## Abstract

Background:

This analysis was carried out to evaluate the cost-effectiveness of adjuvant radiation therapy (ART) versus observation, using a decision analysis model based primarily upon the published results of the Southwest Oncology Group prospective trial (SWOG 8794).

Patients and methods:

A decision analysis model was designed to compare ART versus observation over a 10-year time horizon. Probabilities of treatment success, utilization of salvage treatments, and rates of adverse events were taken from published results of SWOG 8794. Cost inputs were based on 2010 Medicare reimbursement rates. Primary outcome measure was incremental cost per prostate-specific antigen (PSA) success (i.e. serum PSA level <0.4 ng/ml).

Results:

ART results in a higher PSA success rate than observation with probability of 0.43 versus 0.22. The mean incremental cost per patient for ART versus observation was $6023. The mean incremental cost-effectiveness ratio was$26 983 over the 10-year period.

Conclusions:

ART appears cost effective compared with observation based upon this decision analysis model. Future research should consider more costly radiation therapy (RT) approaches, such as intensity-modulated RT, and should evaluate the cost-effectiveness of ART versus early salvage RT.

## introduction

 Cost Variable Value Low High Source Adjuvant radiation therapy $11 295$9036 $13 554 CY 2010 Medicare reimbursement rates Management of Adverse Events Rectal bleeding treatment (2 colonoscopies + 1 YAG laser coagulation)$2660 $2128$3192 CY 2010 Medicare reimbursement rates Bladder neck contracture treatment (2 cystoscopies + cytoscopy with dilation) $4648$3718 $5577 CY 2010 Medicare reimbursement rates ADT (lupron)$8991 $7192$10 788 Krupski et al. [31] Fracture $14 000$11 200 $16 800 Krupski et al. [31] The low and high values shown were used as the range for one-way sensitivity analyses. ### discounting Because the model was over a time span of 10 years, we discounted our outcome measure at a rate of 3%. Costs that were incurred in the first year (ART) were not discounted. All other costs were discounted based on the estimated year they would be incurred in; bleeding and stricture were discounted at 2 years, and fracture and ADT were discounted at 5 years. ### incremental cost-effectiveness ratio The incremental cost-effectiveness ratio (ICER) represents the cost per additional PSA success for ART versus observation. The ICER was calculated as the mean difference in costs between ART and observation divided by the mean difference in effectiveness (probability of PSA success) of ART versus observation. ### sensitivity analysis A series of one-way sensitivity analyses were conducted to evaluate the relative importance of each cost and probability assumption in the model. The standard deviation, when not available, was assumed to be 10% of the mean. A range of two standard deviations (80%–120% of the mean) for each variable was used in the one-way sensitivity analyses (Tables 1 and 2). One-way sensitivity analyses only vary one input at a time. To account for overall variability in the model where each variable changes at the same time, we conducted a probabilistic sensitivity analysis using Monte Carlo simulation in TreeAge Pro Suite 2009 [33]. In a Monte Carlo simulation, each model input is assigned a distribution of values primarily based on the standard deviations (Tables 1 and 2). In cases where the calculated distribution was unavailable, information provided by experts (TNS, LGG, CL, EJT) was utilized. Monte Carlo simulation was then applied to the mean incremental cost and effect and the standard error of the means based on the 1000 microsimulations designed to represent 1000 iterations in which the values for all the variables were randomly selected within the designated distribution. ## results ART results in a higher PSA success rate than observation with a probability of 0.44 versus 0.21 over a 10-year period [23]. Therefore, the mean incremental effect was 0.23. Total cost of ART was$15 900 compared with costs in the observation group of $9876. The mean incremental cost for ART versus observation was$6023 (Table 3). Using the mean incremental cost and effectiveness, the ICER was $26 983 over the 10-year period (Table 3). On a per-year basis, over 10 years, the cost per additional PSA success achieved with ART was$2698.

Table 3.

