• prostate cancer;
  • cardiovascular disease;
  • hormonal therapy;
  • radiation therapy;
  • decision analysis


  1. Top of page
  2. Abstract


Randomized trials have demonstrated improved survival when hormonal therapy (HT) is added to radiation therapy (RT) for high-risk prostate cancer. However, it is still unknown whether men who have a history of myocardial infarction (MI) or MI risk factors achieve a superior outcome from HT.


A Markov decision analysis model was used to compare quality-adjusted life expectancy (QALE) in men aged 50, 60, and 70 years who received RT and no HT, 6 months of HT (short-term), or 3 years of HT (long-term) for high-risk prostate cancer stratified by cardiac risk group.


In men with a history of MI, there was a decrease of 0.1 to 0.2 quality-adjusted life years and 0.5 to 0.6 quality-adjusted life years across all ages with short-term HT and long-term HT, respectively, compared with no HT. In men without MI, receipt of short-term or long-term HT was associated with a QALE benefit versus no HT in all cohorts. Among men without MI, the optimal duration of HT was a function of age and the number of MI risk factors. Long-term HT improved QALE (range, 1.4-5.4 years) for men aged 50 or 60 years except those with MI; whereas, for men aged 70 years with 4 cardiac risk factors, short-term and long-term HT yielded identical QALE.


Men who received RT for high-risk prostate cancer and had a history of MI experienced net harm when they received HT. Men without MI gained a QALE benefit from HT, even if they had up to 4 cardiac risk factors. The optimal duration of HT is a function of patient age and the number of cardiac risk factors. Cancer 2013. © 2013 American Cancer Society.


  1. Top of page
  2. Abstract

The addition of hormonal therapy (HT) to radiation therapy (RT) for the treatment of high-risk localized and locally advanced prostate cancer decreases prostate cancer-specific mortality and prolongs overall survival.1-8 However, several studies suggest harm with regard to the endpoints of cardiovascular morbidity or death in men who receive HT, especially those with a history of myocardial infarction (MI).4, 9-15 Because randomized controlled trials typically enroll otherwise healthy patients, presumably most enrolled patients did not have a previous MI, and this is not the subset in which it has been suggested that harm occurs with the receipt of HT.13 In the cohort of patients who had a previous MI, the potential risk of cardiac mortality may outweigh the survival benefit associated with HT. Furthermore, HT may increase the risk of developing MI in men without pre-existing disease,9 which may further counteract the disease control benefits of HT with longer follow-up.

An analysis of the Framingham Heart and Offspring Studies demonstrated that the degree of elevated cardiac risk factors—blood pressure, low-density and high-density lipoprotein cholesterol levels, glucose intolerance, and smoking—can be grouped into distinct categories (optimal, borderline, and elevated). These categories allow clinicians to estimate the overall risk of developing MI based on the number of elevated risk factors a patient has rather than calculating risk based on the numerical values of individual risk factors.16

In addition to the potential interaction between HT and patient comorbidities, the optimal duration of HT in men with locally advanced prostate cancer has been questioned. A modest but significant survival benefit has been demonstrated using 3 years of HT in these men.1

To our knowledge, there have been no modeling studies examining the impact of 6 months (short-term) or 3 years (long-term) of HT on quality-adjusted life expectancy (QALE) in men who have a history of MI or 1 or more risk factors. Therefore, we developed a Markov decision-analytic model to estimate the effect of HT receipt and its duration on QALE in men with high-risk prostate cancer who received HT and had have a history of MI or least 1 risk factor based on the Framingham studies.


  1. Top of page
  2. Abstract

We developed a state transition Markov model using TreeAge Pro version 2012 (TreeAge Software, Inc., Williamstown, Mass) to estimate QALE predominately in men with localized high-risk or locally advanced prostate cancer representing patients with a prostate-specific antigen level >20 ng/mL, or Gleason scores from 8 to 10, or 2002 American Joint Committee on Cancer clinical tumor (T) classification T2c to T4.0. It is noteworthy that, in the Trans-Tasman Radiation Oncology Group (TROG) 96.01 randomized trial, which was used to derive estimates of biochemical failure and distant metastases for the model, 83% of men had high-risk prostate cancer.5 Men enter the model at ages 50, 60, or 70 years and exit at the time death from prostate cancer, MI, or other causes. Men also enter the model with history of MI, 1 to 4 elevated cardiac risk factors, and borderline or optimal cardiac risk profiles.16 The cycle length was 1 month, and a half-cycle correction was applied.17 The model structure is illustrated in Figure 1.

thumbnail image

Figure 1. Schematic diagram of the model. NED indicates no evidence of disease; MI, myocardial infarction; CV, cardiovascular. Toxicity states include erectile dysfunction, urinary incontinence and gastrointestinal.

