|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 25, No 17 (June 10), 2007: pp. 2420-2425 © 2007 American Society of Clinical Oncology. DOI: 10.1200/JCO.2006.09.3369 Influence of Androgen Suppression Therapy for Prostate Cancer on the Frequency and Timing of Fatal Myocardial Infarctions
From the Departments of Radiation Oncology and Medicine, Cardiovascular Division, Brigham and Women's Hospital and Lank Center for Genitourinary Oncology, Dana-Farber Cancer Institute, Boston, MA; Departments of Radiation Oncology and Statistics, University of Connecticut, Storrs, CT; Trans-Tasman Radiation Oncology Group, Newcastle Mater Hospital, Newcastle, New South Wales; Department of Radiation Oncology, Peter MacCallum Cancer Institute, Melbourne, Victoria, Australia; Princess Margaret Hospital, Toronto, Ontario; Department of Radiation Oncology, Ottawa Regional Cancer Center, Ottawa, Ontario; Department of Radiation Oncology, Vancouver Island Cancer Center, Victoria, British Columbia, Canada; and Department of Radiation Oncology, Wellington Hospital, Wellington, New Zealand Address reprint requests to Anthony V. D'Amico, MD, PhD, Department of Radiation Oncology, Brigham and Women's Hospital, 75 Francis St, L-2 Level, Boston, MA 02215; e-mail: adamico{at}lroc.harvard.edu
Purpose We evaluated whether the timing of fatal myocardial infarction (MI) was influenced by the administration of androgen suppression therapy (AST). Patients and Methods The study cohort comprised 1,372 men who were enrolled onto three randomized trials between February 1995 and June 2001. In the three trials, the men were randomly assigned to receive radiation therapy with 0 versus 3 versus 6, 3 versus 8, or 0 versus 6 months of AST. Fine and Gray's regression was used to determine the clinical factors associated with the time to fatal MI, and estimates of time to fatal MI were calculated using a cumulative incidence method. When comparing the cumulative incidence estimates using Gray's k-sample P values, increased weight was ascribed to the earlier data because recovery of testosterone is expected for most men within 2 years after short-course AST. Results Men age 65 years or older who received 6 months of AST experienced shorter times to fatal MIs compared with men in this age group who did not receive AST (P = .017) and men younger than 65 years (P = .016). No significant difference (P = .97) was observed in the time to fatal MIs in men age 65 years or older who received 6 to 8 months of AST compared with 3 months of AST. Conclusion The use of AST is associated with earlier onset of fatal MIs in men age 65 years or older who are treated for 6 months compared with men who are not treated with AST.
Estrogen therapy in the treatment of prostate cancer was abandoned because of an increased risk of thromboembolic events leading to a shortening of overall survival.1 This form of hormonal therapy was replaced by classes of androgen suppression therapy (AST)2 that competitively inhibit the action of testosterone at the level of the androgen receptor or markedly reduce the secretion of luteinizing hormone that stimulates the production of testosterone by the Leydig cell. When administered in conjunction with external-beam radiation therapy (RT), 6 months to 3 years of AST has been shown to significantly decrease prostate cancer–specific mortality,3,4 all cause mortality, or both5,6 in men with high-grade localized or locally advanced prostate cancer compared with RT alone, making combined RT and AST a current standard of care for these men.3-6 On the basis of this evidence,3-6 the use of AST in conjunction with RT has increased markedly.7 Moreover, an increasing number of men in the United States (US) are receiving AST alone or as part of their treatment for newly diagnosed or recurrent prostate cancer.8 However, AST is associated with anemia, weight gain, insulin resistance, increased arterial stiffness, less favorable lipid profiles, and increased diagnoses of cardiac disease.9-13 These changes can persist after discontinuing AST, especially in men of advanced age.14 This complex of symptoms is consistent with the metabolic syndrome15 that is associated with an increased risk of death as a result of myocardial infarction (MI), even in the absence of known cardiovascular disease or diabetes.16 In addition, an unexpected and unexplained increase in mortality was observed in men with localized prostate cancer undergoing watchful waiting and randomly assigned to receive 150 mg of the antiandrogen bicalutamide compared with men who were assigned to placebo.17 This increase in mortality led to the withdrawal of this antiandrogen at the 150-mg dose level from clinical use in Canada.18 It is possible that hemodynamic and metabolic changes induced by AST may precipitate an MI in a predisposed man. However, this conjecture has not been studied in the setting of randomized trials where risk factors for coronary artery disease (CAD) are balanced between treatment arms. Therefore, in this study, we evaluated whether the timing of a fatal MI was influenced by the administration of AST to men receiving RT to treat prostate cancer in three prospective randomized trials.3,5,14
