Prostate Cancer and Prostatic Diseases (2011) 14, 253–261 & 2011 Macmillan Publishers Limited All rights reserved 1365-7852/11 www.nature.com/pcan

ORIGINAL ARTICLE

Budget impact analysis of a new prostate cancer risk index for prostate cancer detection MB Nichol1,4, J Wu1, JJ An1, J Huang2, D Denham2, S Frencher3 and SJ Jacobsen3 1 Department of Clinical Pharmacy and Pharmaceutical Economics and Policy, School of Pharmacy, University of Southern California, Los Angeles, CA, USA; 2Beckman Coulter, Inc., Brea, CA, USA; 3Kaiser Permanente Southern California, Pasadena, CA, USA and 4School of Policy, Planning, and Development, University of Southern California, CA, USA

The objective of this study was to evaluate the budget impact of a new prostate cancer risk index for detecting prostate cancer. The index is calculated as the combination of serum prostate-specific antigen (PSA), free PSA and a precursor form p2PSA. We constructed two budget impact models using PSA cutoff values of X2 ng ml1 (model #1) and X4 ng ml1 (model #2) for recommending a prostate biopsy in a hypothetical health plan with 100 000 male members aged 50–75 years old. The budgetary impact on the 1-year expected total costs for prostate cancer detection was calculated. Adding the index to the current PSA prostate cancer testing strategies including the total PSA and percent free PSA, the number of detected cancer cases decreased by 20 and 5, in models #1 and #2, respectively. The savings on expected 1-year cost for prostate cancer detection were $356 647 (or $0.30 per-member-per-month (PMPM)) in model #1 and $94 219 ($0.08 PMPM) in model #2. The index produced higher cost savings in the model #1 with PSA cutoff X2 ng ml1 than the model #2 with cutoff X4 ng ml1 with a small short-term reduction in the number of positive tests. Prostate Cancer and Prostatic Diseases (2011) 14, 253–261; doi:10.1038/pcan.2011.16; published online 3 May 2011

Keywords: prostate-specific antigen; free PSA; precursor form p2PSA; prostate cancer risk index; budget impact analysis

Introduction Prostate cancer is the most prevalent malignancy and the second leading cause of death among men in the United States.1 Although routine prostate cancer screening is controversial,2 the guidelines from the American Urological Association (AUA)3 and the American Cancer Society4 recommend serum prostate-specific antigen (PSA) testing for men with counseling regarding risks and benefits of prostate cancer screening. PSA level measurement and digital rectal examination remains the preferred approach for early prostate cancer detection.5–7 A PSA cutoff level X4 ng ml1 is widely recommended for prostate biopsy although this value has B20% sensitivity and 60% specificity for prostate cancer.8,9 As lowering the threshold value may improve sensitivity,10– 12 many clinicians now recommend a prostate biopsy at lower PSA values. Although this may improve the likelihood of detecting cancers, lowering the threshold value increases the identification of clinically insignificant tumors.13 To improve PSA specificity, several PSA derivatives or biomarkers, and test methods have been suggested, Correspondence: Professor MB Nichol, Department of Clinical Pharmacy and Pharmaceutical Economics and Policy, School of Pharmacy, University of Southern California, 1540 Alcazar Street, CHP-140, Los Angeles, CA 90033, USA. E-mail: [email protected] Received 8 September 2010; revised 16 February 2011; accepted 3 March 2011; published online 3 May 2011

including age-specific reference ranges for serum PSA, the ratio of free to total PSA (percent free PSA (%fPSA)), complexed PSA, PSA velocity and PSA density.14–17 In addition, some precursor forms of PSA (p2PSA, for example) have recently demonstrated their potential usefulness in the early detection of prostate cancer.15,18–20 All strategies may significantly reduce false positive results among men with total PSA 2.5–4.0 ng ml1 (ref. 18). Furthermore, previous studies have demonstrated that combinations of PSA, fPSA, and p2PSA in a multivariate model may achieve 41–44% clinical specificity at 90% sensitivity, as compared with individual markers (23% specificity for PSA and 18–33% specificity for %fPSA).19,20 Access Hybritech p2PSA has being investigated for use with Access Hybritech PSA and fPSA to calculate a multivariate index developed by Beckman Coulter (Brea, CA, USA) to determine the individuals’ relative risk of prostate cancer.21–23 Currently, a multicenter clinical study in the United States sponsored by Beckman Coulter suggests that, at 95% sensitivity, this prostate cancer risk index doubled the specificity of prostate cancer detection relative to %fPSA for individuals with PSA 2–10 ng ml1 (ref. 21). Although the index is not currently available for commercial distribution in the United States, an economic evaluation of the index may inform how multiple measures might affect prostate cancer management costs. The present study evaluates the 1-year budget impact to a health plan of adding the index to conventional PSA and fPSA blood tests.

