Diagnostic Value of Prostate Specific Antigen and Its Density in Iranian Men with Prostate Cancer

This Article

Citations


Creative Commons License
Except where otherwise noted, this work is licensed under Creative Commons Attribution-NonCommercial 4.0 International License.

Article Information:


Group: 2009
Subgroup: Volume 11, Issue 2
Date: April 2009
Type: Original Article
Start Page: 170
End Page: 175

Authors:

  • Mehrzad Lotfi
  • Assistant Professor of Department of Radiology, Shiraz University of Medical Sciences, Shiraz, Fars, Iran.
  • R Assadsangabi
  • Department of Radiology, Shiraz University of Medical Sciences, Shiraz, Fars, Iran
  • M Shirazi
  • Department of Urology, Shiraz University of Medical Sciences, Shiraz, Fars, Iran
  • R Jali
  • Department of Radiology, Shiraz University of Medical Sciences, Shiraz, Fars, Iran
  • A Assadsangabi
  • Department of Radiology, Shiraz University of Medical Sciences, Shiraz, Fars, Iran
  • SA Nabavizadeh
  • Department of Radiology, Shiraz University of Medical Sciences, Shiraz, Fars, Iran

      Correspondence:

      Affiliation: Assistant Professor of Department of Radiology, Shiraz University of Medical Sciences
      City, Province: Shiraz, Fars
      Country: Iran.
      Tel: +98-711-2357282
      Fax: +98-711-2307594
      E-mail: lotfim@sums.ac.ir

Abstract:


Background: The specific threshold for prostate-specific antigen and density (PSA, PSAD) to delineate which patients are at the highest risk has been controversial. The purpose of this study was to evaluate the diagnostic value of PSA and PSAD in Iranian patients with prostate cancer.

 

Methods: Three hundred men with the serum PSA greater than 4.0 ng/ml, abnormal digital rectal examination and/or suspicious transrectal ultrasound underwent transrectal ultrasound-guided prostate biopsies. PSAD was calculated by dividing the serum PSA in ng/ml by the volume of the entire prostate in cm3. Correlation with Gleason grade of the tumor was also made. The patients were divided into three groups according to their PSA values. The receiver operator characteristic (ROC) curve was produced from the raw data on all patients.

 

Results: One hundred and two patients showed a cancer rate of 34%. The mean PSA and PSAD of the cancer group were significantly higher than those of the non-cancer group with better performance of PSAD as confirmed by ROC curve. In patients with PSA levels between 4 and 10 ng/ml, mean PSAD values in positive and negative biopsy groups showed a significant difference while mean PSA values between these biopsy groups revealed no significant difference. The PSAD cutoff of more than 0.1 had higher sensitivity than 0.15 at the expense of increasing the number of unnecessary biopsies. Among those the patients with PSA levels above 10ng/ml, both mean PSA and PSAD values of positive and negative biopsy groups had significant differences. The sensitivity of PSAD cutoff of 0.1 was not significantly higher than 0.15 while PSAD of 0.15 showed a higher specificity. PSAD >0.15 missed cancer in 18 out of 102 patients, nearly half of those with clinically significant mid and high grade cancers.

 

Conclusion: Overall, PSAD is a better diagnostic tool for the detection of prostate cancer than PSA, especially in patients with PSA between 4 and 10ng/ml. PSAD cutoff of 0.15 is not inclusive enough in patients with PSA levels between 4 and 10ng/ml and we propose PSAD of 0.1 as a better threshold for prostate biopsy in men with PSA at this range to detect clinically important cancers. Also, we recommend transrectal ultrasound guided biopsy in any patient with PSAD greater than 0.15 and PSA more than 4 ng/ml.