Patient Selection, Random Assignment, and Treatment To address the question of whether the administration of AST influences the timing of a fatal MI, we performed a pooled data analysis of randomized studies that treated men with RT with or without short-course AST or RT and differing durations of short-course AST. Medline was searched from January 1990 to December 2005 to identify studies and principal investigators contacted and invited to collaborate. The study cohort comprised 1,372 men enrolled onto three randomized trials conducted in Australia and New Zealand by the Trans-Tasman Radiation Oncology Group (TROG 96013; n = 802), Canada14 (n = 364), and the United States5 (n = 206) between February 1995 and June 2001. The men in the TROG, Canadian, and US trials were randomly assigned to and received RT and 0 versus 3 versus 6, 3 versus 8, or 0 versus 6 months of AST, respectively. All patients read and signed an approved internal review board consent form before study entry and had a life expectancy of at least 10 years in the US study5 or 5 years in the TROG and Canadian studies.3,14 The median ages of the men in the TROG,3 US,5 and Canadian14 studies were 68 years (interquartile range [IQR], 63 to 72 years), 72.5 years (IQR, 69 to 75 years), and 72 years (IQR, 68 to 75 years), respectively. The patient characteristics before random assignment from the TROG,3 Canadian,14 and US5 studies are listed in Table 1.
The details of the RT techniques, doses, and fields have been previously described.3,5,14 AST consisted of a combination of a luteinizing hormone–releasing hormone (LHRH) agonist (leuprolide acetate) or goserelin in the US study5 and goserelin in the TROG3 and Canadian14 studies coupled with a nonsteroidal antiandrogen (flutamide). Leuprolide acetate was delivered intramuscularly each month at a dose of 7.5 mg or every 3 months at a dose of 22.5 mg. Goserelin was administered subcutaneously each month at a dose of 3.6 mg in the TROG3 and Canadian studies14; goserelin was administered at a dose of 3.6 mg every month or 10.8 mg every 3 months in the US study.5 Both LHRH agonists were permitted in the US study5 because they have been shown to have equivalent efficacy in the treatment of prostate cancer.2 Flutamide was taken orally at a dose of 250 mg every 8 hours and starting 1 to 3 days before the LHRH agonist to block the transient increase in testosterone caused by the LHRH agonist.
Follow-Up Protocol
Determination of the Cause of Death
Statistical Methods
Competing risks regression evaluating associations with time to fatal MI.
A proportional hazards model that adjusted for competing causes of mortality, Fine and Gray's regression,22 was used to determine the clinical factors associated with the time to a fatal MI after random assignment for the pooled TROG/US and TROG/Canadian databases. The databases were grouped so that trials with overlapping randomly assigned treatment arms were compared to take advantage of the ability of random assignment within the individual studies to balance known and unknown confounding factors across treatment arms. To adjust for the potential of trial differences, the categoric covariate of trial was included in addition to duration of AST in addition to age (years) at random assignment, which was evaluated in two ways in two distinct models, continuous and categoric ( Cumulative incidence estimates of fatal MIs. A cumulative incidence methodology23 that adjusted for competing risks of mortality was used for the purpose of illustration. Specifically, cumulative incidence estimates23 of fatal MIs were computed stratified by age at random assignment for the individual treatment arms of the combined data sets containing 0 versus 6 months and 3 versus 6 to 8 months of AST duration. Age 65 years was selected as the cut point based on a prior randomized study24 of radical prostatectomy versus watchful waiting that suggested that treatment may not be beneficial in men age 65 years or older. Given that AST has toxicity,9-13 evaluating the cumulative incidence of fatal MIs in men at or older than this age at which treatment has been questioned is clinically relevant.