Download figure to PowerPoint

Initial Treatment

All men received radiotherapy with 66.0 to 70.4 gray (Gy) to the prostate and seminal vesicles in 1.8-Gy to 2.0-Gy fractions over 6.5 to 7.0 weeks with intensity-modulated radiation therapy, which is considered the standard of care, although patients in TROG 96.01 received conformal radiotherapy. In addition, half of the men in the model received either 6 months of HT, consisting of goserelin and flutamide, or 3 years of HT, consisting of 6 months of flutamide or bicalutamide and triptorelin followed by triptorelin alone for 2.5 years.1, 5 After treatment, men were at risk for biochemical failure for 10 years after treatment.1, 5

Salvage Therapy

Men who developed biochemical failure received lifelong HT.5 These men were at risk for progression to metastatic disease and death from high-risk prostate cancer. We assumed that the risk of distant metastases would decrease 6 years after initial treatment.5

Myocardial Infarction Risk

For men without a history of MI, the risk of MI was based on age and the number elevated risk factors.16 In addition, men who received HT had an elevated risk of developing MI and dying from MI. These risks were elevated for 1.5 years if men received short-term HT and for 2 years if they received long-term HT, based on the duration of testosterone suppression.9, 10, 18, 19 Men with a history of MI were exposed to the same increased risk of dying from MI if they received HT.10 In men who received salvage therapy, the elevated risks of developing MI and MI mortality remained elevated for life.

Model Inputs

Model inputs and assumptions are summarized in Tables 1 and 2. The parameter estimates used in the model are taken from studies comprising level 1 evidence that is the basis for radiotherapeutic management of prostate cancer and its known toxicities. The risks of biochemical failure and distant metastasis with and without 6 months of HT were derived from the TROG 96.01 randomized trial.5 The impact of extending HT to 3 years on biochemical failure was derived from the European Organization for Research and Treatment of Cancer (EORTC) 22961 randomized trial.1 The time interval from the development of distant metastases to death was derived from a large surgical experience at a center of excellence.20 Men who developed MI faced an annual probability of death of 11%, which was a weighted average of mortality after ST elevation and non-ST elevation myocardial infarction in the GUSTO-IIb trial (Global Use of Strategies to Open Occluded Arteries in Acute Coronary Syndromes).21 The risk of MI for men who received HT was assumed to be 1.27 times higher than the risk for men who did not receive HT.9 If men who received HT developed MI or if men with a history of MI received HT, then the risk of death because of MI was assumed to be 1.96 times higher than the risk for men with MI who did not receive HT.10

Table 1. Disease-Related Probabilities
VariableBase-Case EstimateSDRange Used in Sensitivity AnalysisDistribution for PSAAlphaBeta or LambdaReference
  1. Abbreviations: GI, gastrointestinal; HR, hazard ratio; HT, hormonal therapy; MI, myocardial infarction; PSA, prostate-specific antigen; SD, standard deviation.