Keywords: Prostate-specific antigen; Density; Prostate cancer

Manuscript Body:


Introduction

 

The triad of digital rectal examination (DRE), serum prostate-specific antigen (PSA), and transrectal ultrasound-guided prostate biopsy is commonly used in the early detection of prostate cancer. The combination of DRE and serum PSA is the most useful first-line test for prostate cancer screening.1 Prostate ultrasound has been accepted as the appropriate tool for prostate biopsy guidance to detect the presence of prostate cancer if the DRE is abnormal or if the PSA level is not normal.2 Although PSA is secreted only by epithelial cells in the prostate, it is not specific for prostate cancer. A variety of other pathologic conditions cause elevation of PSA which include benign prostatic hyperplasia (BPH), prostatitis, prostate biopsy, and instrumentation, such as cystoscopy and transurethral prostatic resection. So using PSA test alone as an objective criterion may lead to an increased number of unnecessary biopsies. This is more of a concern when interpreting intermediate values of PSA levels between 4 and 10 ng/ml. To identify patients with prostate cancer who have intermediately elevated PSA levels, the concept of PSA density (PSAD) has been described.3 PSAD is the PSA value divided by the prostate volume. This concept emerged from the information that benign prostatic hyperplasia produces 0.3 ng/ml of PSA per gram of prostate tissue and prostate cancer produces 10 folds of this amount.4 Thus, men with intermediately elevated PSA levels associated with a small prostate may have prostate cancer while the same value of PSA in a man with a large prostate may indicate BPH. It has been suggested that a PSAD greater than 0.15 is associated with  25% incidence of cancer, and a PSAD less than 0.10 is associated with 5% incidence of cancer.5

However, the specific cutoff for PSA and PSAD to delineate patients at high risk who should undergo biopsy has been controversial, especially when the PSA and PSAD values are influenced by race and environment. Many investigations have been done to determine the usefulness of PSA and PSAD in detection of prostate cancer; however, few studies have been done to evaluate their performance in diagnosing prostate cancer in Iranian men. We performed a retrospective analysis of 300 men to determine the diagnostic value of PSA and PSAD in Iranian men with prostate cancer.

 

 

Material and Methods

 

This retrospective single institution study was conducted in Shahid Faghihi Hospital affiliated to Shiraz university of Medical Sciences in Shiraz, southern Iran from March 2005 to October 2006. Three hundred patients underwent transrectal ultrasound-guided prostate biopsies for abnormal digital rectal examination, suspicious transrectal ultrasound, elevated PSA, or any combination of these abnormalities. All the patients were referred to our center by their urologists and gave their formal consent for the study. PSA was determined by an immunoradiometric assay, using monoclonal antibodies (Kavoshyar, Tehran, Iran) which could measure PSA levels within 0.01ng/ml. The most recent determination of the serum PSA level in any patient was used for this study. The normal range was defined as PSA values less than 4ng/ml.

Preparation before the ultrasound examination and biopsy included a self-administered cleansing enema and a broad-spectrum prophylactic antibiotic. Men who were treated with 5-alpha reductase inhibitors and received hormonal therapy were excluded from this study. Transrectal ultrasound was performed by an expert radiologist who was unaware of the PSA results, using a GE logique 500 scanner with 6.0 to 8.0 MHZ, high- resolution, near- field, end-firing transducer. Each gland was examined in both axial and sagittal projections. The prostate volume was calculated as follows:

Volume equals (length × width × height) × 0.523, with length being measured in the longitudinal view, and width and height in the axial view.

PSAD          was calculated by dividing PSA by the prostate volume. Biopsies were obtained using an 18-gauge biopsy needle in a spring-driven biopsy gun. All biopsies were performed under ultrasound guidance via the transrectal route on an outpatient basis.

All the patients were subjected to fourteen systemic biopsies (from medial and far lateral aspect of the base, mid and apical areas of the right and left peripheral zones plus one biopsy form each transitional zone on either sides). Additional transrectal ultrasound-guided biopsies were performed at the suspected regions, either hypoechoic or heterogeneous. The core specimens were fixed in separate formalin containers and labeled accordingly to denote the different sites. An expert pathologist examined all the slides. Each biopsy specimen was categorized histologically (normal tissue, hyperplastic changes, acute or chronic inflammation, prostatic intraepithelial neoplasia, or cancer) and a further classification into Gleason grade of the tumor was done as follows: Low-grade cancer=Gleason score 2 to 4, mid-grade cancer=Gleason score 5 to 7 and high-grade cancer=Gleason score 8 to 10.

Mann Whitney test was used to compare the means. Receiver operator characteristic (ROC) curves, comparing PSA and PSAD, were generated. p<0.05 was considered as statistically significant.