Because recovery of testosterone to the prerandomization baseline level is expected for most men within 2 years after 3 to 6 months of AST,25,26 when comparing the cumulative incidence estimates using Gray's k-sample P values,27 increased weight was ascribed to the earlier data. A two-sided P
CAD Risk Factors by Randomized Treatment Arm As presented in Table 2, the CAD risk factors at random assignment were balanced between the randomized treatment arms for the 206 men in the US study, providing evidence that the random assignment worked properly. Although these data were not routinely collected at random assignment in the TROG and Canadian studies, the sample sizes of those studies compared with the US study were approximately four-fold and 1.8-fold larger, respectively. Therefore, one would expect that both known and unknown risk factors for CAD would also be balanced across the randomized treatment arms in the larger randomized studies.
Time to Fatal MI Of the 802, 364, and 206 men randomly assigned on the TROG, Canadian, and US studies, respectively, there were 29 (3.6%), six (1.6%), and 16 (7.8%) fatal MIs that occurred during follow-up. As presented in Table 3, for the combined TROG/US and TROG/Canadian studies, age at random assignment was significantly associated with time to fatal MI when analyzed as a categoric covariate (P = .01 for TROG/US and P = .02 for TROG/Canada) or in a separate model when analyzed as a continuous variable (P = .02 for TROG/US and P = .003 for TROG/Canada). However, trial, duration of AST, and the interaction terms between trial and treatment and age and treatment were not significant in any of the models. Given a lack of interaction between trial and treatment, pooling of the databases by similar or identical treatment arms for the purpose of evaluating the cumulative incidence of time to fatal MIs was reasonable. Although duration of AST was not significantly associated with time to fatal MI, Fine and Gray's regression did not permit analysis ascribing increased weighting of the early data during which time the androgen suppression was in effect.
Cumulative Incidence Estimates of Fatal MI Stratified by AST Duration and Age The median follow-up times of the men in the TROG, US, and Canadian studies were 5.9 years, (IQR, 5.1 to 6.6 years), 6.7 years (IQR, 5.4 to 8.4 years), and 4.8 years (IQR, 3.4 to 6.0 years), respectively. As shown in Figure 1 for the combined TROG/US data, men age 65 years or older who received 6 months of AST experienced a shorter time to fatal MI compared with men in the same age group who did not receive AST (P = .017) and all men age less than 65 years (P = .016). For the combined TROG/Canadian data, no significant difference (P = .97) was observed in the time to fatal MIs in men age 65 years or older who received 6 to 8 months compared with 3 months of AST (Fig 2). The median follow-up time for men in the Canadian study compared with men in the US study was significantly shorter (4.8 v 6.7 years, respectively; P < .001), which explains why the cumulative incidence estimates of fatal MI are numerically smaller in Figure 2 for men age more than 65 years who received 6 to 8 months of AST compared with men age more than 65 years who received 6 months of AST in Figure 1. Men in the combined TROG/Canadian cohort who were less than 65 years old at random assignment experienced longer times to fatal MI compared with men age 65 years or older who received 3 (P = .016) or 6 to 8 months (P = .015) of AST.