Annual probability of developing MI
Aged 50 y       
 Optimal0.0000.0000.000   Vasan 200516
 Borderline0.0030.0010.002-0.004Beta27.78790.8Vasan 200516
 1 High0.00690.0010.005-0.009Beta85.412187.9Vasan 200516
 2 High0.0120.0010.010-0.015Beta113.59102.4Vasan 200516
 3 High0.0200.0020.016-0.025Beta82.63924.1Vasan 200516
 4 High0.0320.0070.019-0.046Beta23.5702.0Vasan 200516
Aged 60 y
 Optimal0.0000.0000.000   Vasan 200516
 Borderline0.0050.0010.003-0.007Beta24.55153.7Vasan 200516
 1 High0.0100.0010.008-0.013Beta78.27375.8Vasan 200516
 2 High0.0180.0020.015-0.022Beta114.66053.6Vasan 200516
 3 High0.0310.0030.024-0.038Beta80.62513.2Vasan 200516
 4 High0.0480.0100.030-0.070Beta20.6402.5Vasan 200516
Aged 70 y
 Optimal0.0000.0000.000   Vasan 200516
 Borderline0.0070.0020.004-0.011Beta53.93353.3Vasan 200516
 1 High0.0160.0020.011-0.020Beta66.62309.0Vasan 200516
 2 High0.0280.0030.021-0.035Beta50.31022.6Vasan 200516
 3 High0.0460.0060.034-0.059Beta15.1191.2Vasan 200516
 4 High0.0710.0180.044-0.108Beta22.13071.3Smith 201113
Annual probability of developing PSA failure0.150.0100.13-0.17Beta194.61102.8Denham 20115
Annual probability of developing distant metastases
 Years 1 through 60.130.0800.09-0.18Beta13.525.6Denham 20115
 Year 7 and later0.02     Denham 20115
Annual probability of death after distant metastases0.300.1160.21-0.54Beta4.710.9Pound 199920
HR for developing biochemical failure with HT
 HT for 6 mo0.570.0880.46-1.0Gamma42.173.9Denham 20115
 HT for 3 y0.380.0290.32-1.0Gamma168.6443.7Bolla 20091
HR for developing distant metastases with HT0.680.1700.45-1.0Gamma15.923.4Denham 20115
HR for developing MI with HT1.270.0891.05-1.53Gamma202.6159.5Keating 20109
HR for the risk of death from MI with HT1.960.4491.00-3.71Gamma19.19.7Nanda 200910
Annual probability of death from MI0.1030.0050.10-0.12Beta512.74208.9Armstrong 199821
Duration of testosterone suppression
 HT for 6 mo1.50.0540-5Beta388.3129.4D'Amico 200918
 HT for 3 y2.00.0740-5Beta723.3361.6Kaku 200619
HR ratio for the impact of dose escalation on biochemical failure1.00.0910.5-1.0Gamma30.260.4Zietman 200528
Table 2. Toxicity Probabilities and Utilities
VariableBase-Case EstimateSDRange Used in Sensitivity AnalysisDistribution for PSAAlphaBetaReference
  1. aAbbreviations: GI, gastrointestinal; MI, myocardial infarction; NED, no evidence of disease; PSA, prostate-specific antigen; SD, standard deviation.

Annual probabilities of toxicities       
 Impotence0.1170.0820.031-0.120Beta1.8813.28Hayes 201025
 Urinary incontinence0.0390.0400.020-0.077Beta3.8091.20Hayes 201025
 GI0.0300.0300.010-0.039Beta8.70281.30Hayes 201025
 NED0.810.180.76-1.0Beta3.040.71Stewart 200526
 MI0.850.0810.75-1.0Beta15.552.74Lazar 201127
 Biochemical failure0.670.240.60-1.0Beta1.900.94Stewart 200526
 Distant metastases0.250.110.22-1.0Beta3.6210.87Stewart 200526
 Impotence0.890.160.86-1.0Beta2.510.31Stewart 200526
 Urinary incontinence0.830.210.79-1.0Beta1.830.37Stewart 200526
 GI problems0.710.260.67-1.0Beta1.450.59Stewart 200526
 Impotence and urinary incontinence0.790.230.72-1.0Beta1.690.45Stewart 200526
 Impotence and GI0.570.260.49-1.0Beta1.501.13Stewart 200526
 Urinary incontinence and GI0.640.330.32-1.0Beta0.710.40Stewart 200526
 Impotence, urinary incontinence, and GI0.450.310.36-1.0Beta0.710.87Stewart 200526

To ensure that the selected inputs would adequately model the natural history of disease, the rates of biochemical failure and distant metastases with and without HT were calculated. Initial probability estimates were calibrated so that model output, including the rates of biochemical failure and distant metastases, matched results of the TROG 96.01 and EORTC 22961 trials.1, 5

Age-specific rates of death from prostate cancer and MI were subtracted from overall age-specific death rates for men from US life tables to generate age-specific death rates from other causes.22-24 The model was run until patients reached age 100 years or until a cycle reward was <0.001 quality-adjusted life years (QALYs).

Complications and Adverse Effects

We assumed that men would be at risk for developing treatment-related complications within 5 years after initial treatment and that all complications would remain stable thereafter. Men who received salvage therapy were assumed to be impotent for life. Complications included urinary incontinence, bowel disturbances, and long-term impotence that are grade 2 or greater on the Radiation Therapy Oncology Group or Common Toxicity Criteria scales. Complication probabilities were based on a meta-analysis of complications from intensity-modulated radiation therapy.25


A utility is a quantitative assessment of a given health state and ranges from 0 (death) to 1 (healthy). For example, the states of distant metastases and no evidence of disease have utility values of 0.25 and 0.81, respectively (Table 2). Utilities associated with prostate cancer states were based on a standard-gamble in men aged ≥60 years, half of whom had been diagnosed with prostate cancer.26 Utilities associated with MI were based on the Coronary Heart Disease Policy Model.27 Because it is believed that utilities in prostate cancer are multiplicative,26 a patient's utility was the product of a patient's prostate cancer state, receipt of HT, presence of MI, and treatment complications.