 

 

Results

 

102 out of 300 patients evaluated had final cancer diagnosis in biopsy, yielding a detection rate of 34 percent. The age range was 42 to 80 with a mean of 67.25 and standard deviation of 9.40. The patients were divided into three groups based on different levels of their serum PSA: serum PSA <4 ng/ml, serum PSA between 4 and 10 ng/ml and serum PSA >10 ng/mL.

25 men had PSA levels of less than 4.0 ng/ml, of whom 3 (12%) had a positive biopsy for prostate cancer. The mean PSA and PSAD for the positive or negative biopsies were 3.07 (SD: 0.73) versus 2.45 (SD: 1.07) (p: 0.497) and 0.16 (SD: 0.077) versus 0.08 (SD: 0.04) (p: 0.01), respectively. A total of 159 patients had PSA levels of 4.0-10.0 ng/ml, of whom, 29 (18.2%) men had a positive biopsy for prostate cancer. The mean PSA and PSAD for the positive or negative biopsies were 7 (SD: 1.89) versus 6.6 (SD: 1.87) (p: 0.256) and 1.45 (SD: 6.83) versus 0.15 (SD: 0.1) (p value: 0.02), respectively. Of 116 men with PSA levels greater than 10.0 ng/ml, 70 (60.3%) men had positive biopsy results. The mean values for PSA and PSAD for men with positive and negative biopsies were 38.7 (SD: 42.28) versus 17.8 (SD: 9.27) (p: 0.0001) and 1.14 (SD: 1.65) versus 0.26 (SD: 0.16) (p: 0.0002), respectively. In general, the mean values for PSA and PSAD for patients in this study with positive or negative biopsies were 28.34 (SD: 37.87) versus 8.77 (SD: 7.01) ng/ml (p: 0.0001) and 1.20 (SD: 3.87) versus 0.16 (SD: 0.13), (p: 0.0007), respectively.

For each group, the mean values of PSA and PSAD in patients with positive and negative biopsy results were calculated and P values were measured. The overall results are summarized in Table 1. A ROC curve was produced from the raw data on all the patients, for both PSA and PSAD (Figure1). The areas under both curves were calculated, being 78.1 percent for serum PSA and 81.3 percent for PSAD, with a statistically significant difference between both areas (p: 0.0002).

 

 

Table 1: Comparison of mean PSA and mean PSAD in patients with positive and negative prostate biopsies in 3 groups.

PSA levels less than 4 ng/ml

PSA levels between 4 and 10 g/ml

PSA levels above 10ng/ml

Biopsy

No. (%)

Mean

PSA

Mean

PSAD

No. (%)

Mean

PSA

Mean

PSAD

No. (%)

Mean

 PSA

Mean

PSAD

Positive

  3 (12)

3.07

0.16

  29 (18.2)

7

1.45

70 (60.3)

38.7

1.14

Negative

22 (88)

2.45

0.08

130 (81.8)

6.6

0.15

46 (39.7)

17.8

0.26

Total

25

p : 0.492

p: 0.01

159

p: 0.256

P: 0.02

116

p: 0.0001

p: 0.0002

PSA: Prostate- specific antigen, PSAD: PSA density

 

 

Fig 1: Comparing PAS and PSAD in detecting prostate cancer using ROC curve.

 

 

The overall results are summarized in Table 1. ROC curves were produced from the raw data on patients in the second (Figure 2) and third (Figure 3) groups. The curves demonstrate the performance of different PSAD ratios in these two PSA ranges. Various sensitivities and specificities are noted for different PSAD cutoff points according to different PSA ranges. Tables 2 and 3 show the effects of changing PSAD cutoff points on the test’s sensitivity and specificity for patients with different PSA ranges.

 

 


Fig 2: PSAD performance in patients with PSA levels between 4 and 10 ng/ml using ROC curve.

 

 

Fig 3: PSAD performance in patients with PSA levels above 10 ng/ml using ROC curve.

 

 

Table 2: Performance of different PSAD cutoff points in patients with serum PSA levels between 4 and 10 ng/ml.