Budget impact of a prostate cancer risk index MB Nichol et al

254

Materials and methods Population estimates The model used a hypothetical health plan with 100 000 men aged 50–75 years old. The number of males undergoing PSA tests for prostate cancer detection was estimated based on the 2009 U.S. Census data for the population age distribution and published annual prevalence data.24,25 Model description The budget impact model used the most recent AUA recommendation for prostate cancer screening.3 Three PSA testing strategies were compared: (1) PSA, (2) %fPSA, and (3) the index. Strategies (1) and (2) represent

current clinical practice, and are reference scenarios in the model. Two PSA threshold values (2 and 4 ng ml1) were used for prostate biopsy recommendation or additional PSA testing. Individuals have three probabilities of test results: (1) The test is negative (PSA ocutoff value). This result indicates no cancer detection and no biopsy is ordered. (2) The test is positive (PSA410 ng ml1). An individual is referred to an urologist and a biopsy confirms prostate cancer. (3) The test is borderline (PSA value is between cutoff and 10 ng ml1). For the PSA strategy, an individual

Table 1 Model probabilities Model parameter Probability of test results for different PSA levels Probability of PSA410 ng ml1 Overall Age 40–49 years Age 50–59 years Age 60–69 years Age 70–75 years Probability of PSAX2 ng ml1 Overall Age 40–49 years Age 50–59 years Age 60–69 years Age 70–75 years Probability of PSAX4 ng ml1 Overall Age 40–49 years Age 50–59 years Age 60–69 years Age 70–75 years Probability of PSA 2 to 10 ng ml1a Overall Age 40–49 years Age 50–59 years Age 60–69 years Age 70–75 years Probability of PSA 4 to 10 ng ml1a Overall Age 40–49 years Age 50–59 years Age 60–69 years Age 70–75 years Probability of having a positive %fPSA test Probability of a positive test for the index

Base case value

Range tested in sensitivity analysis Data source Lacher et al.26

0.011 NA 0.008 0.010 0.033

0.006–0.019 0.001 0.006–0.010 0.003–0.022 0.016–0.058

0.227 NA 0.167 0.272 0.415

0.189–0.270 0.06 0.130–0.211 0.235–0.311 0.368–0.464

0.068 NA 0.038 0.080 0.195

0.049–0.093 0.017 0.023–0.059 0.059–0.106 0.155–0.239

Lacher et al.26

Lacher et al.26

Calculated 0.148 NA 0.092 0.192 0.318

0.041

Calculated 0.057 NA 0.030 0.070 0.162 0.797 0.752

0.325–0.601 0.128–0.425

Provided by Beckman Coulter21b Provided by Beckman Coulter21b

0.667

0.604–0.728

Probability of prostate cancer in a man with a value of PSA 2 to 10 ng ml1c

0.250

0.250–0.323

Probability of prostate cancer in a man with a positive %fPSA test Probability of prostate cancer in a man with a positive test for the index

0.276

0.379–0.313

Andriole et al. (PLCO cancer screening study),33; Crawford et al.,27; Catalona et al.31 Based case data was provided by Beckman Coulter21b. Tested ranges were calculated25,28,32 Provided by Beckman Coulter21b

0.296

0.383–0.516

Provided by Beckman Coulter21b

Probability of prostate cancer detection Probability of prostate cancer in a man with a value of PSA410 ng ml1

0.016

Abbreviations: %fPSA, percent free PSA; NA, not apply; PLCO, prostate, lung, colorectal and ovarian; PSA, prostate-specific antigen. a Probability of PSA 2–10 ng ml1 ¼ (probability of PSAX2 ng ml1 )(probability of PSA410 ng ml1); (probability of PSA 4–10 ng ml1) ¼ (probability of PSAX4 ng ml1)(probability of PSA410 ng ml1) b Data were obtained from a simulation study by Beckman Coulter, which is currently under FDA review. c Probability of prostate cancer in PSA 2–10 ng ml1 ¼ ((Probability of PSAX2 ng ml1) (probability of prostate cancer of PSAX2 ng ml1)(Probability of PSA410 ng ml1) (probability of prostate cancer of PSA 410 ng ml1))/(probability of PSA 2–10 ng ml1).

*

Prostate Cancer and Prostatic Diseases

*

Budget impact of a prostate cancer risk index MB Nichol et al

is referred to an urologist, repeats PSA and receives a biopsy for prostate cancer diagnosis. We assumed the PSA test result is the same as the first test. For the %fPSA testing strategy, fPSA is performed as a reflex test, and %fPSA is calculated. For the index strategy, both fPSA and p2PSA are performed as reflex tests and the index score is calculated. Based on the Beckman Coulter simulation study,21a value of o25% for %fPSA or a score higher than 25 for the index corresponds to B30% of weighted average relative risk of prostate cancer.21 These values represent a positive test used for an urologist visit referral and biopsy. Conversely, a %fPSA value of 25% or higher or an index score of 25 or less is considered negative, with no biopsy.