Sensitivity and Probabilistic Sensitivity Analyses

One-way sensitivity analyses, in which 1 parameter was varied while all others were held constant, were conducted around all model parameters to evaluate the extent to which parameter variability influenced the model results. In addition, we performed a sensitivity analysis to explore the effects of radiation dose escalation on biochemical failure.28 Ranges were based on 95% confidence intervals and are listed in Tables 1 and 2. Utilities were studied to an upper limit of 1, representing the case of perfect health state equivalence. Variable thresholds for which the preferred strategy (defined by the highest gain in QALE) shifted were identified.

We performed probabilistic sensitivity analyses on men aged 60 years to evaluate the impact of uncertainty of individual model parameters. In probabilistic sensitivity analysis, the model is run thousands of times, with each run using the value of a parameter (such as recurrence rate) that is drawn from prespecified distributions based on literature estimates of parameter uncertainty. Parameters that were identified as sensitive in 1-way sensitivity analysis were included in the analysis. Beta distributions were generated for utilities, and gamma distributions were used for hazard ratios. The hazard ratio for biochemical failure with long-term HT was calculated de novo. We constructed a distribution based on bounds 15% above and below its value of 0.38. The model was run over 10,000 trials for the analysis.


  1. Top of page
  2. Abstract

Quality-Adjusted Life Years Stratified by Age, Cardiac Risk Factors, and Hormonal Therapy Duration

In all cohorts, regardless of age, omission of HT was preferred for men with a history of MI. Short-term HT resulted in a decrease of 0.1 to 0.2 QALYs, whereas long-term HT resulted in a decrease of 0.5 to 0.6 QALYs (Table 3).

Table 3. Quality-Adjusted Life Expectancy by Age and Framingham Risk Group
 QALE in Years (Incremental QALYs)
Framingham Risk GroupNo HTHT for 6 MonthsHT for 3 Years
  • Abbreviations: MI, myocardial infarction; HT, hormonal therapy; QALE, quality-adjusted life expectancy; QALYs, quality-adjusted life years.

  • a

    Preferred strategy.

Starting age: 50 y   
 Optimal1215.8 (3.8)17.4 (5.4)a
 Borderline11.514.9 (3.4)16.3 (4.8)a
 1 High10.914 (3.1)15.2 (4.3)a
 2 High10.212.9 (2.7)13.9 (3.7)a
 3 High9.311.6 (2.3)12.4 (3.1)a
 4 High8.510.3 (1.8)10.8 (2.3)a
 MI3.3a3.2 (0.1)2.8 (0.5)
Starting age: 60 y   
 Optimal9.712.2 (2.5)13.2 (3.5)a
 Borderline9.311.6 (2.3)12.5 (3.2)a
 1 High8.911 (2.1)11.7 (2.8)a
 2 High8.410.2 (1.8)10.8 (2.4)a
 3 High7.79.3 (1.6)9.6 (1.9)a
 4 High7.18.3 (1.2)8.5 (1.4)a
 MI3.2a3.1 (0.1)2.7 (0.5)
Starting age: 70 y   
 Optimal7.48.8 (1.4)9.3 (1.9)a
 Borderline7.18.5 (1.4)8.9 (1.8)a
 1 High6.98.1 (1.2)8.4 (1.5)a
 2 High6.67.7 (1.1)7.9 (1.3)a
 3 High6.27.1 (0.9)7.2 (1.0)a
 4 High5.86.5 (0.7)a6.5 (0.7)
 MI3.0a2.8 (0.2)2.4 (0.6)

However, for all other men, some duration of HT led to a QALE benefit, with the optimal duration a function of age and comorbidity. For example, in men ages 50 and 60 years, long-term HT was the preferred strategy (defined as the strategy estimated to yield the highest QALE) in all men except those with a history of MI, producing a benefit of 1.4 to 5.4 QALYs (Table 3). In men aged 70 years, long-term HT was only the preferred strategy for those with optimal, borderline, or, at most, 3 risk factors, producing 1.0 to 1.9 QALYs; whereas short-term HT was identified as equivalent to long-term HT in men aged 70 years who had 4 risk factors, producing 0.7 QALYs.