Mean Incremental cost, Incremental effect, and incremental cost-effectiveness ratio (ICER) of adjuvant radiation therapy (ART) versus observation in the decision analysis model

 Strategy Cost Incremental cost Effect Incremental effect ICER Observation $9876 –– 0.21 –– –– ART$15 900 $6023 0.43 0.22$26 983
 Strategy Cost Incremental cost Effect Incremental effect ICER Observation $9876 –– 0.21 –– –– ART$15 900 $6023 0.43 0.22$26 983

### sensitivity analysis

One-way sensitivity analysis revealed that the model is most sensitive to the likelihood of treatment failure, ADT cost and costs of ART. For example, for the ART to cost as much as observation, then the cost of ART has to decrease to $2000 (from the baseline value of$11 295 in current model) (Figure 2).

The results from the Monte Carlo simulation (Figure 3) showed that ART continues to have higher incremental cost ($6061) and higher incremental effectiveness (0.22) than observation in all calculated conditions. Figure 2. Graph of the linear relationship of incremental cost of adjuvant radiation therapy (versus observation) with incremental cost-effectiveness ratio. The solid horizontal line demonstrates the cost at which the ICER is zero, meaning that adjuvant radiation therapy (RT) costs no more than observation. Figure 2. Graph of the linear relationship of incremental cost of adjuvant radiation therapy (versus observation) with incremental cost-effectiveness ratio. The solid horizontal line demonstrates the cost at which the ICER is zero, meaning that adjuvant radiation therapy (RT) costs no more than observation. Figure 3. Monte Carlo simulation to evaluate probabilistic sensitivity of the incremental cost-effectiveness ratio. The solid oval denotes the 95% confidence interval around the point estimates. Figure 3. Monte Carlo simulation to evaluate probabilistic sensitivity of the incremental cost-effectiveness ratio. The solid oval denotes the 95% confidence interval around the point estimates. ## discussion Based upon our decision analytic model, ART was both more costly and more effective than observation with each additional PSA success achieved at a cost of$26 983 over a 10-year time horizon. Sensitivity analyses confirmed that ART remained a more costly and more effective treatment than observation.

Standard cost-effectiveness analyses employ a quality-adjusted life year (QALY) as the primary outcome measure. However, since utilities are not available in the medical literature for the specific scenario of ART after RP, it was not possible to calculate QALYs for patients receiving ART (or SRT) after RP. A prior cost-effectiveness analysis of risk-prediction tools in selecting patients for early ART after RP used utilities from a group of 17 patients with intermediate-risk PC who received nonsurgical management with intensity-modulated RT (IMRT) as definitive therapy [22]. Due to significant anatomic changes after RP [34], with different rates of toxicity observed for definitive versus postoperative RT for PC [35, 36], it is not clear that utilities from definitive prostate RT should be applied to ART. Therefore, an alternate end point, cost per PSA success, was chosen for the current analysis. In addition to the data being available in the literature, we believe that PSA is an important outcome measure to payers, clinicians, and patients given that treatment decisions are often based on PSA values in this setting [16, 37]. PSA failure after prostatectomy is associated with higher rates of metastasis and mortality, as well as exposure to salvage treatments [25, 38]. Salvage PC treatments for PSA failure after RP are associated with significant adverse effects [30, 31], so the measure used in the current decision analysis model (cost per PSA success) is a relevant consideration. Although the recently published update of the SWOG 8794 trial showed improved overall survival with ART versus observation, overall survival was not reported by treatment arm for the subset of patients with confirmed undetectable PSA [12]. The other two randomized trials of ART showed improved biochemical progression-free survival but not overall survival [9, 10]. Therefore, the end point cost per PSA success permits the current analysis to reflect the shared conclusions of all three trials.

In our analysis, we used PSA success as our outcome measure rather then QALYS. Consequently, there is no published threshold value to determine cost-effectiveness using PSA success as an outcome measure. However, previous decision analytic modeling reports of IMRT [21], proton beam therapy [20], and a risk-prediction tool for ART decision-making [22] for PC while not using QALYS have compared results to the reference $50 000/QALY willingness-to-pay threshold. We believe that our ICER of$26 983 seems reasonable in consideration of the health gains achieved, though cost per PSA success may not be compared with cost per QALY. Furthermore, we believe that our findings lend further support to ART, which has also been shown to result in higher biochemical failure-free survival [9–11, 23], distant metastasis-free survival [12], and overall survival [12] in randomized trials of ART versus observation for patients with APFs.