An increasing number of men in the United States, especially men older than 80 years, are receiving AST alone or as part of their management for newly diagnosed or recurrent prostate cancer.8 Yet, concerns have been raised about the well-being of older men on AST because the adverse effects of AST9-13 can lead to a symptom complex consistent with the metabolic syndrome.15 This syndrome is associated with an increased risk of death as a result of MI, even in the absence of known cardiovascular disease or diabetes.16 Clinical evidence from two prior studies17,29 lend support to this association. The Early Prostate Cancer Trial17 found an unexplained decrease in overall survival in men with a median age of 69 years who were on a watchful waiting program for clinically localized prostate cancer and who were randomly assigned to receive 150 mg of the antiandrogen bicalutamide by mouth daily for a median of 4.4 years compared with placebo. Although bicalutamide can increase serum testosterone levels, evidence now exists to suggest that the protective effect that testosterone may have on the development of atherosclerosis is blocked by bicalutamide's antagonism of the androgen receptor.30 In addition, a significant association was reported30 between neoadjuvant AST use and a shorter time to death in a study or 2,378 men with a median age of 73 years treated with brachytherapy. AST of varying durations (3 to approximately 12 months) was administered to 464 of these men to reduce the prostate gland volume before brachytherapy. The association between AST use and a shorter time to death persisted after adjusting for age, Gleason score, and the prostate-specific antigen level. In the current study of pooled data from three randomized trials from Australia and New Zealand (by TROG),3 Canada,14 and the United States,5 men whose age was 65 years or more and who were randomly assigned to receive 6 months compared with no AST experienced shorter times to fatal MI. The clinical significance of the shortening of the time to fatal MI is illustrated in Figure 1 for men in the combined TROG/US studies. Specifically, before the first fatal MI occurred at 21 months in men randomly assigned to receive no AST and aged 65 years or older, 44% (eight of 18 MIs) of all observed fatal MIs had occurred in men in this age group randomly assigned to receive 6 months of AST. Therefore, of men age 65 years or older who were at risk for having a fatal MI, nearly half of them were observed to experience the MI sooner if they had received a 6-month course of AST. Men age 65 years or older randomly assigned to receive 6 to 8 versus 3 months of AST were observed to have similar times to fatal MI. However, other than the TROG data, no other randomized data of men receiving 3 months versus no AST in a single trial were available, and as a result, the power to measure whether 3 months of AST significantly shortens the time to fatal MIs compared with no AST was limited. Therefore, compared with no AST, 6 months of AST can shorten the time to a fatal MI in men age 65 years or older. The shortening of the time to a fatal MI may occur in men age 65 years or older with as little as 3 months of AST, but additional evidence to support this conjecture is needed. Two points deserve further consideration. First, although the trials included in this pooled data analysis came from three different countries, which potentially enhances the generalizability of the results, the possibility of a selection bias exists given that there were only three studies. Second, insufficient events were observed in men age 65 years or less to do a comparison of the effect of AST duration on time to fatal MI. Therefore, future studies examining the association of AST duration and time to fatal MI are needed to validate the findings of the current and rule out the potential for selection bias, and longer follow-up of the current study is needed in men age less than 65 years to assess whether the increased early fatal MI rate observed for men older than 65 years also exists in younger men. It now remains to be determined what mechanisms are responsible for this effect and what risk factors are operative. We have already speculated that induction or aggravation of the metabolic syndrome15,16 by AST may be responsible. If this were the case, then a shortening in the time to nonfatal MI would also be expected to occur. Although data on nonfatal MIs were not routinely collected in the TROG,3 US,5 and Canadian14 randomized studies, they remain a subject for future studies to explore. However, until prospective studies are conducted, it will remain difficult to know how to prevent this phenomenon. Until then, it is reasonable to recommend that men of advanced age with prostate cancer planning to undergo RT and AST should be considered for cardiovascular evaluation31 before the initiation of AST. If clinically significant CAD is identified, then risk factor modification, appropriate pharmacologic interventions, and mechanical interventions should be considered before initiating AST. In conclusion, 6 months, and possibly as little as 3 months, of AST causes fatal MIs to occur sooner in men age 65 years or older. Prospective studies are necessary to identify the mechanisms and risk factors that are operative so that preventative strategies can be developed. While awaiting the results of such studies, the current study provides evidence to support a cardiovascular evaluation and, if indicated, intervention before the initiation of AST in men age 65 years or older to prevent premature fatal MIs.
The author(s) indicated no potential conflicts of interest.