Evaluation of Key Model Parameters

One-way sensitivity analyses revealed thresholds for several model parameters. In men with a history of MI, the hazard ratio for an increased risk of death from MI with HT was the only significant model parameter (Table 4). At values greater than 1.70 (base case value, 1.96), the omission of HT was favored. This also was the only sensitivity analysis that significantly influenced the decision between short-term HT versus the omission of HT. In all cohorts other than men with a history of MI, various thresholds were encountered for the hazard ratios associated with HT and biochemical failure and distant metastases as well the utility associated with the biochemical failure health state. Finally, a threshold of 0.71 for the hazard ratio for dose escalation was established for men with 4 risk factors, raising the possibility that higher doses of radiation could be preferred over longer durations of HT (Table 4). These cutoff points split the optimal therapy into short-term versus long-term HT.

Table 4. Results of One-Way Sensitivity Analyses
Framingham Risk GroupVariableThresholdPreferred HT Strategy Below ThresholdPreferred HT Strategy Above Threshold
  1. Abbreviations: MI, myocardial infarction; HR, hazard ratio; HT, hormonal therapy.

OptimalHR for 3 y of HT on biochemical failure0.524Long-termShort-term
BorderlineHR for 3 y of HT on biochemical failure0.512Long-termShort-term
1 HighHR for 3 y of HT on biochemical failure0.502Long-termShort-term
2 HighHR for 3 y of HT on biochemical failure0.487Long-termShort-term
3 HighHR for 3 y of HT on biochemical failure0.463Long-termShort-term
3 HighHR for 6 mo of HT on biochemical failure0.488Short-termLong-term
4 HighHR for 3 y of HT on biochemical failure0.428Long-termShort-term
4 HighHR for dose escalation0.710Short-termLong-term
4 HighUtility biochemical failure0.974Long-termShort-term
MIHR for death from MI with HT1.695Short-termNo HT

Although there have been several studies evaluating the impact adding HT to RT for high-risk prostate cancer, we chose the TROG 96.01 and EORTC 22961 randomized trials for our model, because they demonstrated overall survival benefits with short-term and long-term HT. We used sensitivity analyses to test the reductions in biochemical failure based on the RTOG 9202 and 8610 trials and observed no changes in our results.29, 30

Probabilistic Sensitivity Analysis

On probabilistic sensitivity analysis, the preferred strategies were stable, and the magnitude of benefits was essentially unchanged relative to the base case (Table 5). Like in the base case, men with a history of MI experienced the largest decrement of QALE with HT. Similarly, in men aged 60 years with 4 cardiac risk factors, short-term HT was preferred over long-term HT. Taken together, the probabilistic sensitivity analysis supported the model results, emphasizing that uncertainty in key model parameters does not affect its results.

Table 5. Probabilistic Sensitivity Analysis: Starting Age, 60 Years
 No HT6 Months of HT3 Years of HT
Framingham Risk GroupMean ± SD95% CIMean ± SD95% CIMean ± SD95% CI
  1. Abbreviations: MI, myocardial infarction; CI, confidence interval; HT, hormonal therapy; SD, standard deviation.

Optimal9.7 ± 1.17.4-11.212.3 ± 1.19.8-14.013.2 ± 0.811.4-14.3
Borderline9.3 ± 1.07.2-10.711.7 ± 1.09.4-13.212.5 ± 0.711.0-13.5
1 High8.9 ± 0.97.0-10.111.0 ± 1.09.0-12.411.7 ± 0.610.4- 12.6
2 High8.4 ± 4.86.7-9.510.2 ± 0.88.5-11.510.8 ± 0.59.6-11.5
3 High7.7 ± 0.76.3-8.79.3 ± 0.77.8-10.39.6 ± 0.48.7-10.3
4 High7.1 ± 0.55.9-7.98.3 ± 0.57.1-9.28.5 ± 0.37.7-9.3
MI3.2 ± 0.22.8-3.63.1 ±0.42.4-3.82.7 ± 0.51.9-3.7


  1. Top of page
  2. Abstract

Given the delicate balance between prostate cancer and cardiovascular outcomes, we observed that the patient's age and basic cardiac risk factors can be used to derive a theoretical model for the optimal use of HT. In men with a history of MI, QALE decreases with HT in men ages 50 to 70 years. However, we observed that short-term HT was equivalent to short-term HT in men aged 70 years with 4 risk factors. Long-term HT led to the optimal outcome in all other age and risk factor strata.