The current decision analysis model adheres primarily to the design and outcomes of the SWOG 8794 trial and contains limitations regarding the magnitude and scope of cost inputs, as well as the relationship of the model to contemporary practice, which favors decisions of ART versus early salvage therapy rather than observation [37, 39]. Cost inputs for ART were based upon 2010 Medicare reimbursements rates for 3D-conformal radiation therapy (CRT) to a total dose of 64 Gy, similar to the planning and dose-fractionation schedule permitted on SWOG 8794. It is worth noting that reimbursement rates from private payers may exceed the rates used in the model. Furthermore, a national survey of radiation oncologists conducted by our group suggests that IMRT is used as the standard approach for post-RP RT by ∼80% of respondents with total doses commonly ∼68 Gy (manuscript in preparation). Therefore, the current model may significantly underestimate the costs of ART since IMRT is significantly more costly than 3D-CRT. For instance, based on Medicare calendar year 2010 reimbursement, a 38-fraction course of post-RP IMRT (68.4 Gy in 1.8 Gy per fraction) would cost $22 313, nearly two times higher than the estimated cost of$11 295 for 3D-CRT used in our model. Another limitation of the current analysis relates to the scope of cost inputs since the costs of symptom-management medications, outpatient visits, and hospital charges were not included in the model.

Finally, the current model was designed to test ART versus observation rather than ART versus SRT. When SWOG 8794 was designed, the study question reflected the state of PC at that time, and the investigators evaluated ART versus observation. However, much controversy regarding optimal management for PC after RP now centers instead upon ART versus early SRT [16]. Current practice patterns support the trend toward SRT: Ghia et al. recently reported that <20% of patients, who qualify, based on APFs, receive ART [13]. Although the decision analysis model of ART versus observation does not directly inform decisions regarding ART versus early SRT, this cost-effectiveness analysis nevertheless provides a foundation for considering the value of postprostatectomy RT for PC.

In conclusion, ART is an effective and cost-effective strategy for appropriately selected PC patients, based upon this decision analysis model of ART versus observation using published results of SWOG 8794. This model provides a baseline for considering similar management dilemmas in PC and for evaluating changes in the details of ART planning and delivery that might increase (IMRT and image guidance) or decrease (shorter fractionation schedules) the financial costs of RT. Future research should focus on the comparative effectiveness and cost-effectiveness of ART versus early SRT and should incorporate comprehensive, prospectively collected costs incurred by medication charges and outpatient health care utilization, as well as on quality of life gains.

## funding

Kimmel Cancer Center National Institutes of Health Cancer Center Core (P30CA56036).

## disclosure

The authors declare no conflict of interest.

We thank Eileen Comber, Department of Radiation Oncology, Kimmel Cancer Center, Thomas Jefferson University, for her assistance in identifying and collecting medical cost data for the analysis (she received no compensation for her assistance).

This work was presented in part at the 2010 American Society for Radiation Oncology Annual Meeting in San Diego, CA, 31 October–4 November 2010.