Data sources and input data Probabilities of positive PSA test results and cancer detection were derived from the published literature. Where data were unavailable, probabilities were derived, as follows. Probability of PSA test results. Age-specific prevalence for each PSA range was determined according to population-based United States data (Table 1).26 Ninety-five percent confidence intervals were used for the sensitivity analyses. Other published studies with similar data27,33,34 were not applied in the model, because they did not provide age-specific prevalence. As the probability of PSA values in the equivocal ranges 2 or 4 ng ml1 to 10 ng ml1 was not available from the literature, it was calculated as the difference between the probability of PSA4cutoff value and the probability of PSA 410 ng ml1. The positive test probabilities for %fPSA and the index were obtained from the Beckman Coulter simulation study.21

255

Table 2 Health care component costs Cost component

Base case value

Range tested in sensitivity analysis

Primary care physician visit Urologist visit

$37.15 $92.33

$27.86–$46.44 $69.25–$115.41

PSA blood test cost PSA fPSA p2PSAa

$26.85 $26.85 $71.95

$20.14–$33.56 $20.14–$33.56 $49.86–$94.04

$2102.68 $236.24 $183.94

$1577.01–$2628.35

Biopsyb Biopsy of prostate gland Echography-guided prostate biopsy Transrectal ultrasound of prostate gland Tissue exam by pathologist Immunohistochemistry Other diagnostic tests: urinalysis

$141.74 $1246.46 $294.30 $11.00

$8.25–$13.75

Abbreviation: %fPSA, percent free PSA; PSA, prostate-specific antigen. Note: costs data were based on the 2009 Medicare fee schedule. The ranges tested in sensitivity analyses were calculated as decreasing and increasing 25% of base case costs. a The p2PSA test is not currently available for the United States market. Its cost has been estimated from discussions with Beckman Coulter. The cost including p2PSA, fPSA and total PSA is incorporated into the cost of the index. b The biopsy cost was the integrated cost from 12-core prostate biopsy, echographyguided biopsy, transrectal ultrasound, tissue examination by a pathologist and three immunohistochemistry stains.

%fPSA testing strategy; (3) one PSA, one fPSA and one p2PSA for the index testing strategy. We assumed that routine urinalysis was performed for positive PSA blood tests. As prostate biopsies involve multiple procedures, costs included 12-core prostate biopsy, echography-guided biopsy, transrectal ultrasound, tissue examination by a pathologist and three immunohistochemistry stains.

Probability of prostate cancer detection. Among individuals who had a positive PSA test result, the prostate cancer detection rate upon biopsy was determined from published studies (Table 1).

Market share. Base case market share for the PSA strategy was estimated as 90, and 10% for the %fPSA strategy. In the new scenario, we assumed a 10% market share for the index strategy and maintenance of 10% for the %fPSA strategy, and 80% for the PSA strategy.

Costs. Base case costs were obtained from the 2009 Medicare national fee schedule. The sensitivity analysis was performed on the base case costs ±25%. Direct health care costs for relevant physician visits, diagnostic tests and procedures were included. (Table 2). Office visit costs included one primary care visit for the individuals with a negative PSA test, or one primary care visit and two urologist visits for those with a positive test for PSA, or %fPSA, or the index (one visit for biopsy and one for follow-up consultation). We assumed all screened men had a negative digital rectal examination in the model. The cost of PSA blood tests depended on the PSA test result. If PSA was less than the cutoff value or 410 ng ml1, one PSA test was included. If the PSA test result is borderline, the follow-up testing strategy included the following: (1) two PSA test costs (one at the primary care visit and the other at the urologist visit) for the PSA strategy; (2) one PSA and one fPSA for the

Analysis We compared the expected 1-year total costs for prostate cancer detection between the reference and new scenarios. First, the PSA test results were assessed. The total number of positive PSA tests was calculated as the number of men receiving a PSA test multiplied by the probability of a positive PSA test. The number of detected prostate cancers (true positive PSA test) was calculated as the number of positive PSA tests multiplied by the probability of prostate cancer, in which prostate cancer was confirmed via biopsy for the individuals with a positive PSA test. The number of unnecessary biopsies (false positive PSA test) was defined as the individuals who have a positive PSA test, with a subsequent nonconfirmatory biopsy. Second, we calculated the total costs for all strategies, which included costs for office visits, digital rectal examination procedures, PSA blood tests, other laboratory tests and Prostate Cancer and Prostatic Diseases

Prostate Cancer and Prostatic Diseases

$17 681 752 $2 364 176 $976 194 $74 635 $3 864 896 $10 401 851 $14.74 $9 620

Budget component by service Total costs for prostate cancer screening Office visit (primary care+urologist) PSA blood test Lab tests Necessary biopsy Unnecessary biopsy Total costs presented as PMPM Total costs per prostate cancer detectione $1 630 532 $235 842 $108 466 $6 694 $384 277 $895 252 $1.36 $8 922

$1 452 448 $435 685 $1 016 763 $177 668

609 183 426 2715

10%

%fPSA (B)

$19 312 284 $2 600 018 $1 084 660 $81 329 $4 249 173 $11 297 103 $16.09 $9 557

$17 650 307 $4 818 625 $12 831 682 $1 657 820

7 394 2 021 5 373 25 842

100%

Total (C ¼ A+B)

$15 717 113 $2 101 490 $867 728 $66 343 $3 435 463 $9 246 089 $13.10 $9 620

$14 398 097 $3 895 946 $10 502 150 $1 315 690

6 031 1 634 4 397 20 558

80%

PSA (D)

$1 630 532 $235 842 $108 466 $6 694 $384 277 $895 252 $1.36 $8 922

$1 452 448 $435 685 $1 016 763 $177 668

609 183 426 2 715

10%

%fPSA (E)