Indeed, exploring ranges of values for model parameters in 1-way sensitivity analyses indicated that the results were consistent over a wide variety of assumptions. Even in men with a history of MI, there may be a potential benefit with HT, if the model assumes that the hazard of death from MI that is <1.70. Other studies have reported hazard ratios of 1.2,12 1.94,11 and 2.614 in addition to the 1.9610 hazard ratio used in the current study for the risk of cardiovascular mortality with HT. Therefore, under some scenarios, HT would be favored even in this most unfavorable risk group. Otherwise, omission of HT was always preferred in men with MI, regardless of age. Probabilistic sensitivity analysis validated the minimal impact of uncertainty of the model output and demonstrated that, even when considering uncertainty with respect to the risk of cardiovascular mortality from HT, the model results remain robust.

Men with a history of MI likely have a shorter life expectancy relative to men without previous MI. Consequently, the harms associated with HT—including both side effects and potential risks in cardiac mortality—outweigh the improvements in prostate cancer survival. The clinical significance of this finding is that the survival benefits observed from adding HT to RT for men with high-risk or locally advanced prostate cancer in randomized trials probably would not hold in men with a history of MI. For men with 4 risk factors, the utility of administering long-course versus short-course HT and its impact on QALE was age-dependent and favored short-course HT more often as men age. This can be explained from greater competing risks of death and overall shorter life expectancy in older men. For all other men, even those with several risk factors, administering long-term HT was associated with a QALE benefit.

This study had several limitations. First, the underlying risk of MI development and mortality was derived from the Framingham studies, which are relatively old. If modern cardiac management (eg, intensive workup to modify risk factors like hypertension and hypercholesterolemia and assessment for reversible myocardial ischemia) reduces the mortality of MI compared with our reference Framingham cohort, then more men than we have predicted may benefit from HT. Indeed, because our model confirmed the intuitive progressive benefit for men with lower MI risk, we recommend careful cardiovascular examination, including a search for reversible risk factors and for extensive asymptomatic disease, for men with high-risk prostate cancer.31

Moreover, despite the revelation in large database outcomes studies of cardiovascular morbidity and mortality from HT among men with prostate cancer, randomized controlled trials have not reported an increased risk of cardiovascular mortality among men with high-risk prostate cancer who received HT and RT compared with RT alone. A simple explanation for this discrepancy is that those studies did not evaluate cardiovascular mortality in the specific group of men with a documented history of MI.1-8 Similarly, a recent meta-analysis of randomized trials of HT demonstrated that HT was not associated with an increased risk of cardiovascular death. However, none of the included trials stratified patients according to pre-existing cardiovascular comorbidity. The meta-analysis demonstrated that men with moderate-to-severe cardiac comorbidity had a trend toward poorer overall survival with HT.15 Unfortunately, there are no current randomized trials that prestratify men according to cardiovascular risk. Therefore, it remains uncertain whether men with cardiac comorbidity are at an increased risk of cardiovascular mortality with HT. A related concern is whether our results are relevant to the population that may be eligible for HT. Actually, in the study by Nanda et al, in >5000 men who were managed in a community-based cohort that likely represented most of the US population, 52% had no risk factors, 46% had 1 risk factor, and 2% had previous MI or CHF.10

Third, the model was sensitive to several health state utility evaluations. In particular, the quality-of-life evaluation of patients with biochemical recurrence did indicate an influence of short-term versus long-term HT in patients with higher risk factors, in that lower values favored long-term HT, because that approach reduces the likelihood of prostate-specific antigen elevation. Because health state evaluations are patient-dependent, it is critical to assess patient preferences in deciding on the duration of HT in the few strata of men whose decisions are affected by these values.

In addition, the estimates of biochemical failure and distant metastases were based on the receipt of radiation doses from 66 to 0.4 Gy.5 Because it has been demonstrated that radiation dose escalation decreases biochemical failure, we used sensitivity analysis to explore the theoretical effects of dose escalation.28 In fact, dose escalation did not make a meaningful difference in the results except for men with 4 risk factors, in whom the improved local control from dose intensification allowed a shorter course of HT provided the higher dose reduced biochemical failure by ≥29%. In this subset alone, the improvements in biochemical failure with dose escalation offset the increased morbidity and moderate disease control benefits from long-term HT.