## References

1.
Jemal
A
Siegel
R
Xu
J
Ward
E
Cancer statistics, 2010
CA Cancer J Clin
,
2010
, vol.
60
(pg.
277
-
300
)
2.
Cooperberg
MR
Broering
JM
Kantoff
PW
Carroll
PR
Contemporary trends in low-risk prostate cancer: risk assessment and treatment
J Urol
,
2007
, vol.
178
(pg.
S14
-
S19
)
3.
Mariotto
AB
Yabroff
KR
Shao
Y
, et al.  .
Projections of the cost of cancer care in the United States: 2010-2020
J Natl Cancer Inst
,
2011
, vol.
103
(pg.
1
-
12
)
4.
Ollendorf
DA
Hayes
J
McMahon
P
, et al.  .
Management Options for Low-Risk Prostate Cancer: A Report on Comparative Effectiveness and Value
,
2010
Institute for Clinical and Economic Review, Massachusetts General Hospital
Boston, MA
(pg.
5
-
76
)
5.
Penson
DF
Chan
JM
Urologic Diseases in America Project
Prostate cancer
J Urol
,
2007
, vol.
177
(pg.
2020
-
2029
)
6.
Swanson
GP
Riggs
MW
Earle
JD
MG
Long-term follow-up of radical retropubic prostatectomy for prostate cancer
Eur Urol
,
2002
, vol.
42
(pg.
212
-
216
)
7.
Han
M
Partin
AW
Pound
CR
, et al.  .
Long-term biochemical disease-free and cancer-specific survival following anatomic radical retropubic prostatectomy. The 15-year Johns Hopkins experience
Urol Clin North Am
,
2001
, vol.
28
(pg.
555
-
565
)
8.
Swanson
GP
Riggs
M
Hermans
M
Pathologic findings at radical prostatectomy: risk factors for failure and death
Urol Oncol
,
2007
, vol.
25
(pg.
110
-
114
)
9.
Wiegel
T
Bottke
D
Steiner
U
, et al.  .
Phase III postoperative adjuvant radiotherapy after radical prostatectomy compared with radical prostatectomy alone in pT3 prostate cancer with postoperative undetectable prostate-specific antigen: ARO 96-02/AUO AP 09/95
J Clin Oncol
,
2009
, vol.
27
(pg.
2924
-
2930
)
10.
Bolla
M
van Poppel
H
Collette
L
, et al.  .
Postoperative radiotherapy after radical prostatectomy: a randomised controlled trial (EORTC trial 22911)
Lancet
,
2005
, vol.
366
(pg.
572
-
578
)
11.
Thompson
IM
Jr
Tangen
CM
J
, et al.  .
JAMA
,
2006
, vol.
296
(pg.
2329
-
2335
)
12.
Thompson
IM
Tangen
CM
J
, et al.  .
Adjuvant radiotherapy for pathological T3N0M0 prostate cancer significantly reduces risk of metastases and improves survival: long-term followup of a randomized clinical trial
J Urol
,
2009
, vol.
181
(pg.
956
-
962
)
13.
Ghia
AJ
Shrieve
DC
Tward
JD
Urology
,
2010
, vol.
76
(pg.
1169
-
1174
)
14.
Hoffman
KE
Nguyen
PL
Chen
MH
, et al.  .
Recommendations for post-prostatectomy radiation therapy in the United States before and after the presentation of randomized trial
J Urol
,
2011
, vol.
185(1)
(pg.
116
-
120
)
15.
Schreiber
D
Rineer
J
Yu
JB
, et al.  .
Analysis of pathologic extent of disease for clinically localized prostate cancer after radical prostatectomy and subsequent use of adjuvant radiation in a national cohort
Cancer
,
2010
, vol.
116
(pg.
5757
-
5766
)
16.
Nielsen
ME
Trock
BJ
Walsh
PC
J Natl Compr Canc Netw
,
2010
, vol.
8
(pg.
228
-
237
)
17.
Trabulsi
EJ
Valicenti
RK
Hanlon
AL
, et al.  .
A multi-institutional matched-control analysis of adjuvant and salvage postoperative radiation therapy for pT3-4N0 prostate cancer
Urology
,
2008
, vol.
72
(pg.
1298
-
1302
)
18.
Budiharto
T
Perneel
C
Haustermans
K
, et al.  .
A multi-institutional analysis comparing adjuvant and salvage radiation therapy for high-risk prostate cancer patients with undetectable PSA after prostatectomy
,
2010