Upper limit of PSAD cut-off point

Sensitivity

(95% CI)

Specificity

(95% CI)

Positive predictive value

Negative

predictive value

> 0.01

100.0 (87.9-100.0)

  1.5 (0.2-5.5)

18.5

100.0

> 0.05

  96.6 (82.2- 99.4)

  4.6 (1.7-9.8)

18.6

  85.7

> 0.1

  72.4 (52.8-87.2)

36.9 (28.6-45.8)

20.04

  85.7

> 0.15

  55.2 (35.7-73.5)

67.7 (58.9-75.6)

27.6

  87.1

> 0.21*

  44.8 (26.5-64.3)

83.8 (76.4-89.7)

38.2

  87.2

> 0.25

  17.2 (5.9-35.8)

89.2 (82.6-94.0)

26.3

  82.9

PSA: Prostate- specific antigen, PSAD: PSA density, CI: Confidence Interval

 

 

Table 3: Performance of different PSAD cutoff points in patients with serum PSA levels above 10 ng/ml.

Upper limit of

PSAD cut-off point

Sensitivity

(95% CI)

Specificity

(95% CI)

Positive

predictive value

Negative

predictive value

>= 0.06

100.0 (94.8-100.0)

  0.0 (0.0- 7.8)

60.3

 

> 0.1

  98.6 (92.3-99.8)

  8.7 (2.5-25.8)

62.2

80.0

> 0.15

  94.3 (86.0-98.4)

30.4 (17.8-45.8)

67.3

77.8

> 0.21

  88.6 (78.7-94.9)

50.0 (34.9-65.1)

72.9

74.2

> 0.25

  84.3 (73.6-91.9)

67.4 (52.0-80.5)

79.7

73.8

> 0.30

  82.9 (72.0-90.8)

76.1 (61.2-87.4)

84.1

74.5

PSA: Prostate- specific antigen, PSAD: PSA density, CI: Confidence Interval

 

 

Regardless of the PSA or the PSAD values, 80% of the cancers diagnosed were in the mid-grade or high-grade range. Tables 4-6 compare PSA and PSAD levels in men with cancer and with Gleason grade correlation.

 

 

Table 4: Effect of changes in PSAD cutoff point in the number of patients with positive and saved unnecessary biopsies.

PSAD

Positive biopsies

Saved biopsies

0.21

76 (74.5%)

149 (75.3%)

0.15

84 (82.4%)

116 (58.6%)

0.1

97 (95.1%)

  56 (28.3%)

PSAD: Prostate-specific antigen density

 

 

Table 5: Correlation between PSA levels and Gleason grade in men with cancer.

PSA

Men with

cancer N0. (%)

Low grade

Mid-grade

High grade

<4

    3/25 (12)

  2

  1

  0

4-10

  29/159 (18.2)

11

13

  5

> 10

  70/116 (60.3)

  8

41

21

Total

102

21

55

26

PSA: Prostate- specific antigen

 

 

Table 6: Correlation between PSAD values and Gleason grade in men with cancer

PSAD

Men with cancer

Low grade

Mid grade

High grade

< 0.1

  5/61

  2

  1

2

<0.15

18/134

  8

  6

4

< 0.21

25/175

10

10

5

PSAD: Prostate- specific antigen density

 

 

Discussion

 

Prostate cancer is a commonly diagnosed cancer in men, making this disease a significant public health issue. Serum prostate specific antigen is an important tool in screening and early detection of prostate cancer, and is considered one of the most important prognostic factors amongst patients with prostate cancer.6 Racial differences in PSA levels have been found. Two studies performed to determine the reference values of serum PSA for Iranian men revealed that the PSA values in them were significantly lower than those for white and black Western men, and slightly lower than those for Japanese men.7,8

Attempts have been made to refine the use of PSA in detection strategies to decrease the number of unnecessary biopsies while minimizing the number of missed cancers. Since serum PSA levels can be increased by a number of processes besides prostate cancer, the concept of PSAD is described as a potentially better predictor of prostate cancer.3 Although many studies have looked at the performance of PSA and PSAD in prostate cancer detection in afro-Caribbean and Caucasian populations,9 few investigations have been done on Iranian population.

The patients in our study had a mean PSA level of 15.4 as compared with other studies9 with a mean PSA level of 21.56 and 10.96 in Afro-Caribbeans and Caucasians, respectively. On the other hand, the mean PSAD for the studied group was 0.39 compared to 0.68 in Afro-Caribbean populations and 0.34 in Caucasians.9 These possible differences may be accounted for by socioeconomic status, race, culture, health care access and differences in tumor biology; however, these issues need to be clarified in well- designed prospective studies.