$1 607 992 $229 892 $159 990 $6 339 $387 850 $823 921 $1.34 $8 718

$1 414 203 $451 203 $963 000 $193 374

576 184 392 2 747

10%

The Index (F)

New Scenario (G)

$18 955 637 $2 567 224 $1 136 184 $79 376 $4 207 591 $10 965 263 $15.80 $9 473

$17 264 748 $4 782 834 $12 481 914 $1 686 732

7 216 2 001 5 215 26 020

100%

Total (G ¼ D+E+F)

$356 647 $32 795 $51 524 $1 954 $41 582 $331 840 $0.30 $84

$385 559 $35 790 $349 769 $28 912

178 20 158 178

Budget impact (GC)a

Abbreviations: %fPSA, percent free PSA; PMPM, per-member-per-month; PSA, prostate-specific antigen. a Budget impact was calculated by the total costs of the new scenario minus the total costs of the reference scenario. b Four PSA test results are considered as positive test: (1) PSA 410 ng ml1, (2) PSAX2 ng ml1 for the PSA strategy, (3) PSA at 2–10 ng ml1 and a positive %fPSA test for the %fPSA strategy, (4) PSA at 2–10 ng ml1 and a positive index for the index strategy. c Detected prostate cancer was defined as the true positive PSA test, in which prostate cancer is confirmed via biopsy for the individuals who have a positive PSA test. d Unnecessary biopsy was defined as a biopsy performed due to the false-positive PSA test, in which prostate cancer was not confirmed via biopsy for the individuals who have a positive PSA test. e Total costs per prostate cancer detection was calculated as total costs for prostate cancer screening divided by total number of detected prostate cancers. Note: Costs were presented in 2009 US dollars.

$16 197 859 $4 382 940 $11 814 919 $1 480 152

6 785 1 838 4 947 23 127

90%

Budget component by PSA test results Total costs for positive PSA tests Total cost for detected prostate cancers Total cost for unnecessary biopsies Total cost for negative tests

PSA test results Total number of positive PSA testsb Total number of detected prostate cancersc Total number of unnecessary biopsiesd Total number of negative PSA tests

Market share

PSA (A)

Reference scenario (C)

256

Parameters

Table 3a Base case budget impact analysis results for PSA biopsy threshold of 2 ng ml1 Budget impact of a prostate cancer risk index MB Nichol et al

$6 654 971 $1 488 456 $848 863 $22 470 $1 372 005 $2 923 177 $5.55 $10 199

Budget component by service Total costs for prostate cancer screening Office visit (primary care+urologist) PSA blood test Lab tests Necessary biopsy Unnecessary biopsy Total costs presented as PMPM Total costs per prostate cancer detectione $651 177 $158 292 $94 318 $2 074 $140 516 $255 977 $0.54 $9 744

$449 088 $158 708 $290 380 $201 673

189 67 122 3135

10%

%fPSA (B)

$7 306 148 $1 646 748 $943 181 $24 544 $1 512 520 $3 179 154 $6.09 $10 157

$5 316 658 $1 709 076 $3 607 582 $1 985 333

2 231 719 1 512 31 005

100%

Total (C ¼ A+B)

Reference scenario (C)

$5 915 530 $1 323 072 $754 545 $19 973 $1 219 560 $2 598 380 $4.93 $10 199

$4 326 729 $1 378 105 $2 948 624 $1 585 475

1 816 580 1 236 24 773

80%

PSA (D)

$651 177 $158 292 $94 318 $2 074 $140 516 $255 977 $0.54 $944

$449 088 $158 708 $290 380 $201 673

189 67 122 3135

10%

%fPSA (E)

$645 223 $156 720 $107 930 $1 981 $141 460 $237 132 $0.54 $9 591

$438 985 $162 808 $276 177 $205 822

180 67 113 3144

10%

The index (F)

New scenario (G)

$7 211 929 $1 638 085 $956 793 $24 028 $1 501 535 $3 091 489 $6.01 $10 099

$5 214 801 $1 699 621 $3 515 180 $1 992 971

2 184 714 1 470 31 052

100%

Total (G ¼ D+E+F)

$94 219 $8 664 $13 611 $516 $10 985 $87 665 $0.08 $58

$101 857 $9 455 $92 402 $7 638

47 5 42 47

Budget impact (GC)a

Abbreviations: %fPSA, percent free PSA; PMPM, per-member-per-month; PSA, prostate-specific antigen. Note: Costs were presented in 2009 US dollars. a Budget impact was calculated by the total costs of the new scenario minus the total costs of the reference scenario. b Four PSA test results are considered as positive test, (1) PSA 410 ng ml1, (2) PSAX4 ng ml1 for the PSA strategy, (3) PSA at 4 to 10 ng ml1 and a positive %fPSA test for the %fPSA strategy, (4) PSA at 4–10 ng ml1 and a positive index for the index strategy. c Detected prostate cancer was defined as the true positive PSA test, in which prostate cancer is confirmed via biopsy for the individuals who have a positive PSA test. d Unnecessary biopsy was defined as a biopsy performed due to the false positive PSA test, in which prostate cancer was not confirmed via biopsy for the individuals who have a positive PSA test. e Total costs per prostate cancer detection was calculated as total costs for prostate cancer screening divided by total number of detected prostate cancers.