Conception and design: Anthony V. D'Amico Administrative support: Anthony V. D'Amico, Ming-Hui Chen, David Joseph, Allison Steigler Provision of study materials or patients: Anthony V. D'Amico, James W. Denham, Juanita Crook, Samuel Z. Goldhaber, Keen-Hun Tai, Shawn Malone, Charles Ludgate, Philip W. Kantoff Collection and assembly of data: Anthony V. D'Amico, James W. Denham, Juanita Crook, Ming-Hui Chen, David S. Lamb, Allison Steigler Data analysis and interpretation: Anthony V. D'Amico, James W. Denham, Juanita Crook, Ming-Hui Chen, David Joseph, Philip W. Kantoff Manuscript writing: Anthony V. D'Amico, James W. Denham, Juanita Crook, Ming-Hui Chen, Samuel Z. Goldhaber, David S. Lamb, Keen-Hun Tai, Shawn Malone, Charles Ludgate, Philip W. Kantoff Final approval of manuscript: Anthony V. D'Amico, James W. Denham, Juanita Crook, Ming-Hui Chen, Samuel Z. Goldhaber, David S. Lamb, Keen-Hun Tai, Shawn Malone, Charles Ludgate, Allison Steigler, Philip W. Kantoff
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Byar DP, Corle DK: Hormone therapy for prostate cancer: Results of the Veterans Administration Cooperative Urological Research Group studies. NCI Monogr 7:165-170, 1988[Medline] 2. Agency for Health Care and Research: Relative effectiveness and cost-effectiveness of methods of androgen suppression in the treatment of advanced prostatic cancer. http://www.ahrq.gov/clinic/epcsums/prossumm.htm 3. Denham J, Steigler A, Lamb DS, et al: Short-term androgen deprivation and radiotherapy for locally advanced prostate cancer: Results from the Tran-Tasman Radiation Oncology Group 96.01 randomised controlled trial. Lancet Oncol 6:841-850, 2005[CrossRef][Medline] 4. 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 21:3972-3978, 2003 5. D'Amico AV, Manola J, Loffredo M, et al: 6-month androgen suppression plus radiation therapy vs. radiation therapy alone for patients with clinically localized prostate cancer: A randomized controlled trial. JAMA 292:821-827, 2004 6. Bolla M, Collette L, Blank L, et al: Long-term results with immediate androgen suppression therapy and external irradiation in patients with locally advanced prostate cancer (an EORTC study): A phase III randomized trial. Lancet 360:103-108, 2002[CrossRef][Medline] 7. Park S, Meng MV, Elkin EP, et al: Androgen deprivation use with external beam radiation for prostate cancer: Results from CaPSURE. J Urol 174:1802-1807, 2005[CrossRef][Medline] 8. Shahinian VB, Kuo YF, Freeman JL, et al: Increasing use of gonadotropin-releasing hormone agonists for the treatment of localized prostate cancer. Cancer 103:1615-1624, 2005[CrossRef][Medline] 9. Holzbeierlein JM, Castle E, Thrasher JB: Complications of androgen deprivation therapy: Prevention and treatment. Oncology 18:303-309, 2004[Medline] 10. Malone S, Perry G, Segal R, et al: Long-term side-effects of intermittent androgen suppression therapy in prostate cancer: Results of a phase II study. BJU Int 96:514-520, 2005[CrossRef][Medline] 11. Smith JC, Bennett S, Evans LM, et al: The effects of induced hypogonadism on arterial stiffness, body composition, and metabolic parameters in males with prostate cancer. J Clin Endocrinol Metab 86:4261-4267, 2001 12. Hougaku H, Fleg JL, Najjar SS, et al: Relationship between androgenic hormones and arterial stiffness, based no longitudinal hormone measurements. Am J Physiol Endocrinol Metab 290:E234-E242, 2006 13. Keating NL, O'Malley AJ, Smith MR: Diabetes and cardiovascular disease during androgen deprivation therapy for prostate cancer. J Clin Oncol 24:4448-4456, 2006 14. Crook J, Ludgate C, Malone S, et al: Report of a multi-center Canadian phase III randomized trial of 3 months vs. 8 months neoadjuvant androgen deprivation before standard-dose radiotherapy for clinically localized prostate cancer. Int J Radiat Oncol Biol Phys 60:15-23, 2004[CrossRef][Medline] 15. Cleeman