Finally, we caution that a Markov model does not substitute for a prospective randomized trial. A randomized trial comparing HT of various duration with RT alone stratified according to cardiac risk is the ideal trial for assessing the impact of HT on cardiovascular events. However, because several published trials previously investigated the length of HT, it is unlikely that an additional study will be performed, emphasizing the importance of decision analysis.

In conclusion, most men with high-risk prostate cancer will benefit from long-term HT. However, compared with long-term HT, short-term HT maximizes QALE in a subset of men with MI risk factors. Given the potential interaction between HT and cardiovascular disease, it is important for the patient to recognize the importance of achieving optimal cardiovascular health. We recommend pretherapy consultation with a cardiologist to quantify and, if possible, reverse MI risk factors.


  1. Top of page
  2. Abstract

No specific funding was disclosed.


Samuel Z. Goldhaber has received research grants from Daiichi Sankyo, Eisai, EKOS, Johnson & Johnson, and Sanofi Aventis and is a paid consultant for Baxter, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi Sankyo, Eisai, Medscape, Merck, Portola, and Sanofi Aventis.


  1. Top of page
  2. Abstract
  • 1
    Bolla M, de Reijke TM, Van Tienhoven G, et al. Duration of androgen suppression in the treatment of prostate cancer. N Engl J Med. 2009; 360: 25162527.
  • 2
    Bolla M, Van Tienhoven G, Warde P, et al. External irradiation with or without long-term androgen suppression for prostate cancer with high metastatic risk: 10-year results of an EORTC randomised study. Lancet Oncol. 2010; 11: 10661073.
  • 3
    Crook J, Ludgate C, Malone S, et al. Final report of multicenter Canadian phase III randomized trial of 3 versus 8 months of neoadjuvant androgen deprivation therapy before conventional-dose radiotherapy for clinically localized prostate cancer. Int J Radiat Oncol Biol Phys. 2009; 73: 327333.
  • 4
    D'Amico AV, Chen M-H, Renshaw AA, Loffredo M, Kantoff PW. Androgen suppression and radiation vs radiation alone for prostate cancer: a randomized trial. JAMA. 2008; 299: 289295.
  • 5
    Denham JW, Steigler A, Lamb DS, et al. Short-term neoadjuvant androgen deprivation and radiotherapy for locally advanced prostate cancer: 10-year data from the TROG 96.01 randomised trial. Lancet Oncol. 2011; 12: 451459.
  • 6
    Laverdiere J, Gomez JL, Cusan L, et al. Beneficial effect of combination hormonal therapy administered prior and following external beam radiation therapy in localized prostate cancer. Int J Radiat Oncol Biol Phys. 1997; 37: 247252.
  • 7
    Pilepich MV, Caplan R, Byhardt RW, et al. Phase III trial of androgen suppression using goserelin in unfavorable-prognosis carcinoma of the prostate treated with definitive radiotherapy: report of Radiation Therapy Oncology Group Protocol 85-31. J Clin Oncol. 1997; 15: 10131021.
  • 8
    Widmark A, Klepp O, Solberg A, et al. Endocrine treatment, with or without radiotherapy, in locally advanced prostate cancer (SPCG-7/SFUO-3): an open randomised phase III trial. Lancet. 2009; 373: 301308.
  • 9
    Keating NL, O'Malley AJ, Freedland SJ, Smith MR. Diabetes and cardiovascular disease during androgen deprivation therapy: observational study of veterans with prostate cancer. J Natl Cancer Inst. 2010; 102: 3946.
  • 10
    Nanda A, Chen M-H, Braccioforte MH, Moran BJ, D'Amico AV. Hormonal therapy use for prostate cancer and mortality in men with coronary artery disease-induced congestive heart failure or myocardial infarction. JAMA. 2009; 302: 866873.
  • 11
    Punnen S, Cooperberg MR, Sadetsky N, Carroll PR. Androgen deprivation therapy and cardiovascular risk. J Clin Oncol. 2011; 29: 35103516.
  • 12
    Saigal CS, Gore JL, Krupski TL, Hanley J, Schonlau M, Litwin MS. Androgen deprivation therapy increases cardiovascular morbidity in men with prostate cancer. Cancer. 2007; 110: 14931500.