159 men had PSA values between 4.0 and 10.0 ng/ml, of whom 29 (18.2%) had positive biopsies for prostate cancer. For this group as a whole, the mean PSA values of the positive and negative biopsy groups showed no significant difference. The mean PSAD was significantly different (p: 0.02) between the positive and negative biopsy groups. One hundred sixteen men had PSA values greater than 10 ng/ml. Both PSA and PSAD were effective in distinguishing men with positive or negative biopsies for prostate cancer in the entire group, being statistically significant.

By comparing the areas under the ROC curves which were produced from the raw data on all patients for both PSA and PSAD, it was found that PSAD was a better diagnostic tool for the detection of prostatic carcinoma. (p: 0.0002). Brawer et al. did not observe a benefit of PSAD over PSA in cancer detection.10 This finding is contradictory to those of Benson et al.5 In a current study by Gregorio et al., it was also found that the combination of free PSA and PSAD could allow better discrimination between benign and malignant causes of increased PSA.11

Several series have compared PSAD cutoff values of 0.1 and 0.15 in prostate cancer detection. Bazinet et al. found that a ROC curve confirmed a PSAD of 0.15 as the best cutoff in patients with a negative DRE result, negative ultrasound result, and PSA levels between 4.1 and 10 ng/ml.12 Catalona et al., using similar group criteria (PSA 4.1-9.9 ng/ml, normal DRE and normal ultrasound result) found that half the tumors were missed by using a PSAD cutoff point of more than 0.15.13

Our analysis revealed that out of 102 patients 84 and 97 diagnosed with prostate cancer had PSAD greater than 0.15 and 0.1, respectively. We identified 18 out of 102 patients with cancer that would have been missed by using a PSAD cutoff of 0.15 in comparison to missing 5 patients with cancer in cutoff of 0.1.

In our series, the test results for prostate cancer detection in men with PSA levels between 4 and 10 ng/ml at PSAD cutoffs of 0.15 and 0.1 showed sensitivities of 55.2% vs. 72.4% and specificities of 67.7% vs. 36.9%, respectively. Calalona et al. observed sensitivities of 52% vs. 79% and specificities of 81% vs. 47% when comparing PSAD cutoffs of 0.15 and 0.1, respectively.13 Using a PSAD cutoff of 0.15, Cookson et al. observed a sensitivity of 61% and a specificity of 22% in 44 patients with a normal DRE and PSA of 4 to 10 ng/ml.14 However, this study and the previously mentioned ones failed to consider variations in ethnicity.10

Overall, at PSA values between 4 and 10 ng/ml, cutoff of 0.1 shows a better sensitivity than that of 0.15 at the expense of increasing the number of unnecessary biopsies and diminishing positive predictive value of the test.  However, our data suggest that more than half of all men with cancer who have a PSAD less than 0.15 demonstrate mid to high grade cancers which are clinically significant.

In patients with serum PSA more than 10 ng/ml, applying PSAD of 0.15 will increase the specificity and positive predictive value of the test without significantly compromising the test sensitivity.  In addition, it seems reasonable that all men with PSA greater than 4 ng/ml should have an ultrasound-guided biopsy if their PSAD is greater than 0.15.

This work is the first study which deals with diagnostic value of PSA and PSAD for detection of prostate cancer in Iran. Furthermore, few articles have analyzed and compared the accuracy of PSA and PSAD in the range of PSA between 4-10ng/ml through the ROC curve,11,15 as was performed in this work. Therefore, we believe PSAD of 0.15 is not an ideal cutoff point in patients with PSA between 4 and 10 ng/ml and that of 0.1 or greater and above should be used as an indicator for biopsy consideration in this PSA range.

 

Acknowledgement

 

The authors would like to thank the Office of Vice Chancellor of Shiraz University of Medical Sciences for financial support.

 

Conflict of interest: None declared.