$4 867 570 $1 550 368 $3 317 202 $1 783 660

2 043 653 1 390 27 870

90%

PSA (A)

Budget component by PSA test results Total costs for positive PSA tests Total cost for detected prostate cancers Total cost for unnecessary biopsies Total cost for negative tests

PSA test results Total number of positive PSA testsb Total number of detected prostate cancersc Total number of unnecessary biopsiesd Total number of negative PSA tests

Market share

Parameters

Table 3b Base case budget impact analysis results for PSA biopsy threshold of 4 ng ml1

Budget impact of a prostate cancer risk index MB Nichol et al

257

Prostate Cancer and Prostatic Diseases

Budget impact of a prostate cancer risk index MB Nichol et al

258

biopsy costs according to the assumptions previously defined. Costs are presented as total costs for prostate cancer detection, total costs per prostate cancer case detected and per-member-per-month (PMPM) cost. All costs are expressed in 2009 US dollars. Budget impact is calculated as the difference between the new and reference scenarios for all budget components and total costs. We presented the budget component by PSA test results and by service used. Finally, sensitivity analyses were performed to determine the influence of varied inputs on the total costs. We assumed the health plan perspective for a 1-year horizon. Costs were not discounted due to the short length of the analysis.35 Analyses were repeated for two models of PSA cutoffs (2 and 4 ng ml1) using Microsoft Office Excel 2003, and followed the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) principles of good practice for budget impact analysis.35 The study was approved by the Institutional Review Board of the principal investigator’s institution. Two coauthors from Beckman Coulter participated in study design and manuscript development. Final approval of the manuscript was the responsibility of the corresponding author.

Results Budget impact analysis The model evaluated the diagnostic pathway for 33 236 members receiving at least one PSA blood test for detecting prostate cancer. The new scenario resulted in a decrease of 20 and 5 detected prostate cancers, as well as 158 and 42 cases of unnecessary biopsies, which corresponded to the decrease in the expected total number of false positive tests by 3.0, and 2.8%, in models of PSA cutoff 2 and 4 ng ml1, respectively (Tables 3a and b). When the budget component was presented by PSA test results, the savings on the total costs for the individuals with a positive PSA test were $385 559 and $101 857 at the two thresholds, respectively. The total budget impact on total costs for individuals with a negative test ranged from $28 912 to $7638 (Figure 1). The incremental costs for PSA blood test ranged from $51 524 to $13 611; whereas cost savings were seen for office visits ($32 795 to $8664), laboratory tests ($1954 to $516), necessary biopsies ($41 582 to $10 985) and unnecessary biopsies ($331 840 to $87 665), depending on threshold. Ninety-three percent of the overall savings came from avoiding unnecessary biopsies. The savings on expected total 1-year costs for prostate cancer detection for each model were $356 647 (0.30 PMPM) and $94 219 (0.08 PMPM). The savings on total costs per prostate cancer case detected were $84 and $58 at the respective thresholds. Sensitivity analysis Changing the index cutoff score from 25 to 55 for the models of PSA cutoff 2 and 4 ng ml1 increased the savings to $1 027 009 and $271 314, respectively. The probability of having a positive PSA test, biopsy rate and biopsy costs were major factors that affected the cost Prostate Cancer and Prostatic Diseases

Figure 1 Data are presented as the comparison of total costs of prostate cancer detection between the reference and new scenarios by service category. (a) Displays the budget impact in the model with PSA cutoff 2 ng ml1. (b) Displays the budget impact in the model with PSA cutoff 4 ng ml1. PSA, prostate specific antigen.

differences significantly (Figure 2). Cost savings decreased when the screening age threshold was reduced to 40 years, the biopsy compliance rate decreased or when p2PSA cost increased. The incremental cost of generating the index could increase to $500 from the base case assumption of $72, yet there would still be cost savings at a PSA threshold of 4 ng ml1. Increasing the market share of the index strategy from 10 to 90% would increase total cost savings from $94 219 (0.08 PMPM) to $847 967 (0.71 PMPM), although this would also require a significant organizational investment for testing. When the index strategy in the new scenario has an equivalent market share of fPSA in the reference scenario, the savings of total cost would be $22 540 (0.02 PMPM) to $45 079 (0.04 PMPM) in PSA cutoff 2 ng ml1 for 10–20% market share, respectively.

Discussion Adding the index to the current PSA testing strategies for prostate cancer detection could produce total cost savings at both thresholds of PSA cutoff 2 ng ml1 (0. 30 PMPM) and 4 ng ml1 (0.08 PMPM). The index reduced the total number of both true and false positive tests in the tested population, which may decrease the number of office visits, laboratory tests and prostate

Budget impact of a prostate cancer risk index MB Nichol et al

259

Figure 2 One-way sensitivity analysis on total healthcare costs difference between the new and reference scenarios. (a) Displays the results for the model with PSA cutoff 2 ng ml1. (b) Displays the results for the model with PSA cutoff 4 ng ml1. fPSA, free PSA; PSA, prostate specific antigen; tPSA, total PSA.

