRUS was small in our data, therefore, further studies should be conducted.

Table 1: Clinical findings in All participants and clinico-pathologic features in subjects with prostate cancer

Table 2: Sensitivity, specificity and efficiency relative to various cut-offs on prostate volume related PSA parameters

References

1. Partin AW, Mangold LA, Lamm DM, Walsh PC, Epstein JI, Pearson JD. Contcmporary updatc of prostate cancer staging nomograms (Partin Tables) for the new millennium. Urology 2001; 580: 843-348.
2. Daniel WL. Beyond PSA: utility of novel tumor markers in the setting of elevated PSA. Urol Oncol 2009; 27 (3): 315- 321.
3. Catalona WJ, Partin AW, Finlay JA, Chan DW, Rittcnhousc HG, Wolfcrt RL. et al. Use of percentage of free prostate-specific antigen to identify men al high risk of prostate cancer when PSA levels are 2.51 to 4 ng/ml ,wd digital rectal examination is not suspicious for prostate cancer: an alternative model. Urology 1999; 54: 20-24.
4. Catalona WJ, Smith OS, Ornstein DK. Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/ml and benign prostate examination. Enhancement of specificity with free PSA measurements. JAMA 1997; 277: 1452-1455.
5. Ito K, Yamamoto T, Kubota Y, Suzuki K, Fukabori Y, Kurokawa K, et al. Usefulness of age-specific reference range of prostate-specific antigen for Japanese men older than 60 years in mass screening for prostate cancer. Urology 2000; 56:278-82.
6. Raaijmakers R, Blijenberg H, Rittcnhouse HG, Roobol MA, Finlay JA, Schroder FH. Prostate cancer (PC) detection in the PSA range 2-3 ng/ml-value of PSA, FPSA, HK2 and related algorithms. J Urol 2002; 167(Suppl): 286.
7. Horninger W, Voigger H, Rogatsch H, Oschweendtner A, Bartsch G. Consideration of low PSA cut-off levels to optimize the detection of cumble prostate cancer. Eur Urol 2001; 39(Suppl4):43--6.
8. Matthew B, Gretzer A, Alan W, Partin AW. PSA levels and the probability of prostate cancer on biopsy Euro Urol Supplements 2008; 1(6): 21-27
9. Giannarini G, Scoot CA, Moro U, Pertoldi B, Beltrami CA, Selli C. Are PSA density and PSA density of the transition zone more accurate in predicting the pathological stage of clinically localized prostate cancer. Urol Oncol 2008; 26(4): 353-363.
10. Ito K, Kubota Y, Suzuki K, Shimizu N, Fukabori Y, Kurokawa K, et al. Correlation of prostate-specific antigen before prostate cancer detection and clinicopathologic leatures: evaluation of mass screening populations. Urology 2000; 55:705-709. [
11. Yamamoto T, Ito K, Ohi M, Kubota Y, Suzuki K, Fukabori Y, et al. Diagnostic significance of digital rectal examination and transrectal ultrasonography in men with prostate-specific antigen levels of 4 ng/ ml or less. Urology 2001; 58:994-998.
12. Kandirali E, Boran C, Serin E, Semircioz A, Metin A. Association of extent and aggressiveness of inflammation with serum PSA levels and PSA density . Urology 2007; 70 (4):743-747.
13. Djavan B, Zlolla AR, Remzi M, Ghawidel K, Bursa B, Hruby 5, et al. Total and transition zone prostate volume and age: how do they affect the Utility of PSA-based diagnostic parameters for carly prostate cancer detection? Urology 1999; 54: 846-852.
14. Kikuchi E, Nakashima J, Ishibashi M, Ohigashi T Asakura H, Tachibana M, et al. Prostate specific antigen adjusted for transition zone volume: the most powerful method for detecting prostate carcinoma. Cancer 2000; 89:842-849.
15. Loeb S, Sara N, Catalona WJ. Exclusion of inflammation in the differential diagnosis of an an elevated PSA. Urol Oncol 2009 27 (1): 64-66.
16. Djavan B, Zlolla A, Kratzik C, Remzi M, Seitz C, Schulman CC. Relationship between initial PSA density with future PSA Kinetics and repeat biopsies in men with prostate cancer on active surveillance. J Urol; 2010; 183(4):673-674.

Written by:
Prof. Abdelatif Janane, MD as part of Beyond the Abstract on UroToday.com. This initiative offers a method of publishing for the professional urology community. Authors are given an opportunity to expand on the circumstances, limitations etc... of their research by referencing the published abstract.

Usefulness and predictive value of PSA density, adjusted by transition zone volume, in men with PSA levels between 2 and 4 ng/ml - Abstract

More Information about Beyond the Abstract