19.
Fleming
C
Wasson
JH
Albertsen
PC
, et al.  .
A decision analysis of alternative treatment strategies for clinically localized prostate cancer. Prostate Patient Outcomes Research Team
JAMA
,
1993
, vol.
269
(pg.
2650
-
2658
)
20.
Konski
A
Speier
W
Hanlon
AL
, et al.  .
Is proton beam therapy cost effective in the treatment of adenocarcinoma of the prostate?
J Clin Oncol
,
2007
, vol.
25
(pg.
3603
-
3608
)
21.
Konski
A
Watkins-Bruner
D
Feigenberg
S
, et al.  .
Using decision analysis to determine the cost-effectiveness of intensity-modulated radiation therapy in the treatment of intermediate risk prostate cancer
Int J Radiat Oncol Biol Phys
,
2006
, vol.
66
(pg.
408
-
415
)
22.
Zubek
VB
Konski
A
Cost effectiveness of risk-prediction tools in selecting patients for immediate post-prostatectomy treatment
Mol Diag Ther
,
2009
, vol.
13
(pg.
31
-
47
)
23.
Swanson
GP
Hussey
MA
Tangen
CM
, et al.  .
Predominant treatment failure pattern in postprostatectomy patients is local: analysis of patterns of treatment failure in SWOG 8794
J Clin Oncol
,
2007
, vol.
25
(pg.
2225
-
2229
)
24.
Thompson
IM
Tangen
CM
Klein
EA
Is there a standard of care for pathologic stage T3 prostate cancer?
J Clin Oncol
,
2009
, vol.
27
(pg.
2898
-
2899
)
25.
Agarwal
PK
N
Konety
BR
, et al.  .
Treatment failure after primary and salvage therapy for prostate cancer: likelihood, patterns of care, and outcomes
Cancer
,
2008
, vol.
112
(pg.
307
-
314
)
26.
Grossfeld
GD
Stier
DM
Flanders
SC
, et al.  .
Use of second treatment following definitive local therapy for prostate cancer: data from the caPSURE database
J Urol
,
1998
, vol.
160
(pg.
1398
-
1404
)
27.
Mehta
SS
Lubeck
DP
N
, et al.  .
Patterns of secondary cancer treatment for biochemical failure following radical prostatectomy: data from CaPSURE
J Urol
,
2004
, vol.
171
(pg.
215
-
219
)
28.
Hayes
SB
Pollack
A
Parameters for treatment decisions for salvage radiation therapy
J Clin Oncol
,
2005
, vol.
23
(pg.
8204
-
8211
)
29.
Weinstein
MC
O'Brien
B
Horberger
J
, et al.  .
Principles of good practice for decision analytic modeling in health-care evaluation: report of the ISPOR Task Force on Good Research Practices–Modeling Studies
Value Health
,
2003
, vol.
64
(pg.
9
-
17
)
30.
Feng
M
Hanlon
AL
Pisansky
TM
, et al.  .
Predictive factors for late genitourinary and gastrointestinal toxicity in patients with prostate cancer treated with adjuvant or salvage radiotherapy
Int J Radiat Oncol Biol Phys
,
2007
, vol.
68
(pg.
1417
-
1423
)
31.
Krupski
TL
Foley
KA
Baser
O
, et al.  .
Health care cost associated with prostate cancer, androgen deprivation therapy and bone complications
J Urol
,
2007
, vol.
178
(pg.
1423
-
1428
)
32.
Frizzera
C
Sebelius
K
Medicare Program: changes to the hospital outpatient prospective payment system and CY 2010 payment rates; changes to the ambulatory surgical center payment system and CY 2010 payment rates. Final rule with comment period
Fed Regist
,
2009
, vol.
74
(pg.
60315
-
60983
)
33.
TreeAge Pro 2009 Suite [Computer Program]. In Edition 1.0.2
,
2009
Williamstown, MA
TreeAge Software, Inc.
34.
Sanguineti
G
Castellone
P
Foppiano
F
, et al.  .
Anatomic variations due to radical prostatectomy: impact on target volume definition and dose-volume parameters of rectum and bladder
Strahlenther Onkol
,
2004
, vol.
180
(pg.
563
-
572
)
35.
Chen
JC
Schultheiss
TE
Nguyen
KH
Wong
JY
Acute toxicity in definitive versus postprostatectomy image-guided radiotherapy for prostate cancer
Int J Radiat Oncol Biol Phys
,
2008
, vol.
71
(pg.
351
-
357
)
36.
Teh
BS
Wei-Yuan
M
Augspurger
ME
, et al.  .
Intensity modulated radiation therapy (IMRT) following prostatectomy: more favorable acute genitourinary toxicity profile compared to primary IMRT for prostate cancer
Int J Radiat Oncol Biol Phys
,
2001
, vol.
49
(pg.
465
-
472
)
37.
Trock
BJ
Han
M
Freedland
SJ
, et al.  .
Prostate cancer-specific survival following salvage radiotherapy vs observation in men with biochemical recurrence after radical prostatectomy
JAMA
,
2008
, vol.
299
(pg.
2760
-
2769
)
38.
Pound
CR
Partin
AW
Eisenberger
MA
, et al.  .
Natural history of progression after PSA elevation following radical prostatectomy
JAMA
,
1999
, vol.
281
(pg.
1591
-
1597
)
39.
Raldow
A
Hamstra
DA
Kim
SN
Yu
JB