biopsies. Where p2PSA is commercially available, the unit cost is B2–3 times higher than either PSA or fPSA, so the use of the index for prostate cancer detection increased the total blood test costs by $51 524 and $13 611 for the models with PSA cutoff of 2 and 4 ng ml1, respectively. However, these additional blood test costs were offset by the savings on the costs of office visits ($32 795 to $8664), laboratory tests ($1954 to $516) and biopsies ($373 422 to $98 650), depending on the thresholds. Comparing the models between PSA cutoff 2 and 4 ng ml1, the savings in the model of 2 ng ml1 were B2.6 times of those in the model of 4 ng ml1. Lowering the PSA cutoff increased test sensitivity, therefore increasing the total number of positive tests overall. However, compared with the reference scenario, the introduction of the index reduced more cases of positive tests in the 2 ng ml1 model than in the 4 ng ml1 model (that is, a difference of 178 and 47 cases between scenarios, respectively). This may have led to the higher savings in the threshold 2 ng ml1 analysis. Most of the savings (92%) were realized by reducing unnecessary biopsies through increased index specificity. The overall reduction in biopsies resulting from a decrease in the number of positive PSA tests yielded a modest 2% decline for both screening thresholds. There were few missed cases, but these produced 12% of the

overall savings. These missed cases came from two sources. First, the models were constructed using threshold values to determine a positive test. Other modeling methods, such as using a continuous cancer risk score to determine relative cancer risk in the model could limit missed cases. Second, false negative tests produced missed cases, which could be minimized by increasing specificity. Cases missed at initial screening generate relatively few poor outcomes, as the false negative cases are likely to be found in subsequent screenings.28 There is some evidence that missed prostate cancers have relatively low Gleason scores (average 6 (3 þ 3)).36 Moreover, some studies indicate that most cancers detected at 2 to 4 years after an initial screen (first round) will be curable.37–42 A study of an individualized screening algorithm showed that very few important prostate cancers, for which diagnosis at a subsequent screening visit might be too late for treatment with curative intent, would be missed. When reducing the screening age threshold to 40 years, consistent with the proposed guidelines from AUA, the total screening costs and the savings from adding the index were smaller than those for the screening population with age 50–75 years. Younger men of age 40–49 years old were less likely to have a PSA test and to have a positive test result than older men. The number of tested men and required biopsies were Prostate Cancer and Prostatic Diseases

Budget impact of a prostate cancer risk index MB Nichol et al

260

smaller in the screening population with lower age threshold, which produced less total screening costs and savings. This analysis has several limitations. First, we have not counted the potential clinical and economic impact from missed diagnoses due to PSA false negative test results, as we assumed that all negative PSA tests were correct. However, the missed cases are usually detected in subsequent screenings.28,36 Second, we modeled prostate cancer detection using biopsy threshold values instead of the probability of prostate cancer risk. Research suggests that the higher levels of PSA are associated with increased risk of cancer.32,43 Although these values are used to determine the need for biopsies, studies have demonstrated that the use of a 4 ng ml1 cutoff was unable to detect all prostate cancers.13,43 Moreover, some cancers detected at very low PSA levels have proven incurable.12 Although there is some evidence that the field is now using a 2 ng ml1 cutoff, lowering the biopsy threshold may increase detection of ‘indolent’ lesions.13,29,30,43–45 Therefore, the 2009 AUA PSA Best Practice Policy3 emphasizes that the prostate biopsy decision should consider PSA and digital rectal examination results, as well as multiple factors. Two recent studies22,23 reported that a mathematical formula combining total PSA, fPSA and p2PSA to calculate the index shows improved specificity for detecting prostate cancer. Jansen et al.22 reported a specificity of 31% for the index at 90% sensitivity, compared with total PSA (10%) and %fPSA (12%). Comprehensive prostate cancer risk calculators have been developed to assess individual risk32,46 of cancer or need for biopsy and follow-up. Although applying a risk score within the budget impact model may have provided a more precise economic evaluation, the lack of available data restricted such an application. In summary, the index shows considerable promise as a complementary approach to current prostate cancer screening strategies as one method for cost reduction. The 1-year budget impact model showed that the increased laboratory costs required for this application resulted in health care cost savings from reductions in physician office visits, and the avoidance of unnecessary biopsy procedures. Future studies should address the impact of the new screening strategy on identifying prostate cancers over multiple years, and the consequent treatment costs and mortality.

Conflict of interest Research funding for this study was provided by Beckman Coulter. Dr Michael B Nichol has received research grant from Beckman; Joanne Wu and Jae Jin An have been paid from the research funding from Beckman Coulter in connection with this paper; Joice Huang and Dwight Denham are both employees of Beckman Coulter.

Acknowledgements Partial contents of this research have been presented at the International Society for Pharmacoeconomics and Prostate Cancer and Prostatic Diseases

Outcomes Research (ISPOR) 15th Annual International Meeting, Atlanta, GA, May 2010. This study was funded by Beckman Coulter.