JI, Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults: Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 285:2486-2497, 2001 16. Lakka HM, Laaksonen DE, Lakka TA, et al: The metabolic syndrome and total and cardiovascular disease mortality in middle-aged men. JAMA 288:2709-2716, 2002 17. Wirth MP, See WA, McLeod DG, et al: Bicalutamide 150 mg in addition to standard care in patients with localized or locally advanced prostate cancer: Results from the second analysis of the early prostate cancer program at median follow up of 5.4 years. J Urol 172:1865-1870, 2004[CrossRef][Medline] 18. Health Canada: Accelerated deaths using Casodex 150 mg in patients with localized prostate cancer. http://www.hc-sc.gc.ca/dhp-mps/alt_formats/hpfb-dgpsa/pdf/medeff/casodex_hpc-cps_e.pdf 19. Grundy SM, Pasternak R, Greenland P, et al: Assessment of cardiovascular risk by use of multiple-risk-factor assessment equations: A statement for healthcare professionals from the American Heart Association and the American College of Cardiology. J Am Coll Cardiol 34:1348-1359, 1999 20. Greene FL: Prostate, in Greene FL, Page DL, Fleming ID, et al (eds): AJCC Cancer Staging Manual (ed 6). New York, NY, Springer, 2002, pp 309-316 21. Agresti A: An Introduction to Categorical Data Analysis (ed 2). New York, NY, John Wiley & Sons, 2002, pp 16-52 22. Fine JP, Gray RJ: A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 94:496-509, 1999[CrossRef] 23. Gaynor JJ, Feur EJ, Tan CC, et al: On the use of cause-specific failure and conditional failure probabilities: Examples from clinical oncology data. J Am Stat Assoc 88:400-409, 1993[CrossRef] 24. Bill-Axelson A, Holmberg L, Ruutu M, et al: Radical prostatectomy versus watchful waiting in early prostate cancer. N Engl J Med 352:1977-1984, 2005 25. Oefelein MG: Time to normalization of serum testosterone after 3-month luteinizing hormone-releasing hormone agonist administered in the neoadjuvant setting: Implications for dosing schedule and neoadjuvant study consideration. J Urol 160:1685-1688, 1996[CrossRef] 26. D'Amico AV, Renshaw AA, Loffredo B, et al: Duration of androgen suppression and the risk of death from prostate cancer in men treated using radiation and 6 months of hormonal therapy. Presented at the 3rd Annual American Society of Clinical Oncology Prostate Cancer Symposium, Orlando, FL, February, 22-24, 2007 27. Gray RJ: A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat 16:1141-1154, 1988[CrossRef] 28. Neter J, Wassermann W, Kutner M (eds): Simultaneous inferences and other topics in regression analyses-1, in Applied Linear Regression Models. Homewood, IL, McGraw-Hill, 1983, pp 150-153 29. Beyer DC, McKeough T, Thomas T: Impact of short course hormonal therapy on overall and cancer-specific survival after permanent prostate brachytherapy. Int J Radiat Oncol Biol Phys 61:1299-1305, 2005[CrossRef][Medline] 30. Norata GD, Tibolla G, Seccomandi PM, et al: Dihydrotestosterone decreases tumor necrosis factor-alpha and lipopolysaccharide-induced inflammatory response in human endothelial cells. J Clin Endocrinol Metab 91:546-554, 2006 31. Beller GA, Zaret BL: Contributions of nuclear cardiology to diagnosis and prognosis of patients with coronary artery disease. Circulation 101:1465-1478, 2000 Submitted September 25, 2006; accepted March 21, 2007.
Related Correspondence
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 metric.
65 v < 65 years old), and the interaction terms between trial and treatment and between age and treatment. The interaction term between trial and treatment was added to assess whether the same treatment from two different trials had a differential effect on the end point of time to fatal MI after random assignment. Unadjusted and adjusted hazard ratios
.05 was considered statistically significant, except in the case of multiple comparisons where a Bonferroni correction