  • 13
    Smith MR, Klotz L, van der Meulen E, Colli E, Tanko LB. Gonadotropin-releasing hormone blockers and cardiovascular disease risk: analysis of prospective clinical trials of degarelix. J Urol. 2011; 186: 18351842.
  • 14
    Tsai HK, D'Amico AV, Sadetsky N, Chen M-H, Carroll PR. Androgen deprivation therapy for localized prostate cancer and the risk of cardiovascular mortality. J Natl Cancer Inst. 2007; 99: 15161524.
  • 15
    Nguyen PL, Je Y, Schutz Y, et al. Association of androgen deprivation therapy with cardiovascular death in patients with prostate cancer: a meta-analysis of randomized trials. JAMA. 2011; 306: 23592366.
  • 16
    Vasan RS, Sullivan LM, Wilson PWF, et al. Relative importance of borderline and elevated levels of coronary heart disease risk factors. Ann Intern Med. 2005; 142: 393402.
  • 17
    Naimark DMJ, Bott M, Krahn M. The half-cycle correction explained: 2 alternative pedagogical approaches. Med Decis Making. 2008; 28: 706712.
  • 18
    D'Amico AV, Chen M-H, Renshaw AA, Loffredo M, Kantoff PW. Interval to testosterone recovery after hormonal therapy for prostate cancer and risk of death. Int J Radiat Oncol Biol Phys. 2009; 75: 1015.
  • 19
    Kaku H, Saika T, Tsushima T, et al. Time course of serum testosterone and luteinizing hormone levels after cessation of long-term luteinizing hormone-releasing hormone agonist treatment in patients with prostate cancer. Prostate. 2006; 66: 439444.
  • 20
    Pound CR, Partin AW, Eisenberger MA, Chan DW, Pearson JD, Walsh PC. Natural history of progression after PSA elevation following radical prostatectomy. JAMA. 1999; 281: 15911597.
  • 21
    Armstrong PW, Fu Y, Chang WC, et al. Acute coronary syndromes in the GUSTO-IIb trial: prognostic insights and impact of recurrent ischemia. The GUSTO-IIb Investigators. Circulation. 1998; 98: 18601868.
  • 22
    Altekruse SF, Kosary CL, Krapcho M, et al., eds. Trends in SEER Incidence and US Mortality. SEER Cancer Statistics Review 1975-2007 [based on November 2009 SEER data submission, posted to the SEER website 2010]. Bethesda, MD: National Cancer Institute; 2009. Available at: [Accessed October 28, 2011.]
  • 23
    Keenan NL, Shaw KM. Coronary heart disease and stroke deaths—United States, 2006. MMWR. 2011; 60( suppl): 6266.
  • 24
    Xu J, Kochanek KD, Murphy SL, et al. Deaths: final data for 2007. Natl Vital Stat Rep. 2010; 58: 1136.
  • 25
    Hayes JH, Ollendorf DA, Pearson SD, et al. Active surveillance compared with initial treatment for men with low-risk prostate cancer: a decision analysis. JAMA. 2010; 304: 23732380.
  • 26
    Stewart ST, Lenert L, Bhatnagar V, Kaplan RM. Utilities for prostate cancer health states in men aged 60 and older. Med Care. 2005; 43: 347355.
  • 27
    Lazar LD, Pletcher MJ, Coxson PG, Bibbins-Domingo K, Goldman L. Cost-effectiveness of statin therapy for primary prevention in a low-cost statin era. Circulation. 2011; 124: 146153.
  • 28
    Zietman AL, DeSilvio ML, Slater JD, et al. Comparison of conventional-dose vs high-dose conformal radiation therapy in clinically localized adenocarcinoma of the prostate: a randomized controlled trial. JAMA. 2005; 294: 12331239.
  • 29
    Roach M3rd, Bae K, Speight J, et al. Short-term neoadjuvant androgen deprivation therapy and external-beam radiotherapy for locally advanced prostate cancer: long-term results of RTOG 8610. J Clin Oncol. 2008; 26: 585591.
  • 30
    Hanks GE, Pajak TF, Porter A, et al. Phase III trial of long-term adjuvant androgen deprivation after neoadjuvant hormonal cytoreduction and radiotherapy in locally advanced carcinoma of the prostate: the Radiation Therapy Oncology Group Protocol 92-02. J Clin Oncol. 2003; 21: 39723978.
  • 31
    Nguyen PL, Chen MH, Goldhaber SZ, et al. Coronary revascularization and mortality in men with congestive heart failure or prior myocardial infarction who receive androgen deprivation. Cancer. 2011; 117: 406413.