References 1 Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics, 2009. CA Cancer J Clin 2009; 59: 225–249. 2 Stark JR, Mucci L, Rothman KJ, Adami HO. Screening for prostate cancer remains controversial. BMJ 2009; 339: b3601. 3 Greene KL, Albertsen PC, Babaian RJ, Carter HB, Gann PH, Han M et al. Prostate specific antigen best practice statement: 2009 update. J Urol 2009; 182: 2232–2241. 4 Wolf AM, Wender RC, Etzioni RB, Thompson IM, D’Amico AV, Volk RJ et al. American Cancer Society guidelines for the early detection of cancer: update 2010. Ca Cancer J Clin 2010; 60: 70–98. 5 Bretton PR. Prostate-specific antigen and digital rectal examination in screening for prostate cancer—a community-based study. Southern Med J 1994; 87: 720–723. 6 Muschenheim F, Omarbasha B, Kardjian PM, Mondou EN. Screening for carcinoma of the prostate with prostate specific antigen. Ann Clin Lab Sci 1991; 21: 371–380. 7 Richie JP, Catalona WJ, Ahmann FR, Hudson MA, Scardino PT, Flanigan RC et al. Effect of patient age on early detection of prostate cancer with serum prostate-specific antigen and digital rectal examination. Urology 1993; 42: 365–374. 8 Brawer MK. Prostate-specific antigen: current status. Ca-Cancer J Clin 1999; 49: 264–281. 9 Thompson IM, Ankerst DP, Chi C, Lucia MS, Goodman PJ, Crowley JJ et al. Operating characteristics of prostate-specific antigen in men with an initial PSA level of 3.0 ng/ml or lower. JAMA 2005; 294: 66–70. 10 Nadler RB, Loeb S, Roehl KA, Antenor JA, Eggener S, Catalona WJ. Use of 2.6 Ng/Ml prostate specific antigen prompt for biopsy in men older than 60 years. J Urol 2005; 174: 2154–2157. 11 Makarov DV, Humphreys EB, Mangold LA, Walsh PC, Partin AW, Epstein JI et al. Pathological outcomes and biochemical progression in men with T1c prostate cancer undergoing radical prostatectomy with prostate specific antigen 2.6 to 4.0 vs 4.1 to 6.0 ng/ml. J Urol 2006; 176: 554–558. 12 Lucia MS, Darke AK, Goodman PJ, La Rosa FG, Parnes HL, Ford LG et al. Pathologic characteristics of cancers detected in The Prostate Cancer Prevention Trial: implications for prostate cancer detection and chemoprevention. Cancer Prev Res (Phila Pa) 2008; 1: 167–173. 13 Thompson IM, Ankerst DP. Prostate-specific antigen in the early detection of prostate cancer. CMAJ 2007; 176: 1853–1858. 14 Brawley OW, Ankerst DP, Thompson IM. Screening for prostate cancer. CA Cancer J Clin 2009; 59: 264–273. 15 Makarov DV, Loeb S, Getzenberg RH, Partin AW. Biomarkers for prostate cancer. Annu Rev Med 2009; 60: 139–151. 16 Bangma CH, van Schaik RH, Blijenberg BG, Roobol MJ, Lilja H, Stenman UH. On the use of prostate-specific antigen for screening of prostate cancer in European Randomised Study for Screening of Prostate Cancer. Eur J Cancer 2010; 46: 3109–3119. 17 Vickers A, Cronin A, Roobol M, Savage C, Peltola M, Pettersson K et al. Reducing unnecessary biopsy during prostate cancer screening using a four-kallikrein panel: an independent replication. J Clin Oncol 2010; 28: 2493–2498. 18 Sokoll LJ, Chan DW, Mikolajczyk SD, Rittenhouse HG, Evans CL, Linton HJ et al. Proenzyme PSA for the early detection of prostate cancer in the 2.5–4.0 ng/ml total PSA range: preliminary analysis. Urology 2003; 61: 274–276. 19 Khan MA, Partin AW, Rittenhouse HG, Mikolajczyk SD, Sokoll LJ, Chan DW et al. Evaluation of proprostate specific antigen for early detection of prostate cancer in men with a total prostate specific antigen range of 4.0 to 10.0 ng/ml. J Urol 2003; 170: 723–726.