References: (15)

  1. Catalona WJ, Richie JP, Ahmann FR, Hudson MA, Scardino PT, Flanigan RC, deKernion JB, Ratliff TL, Kavoussi LR, Dalkin BL, et al. Comparison of digital rectal examination and serum prostate specific antigen in the early detection of prostate cancer: results of a multicenter clinical trial of 6,630 men. J Urol 1994;151(5):1283-90. [7512659]
  2. Littrup PJ, Sparschu R. Transrectal ultrasound and prostate cancer risks. Cancer 1995;755:1805-13. [doi:10.1002/1097-0142(19950401)75:7+<1805::AID-CNCR2820751609>3.0.CO;2-K]
  3. Benson MC, Whang IS, Olsson CA, McMahon DJ, Cooner WH. The use of prostate specific antigen density to enhance the predictive value of intermediate levels of serum prostate specific antigen. J Urol 1992;147:817-21. [1371555]
  4. Stamey TA, Kabalin JN, McNeal JE, Johnstone IM, Freiha F, Redwine EA, Yang N. Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate. II. Radical prostatectomy treated patients. J Urol 1989;141:1076-83. [2468795]
  5. Narayan P, Ashutosh, Badrinath K. Prostate cancer: Part I. Diagnosis and staging. Hospital Physician 1996;34:10-24.
  6. Partin AW, Kattan MW, Subong EN, Walsh PC, Wojno KJ, Oesterling JE, Scardino PT, Pearson JD. Combination of prostate-specific antigen, clinical stage, and Gleason score to predict pathological stage of localized prostate cancer. A multi-institutional update. JAMA 1997;277:1445-51. [9145716] [doi:10.1001/jama.277.18.1445]
  7. Hosseini SY, Moharramzadeh M, Ghadian AR, Hooshyar H, Lashay AR, Safarinejad MR. Population-based screening for prostate cancer by measuring total serum prostate-specific antigen in Iran. Int J Urol 2007;14:406-11. [17511722] [doi:10.1111/j.1442-2042.2006.01729.x]
  8. Mehrabi S, Ghafarian Shirazi HR, Rasti M, Bayat B. Analysis of serum prostate-specific antigen levels in men aged 40 years and older in yasuj, iran. Urol J 2005;2:189-92. [17602427]
  9. Abdalla I, Ray P, Vaida F, Vijayakumar S. Racial differences in prostate-specific antigen levels and prostate-specific antigen densities in patients with prostate cancer. Am J Clin Oncol 1999;22:537-41. [10597735] [doi:10.1097/00000421-199912000-00001]
  10. Brawer MK, Aramburu EA, Chen GL, Preston SD, Ellis WJ. The inability of prostate specific antigen index to enhance the predictive the value of prostate specific antigen in the diagnosis of prostatic carcinoma. J Urol 1993;150:369-73. [7686982]
  11. Gregorio EP, Grando JP, Saqueti EE, Almeida SH, Moreira HA, Rodrigues MA. Comparison between PSA density, free PSA percentage and PSA density in the transition zone in the detection of prostate cancer in patients with serum PSA between 4 and 10 ng/mL. Int Braz J Urol 2007;33:151-60. [17488533] [doi:10.1590/S1677-55382007000200004]
  12. Bazinet M, Meshref AW, Trudel C, Aronson S, Péloquin F, Nachabe M, Bégin LR, Elhilali MM. Prospective evaluation of prostate-specific antigen density and systematic biopsies for early detection of prostatic carcinoma. Urology 1994;43:44-51. [7506853] [doi:10.1016/S0090-4295(94)80260-2]
  13. Catalona WJ, Richie JP, deKernion JB, Ahmann FR, Ratliff TL, Dalkin BL, Kavoussi LR, MacFarlane MT, Southwick PC. Comparison of prostate specific antigen concentration versus prostate specific antigen density in the early detection of prostate cancer: receiver operating characteristic curves. J Urol 1994;152:2031-6. [7525994]
  14. Cookson MS, Floyd MK, Ball TP Jr, Miller EK, Sarosdy MF. The lack of predictive value of prostate specific antigen density in the detection of prostate cancer in patients with normal rectal examinations and intermediate prostate specific antigen levels. J Urol 1995;154:1070-3. [7543601] [doi:10.1016/S0022-5347(01)66979-9]
  15. Okegawa T, Noda H, Nutahara K, Higashihara E. Comparison of two investigative assays for the complexed prostate-specific antigen in total prostate-specific antigen between 4.1 and 10.0 ng/mL. Urology 2000;55:700-4. [10792084] [doi:10.1016/S0090-4295(99)00589-0]