Budget impact of a prostate cancer risk index MB Nichol et al 20 Sokoll LJ, Wang Y, Feng Z, Kagan J, Partin AW, Sanda MG et al. 2]proenzyme prostate specific antigen for prostate cancer detection: a national cancer institute early detection research network validation study. J Urol 2008; 180: 539–543. 21 Catalona WJ, Partin AW, Sanda MG, Wei JT, Klee GG, Bangma CH et al. A multicenter study of [-2]Pro-prostatespecific antigen combined with prostate specific antigen and free prostate specific antigen for prostate cancer detection in the 2.0 to 10.0 ng/ml prostate specific antigen range. J Urol 2011; 185: 1650–1655. 22 Jansen FH, van Schaik RH, Kurstjens J, Horninger W, Klocker H, Bektic J et al. Prostate-specific antigen (PSA) isoform p2PSA in combination with total PSA and free PSA improves diagnostic accuracy in prostate cancer detection. Eur Urol 2010; 57: 921–927. 23 Le BV, Griffin CR, Loeb S, Carvalhal GF, Kan D, Baumann NA et al. [-2]Proenzyme prostate specific antigen is more accurate than total and free prostate specific antigen in differentiating prostate cancer from benign disease in a prospective prostate cancer screening study. J Urol 2010; 183: 1355–1359. 24 Ross LE, Coates RJ, Breen N, Uhler RJ, Potosky AL, Blackman D et al. Prostate-specific antigen test use reported in the 2000 National Health Interview Survey. Prev Med 2004; 38: 732–744. 25 Ross LE, Berkowitz Z, Ekwueme DU. Use of the prostate-specific antigen test among US men: findings from the 2005 national health interview survey. Cancer Epidem Biomar 2008; 17: 636–644. 26 Lacher DA, Thompson TD, Hughes JP, Saraiya M. Total, free, and percent free prostate-specific antigen levels among U.S. men, 2001–04. Adv Data 2006; 379: 1–12. 27 Crawford ED, DeAntoni EP, Etzioni R, Schaefer VC, Olson RM, Ross CA. Serum prostate-specific antigen and digital rectal examination for early detection of prostate cancer in a national community-based program. The Prostate Cancer Education Council. Urology 1996; 47: 863–869. 28 Crawford ED, Abrahamsson PA. PSA-based screening for prostate cancer: how does it compare with other cancer screening tests? Eur Urol 2008; 54: 262–273. 29 Catalona WJ, Bartsch G, Rittenhouse HG, Evans CL, Linton HJ, Horninger W et al. Serum pro-prostate specific antigen preferentially detects aggressive prostate cancers in men with 2 to 4 ng/ml prostate specific antigen. J Urol 2004; 171 (Pt 1): 2239–2244. 30 Catalona WJ, Loeb S. Prostate cancer screening and determining the appropriate prostate-specific antigen cutoff values. J Natl Compr Canc Netw 2010; 8: 265–270. 31 Catalona WJ, Smith DS, Ratliff TL, Dodds KM, Coplen DE, Yuan JJ et al. Measurement of prostate-specific antigen in serum as a screening-test for prostate-cancer. N Engl J Med 1991; 324: 1156–1161. 32 Thompson IM, Ankerst DP, Chi C, Goodman PJ, Tangen CM, Lucia MS et al. Assessing prostate cancer risk: results from the Prostate Cancer Prevention Trial. J Natl Cancer Inst 2006; 98: 529–534. 33 Andriole GL, Levin DL, Crawford ED, Gelmann EP, Pinsky PF, Chia D et al. Prostate cancer screening in the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial: findings

34

35

36

37

38

39

40

41

42

43

44

45

46

from the initial screening round of a randomized trial. J Natl Cancer Inst 2005; 97: 433–438. Smith DS, Bullock AD, Catalona WJ. Racial differences in operating characteristics of prostate cancer screening tests. J Urol 1997; 158: 1861–1865. Mauskopf JA, Sullivan SD, Annemans L, Caro J, Mullins CD, Nuijten M et al. Principles of good practice for budget impact analysis: report of the ISPOR Task Force on good research practices—budget impact analysis. Value Health 2007; 10: 336–347. Wolters T, van der Kwast TH, Vissers CJ, Bangma CH, Roobol M, Schro¨der FH et al. False-negative prostate needle biopsies: frequency, histopathologic features, and follow-up. Am J Surg Pathol 2010; 34: 35–43. Hugosson J, Aus G, Lilja H, Lodding P, Pihl CG, Pileblad E. Prostate specific antigen based biennial screening is sufficient to detect almost all prostate cancers while still curable. J Urol 2003; 169: 1720. van der Cruijsen-Koeter IW, van der Kwast TH, Schroder FH. Interval carcinomas in the European Randomized Study of Screening for Prostate Cancer (ERSPC)-Rotterdam. J Natl Cancer Inst 95; 1462: 2003. Hoedemaeker RF, van der Kwast TH, Boer R, de Koning HJ, Roobol M, Vis AN et al. Pathologic features of prostate cancer found at population-based screening with a four-year interval. J Natl Cancer Inst 2001; 93: 1153. Postma R, Roobol M, Schroder FH, van der Kwast TH. Potentially advanced malignancies detected by screening for prostate carcinoma after an interval of 4 years. Cancer 2004; 100: 968. Hugosson J, Aus G, Lilja H, Lodding P, Pihl CG. Results of a randomized, population-based study of biennial screening using serum prostate-specific antigen measurement to detect prostate carcinoma. Cancer 2004; 100: 1397. Schro¨der FH, Bangma CH, Roobol MJ. Is it necessary to detect all prostate cancers in men with serum PSA levels o3.0 ng/ml? A comparison of biopsy results of PCPT and outcome-related information from ERSPC. Eur Urol 2008; 53: 901–908. 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 o ¼ 4.0 ng per milliliter. N Engl J Med 2004; 350: 2239–2246. 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: 469–473 discussion 473–4. Meeks JJ, Loeb S, Helfand BT, Kan D, Smith ND, Catalona WJ. Characteristics of prostate cancers detected at prostate specific antigen levels less than 2.5 ng/ml. J Urol 2009; 181: 2515–2518. Kaplan DJ, Boorjian SA, Ruth K, Egleston BL, Chen DY, Viterbo R et al. Evaluation of the prostate cancer prevention trial risk calculator in a high-risk screening population. Bju Int 2010; 105: 334–337.

261

Prostate Cancer and Prostatic Diseases