Re: Angiotensin converting enzyme inhibitors and risk of lung cancer: population based cohort study
In the prospective registry by Hicks et al. of new prescriptions of ACEi and ARB for hypertension, the “absolute” risk of lung cancer over 6.4 years is relatively small at 0.16% vs. 0.12% respectively, with ACEi having a 0.04% absolute greater risk of lung cancer than ARB, which translate into a 14% increase in relative risk (HR 1.14; 95% CI 1.01-1.29). Both Hicks and the accompanying editorial outline the limitations of this registry which makes clear that there can be no conclusion for causality of lung cancer by ACEi – a hypothesis that would need to be tested in a prospective randomized controlled trial. Any potential increase risk of lung cancer with ACEi as compared to an ARB – albeit small – reminds us that drug safety is paramount, and therefore must also be assessed in the context of the most inclusive and robust endpoint for safety, as well as drug efficacy - all-cause mortality. All-cause mortality was not reported by Hicks although patients were “censored” at the time of death. In the randomized ACEi/ARB trials, all-cause mortality is 50 times more common than the incidence of lung cancer reported in the registry (all-cause mortality rates in placebo arms of parallel meta-analysis of ACEi/ARB trials - 7.8 and 9%, respectively1). This is relevant as there is an ongoing controversy – the “ARB MI Paradox”2 - that proposes that ARB, despite good tolerability and effective blood pressure lowering, do not reduce the risk of MI or death as do ACEi3. If this hypothesis is valid4, then the Hick registry must be viewed in the context of their yet to be reported all-cause mortality.
The clinical trial evidence that supports the “ARB MI Paradox” is extensive and cogent. In a parallel meta-analysis of ACEi and ARB trials in high-risk patients2, ACEi vs. all comparators (39 trials, n=150,943) reduce the relative risk of all-cause mortality 9% (p<0.0001) and MI 14% (p<0.0001)2. In contrast, the parallel meta-analysis of ARB vs. all comparators (11 trials, n=55,050) demonstrates no reduction in mortality (OR 1.01; 95% CI 0.96-1.06, p=0.8) and suggested that the risk of MI actually increased 8% (OR 1.08; 95% CI 1.01-1.16; p=0.03). When systolic blood pressure reductions within ACEi and ARB trials were regressed against the risk of MI and death (CHD –coronary heart disease) (N=137,356) by the Blood Pressure Lowering Treatment Trialists Collaboration, ACEi reduce the risk of CHD an additional 9% above and “independent” of the effects of blood pressure lowering, with that risk reduction apparent even in the absence of any blood pressure reductions (p=0.002) – ARB did not. Furthermore, for any given blood pressure reduction, ACEi reduce the risk of MI and death an additional 15% (p=0.001) above that of an ARB.
Drug safety as well as drug efficacy, is best evaluated when the comparator is a placebo – the gold standard. There are 4 large parallel meta-analyses of ACEi and ARB vs. placebo randomized controlled trials that include up to 130,000 patients1! ACEi demonstrate robust risk reductions of 9 to 11% in all-cause mortality, 13-17% in CV mortality, and 17-21% in MI (all p<0.05; Thomopoulos5 – hypertension; Savarese6 - high risk; Messerli1 - high risk; Cheng7 - diabetes mellitus)4. In the parallel meta-analyses of ARB vs. placebo-controlled trials, ARB had a complete lack of risk reduction in both all-cause and CV mortality (HR 1.01-1.03, HR 1.02-1.21; p=ns, respectively). ARB absolute lack of cardiovascular protection was apparent despite a robust average blood pressure reduction of 3.7/2.0 mmHg, compared to placebo in the hypertension trials5 and similar event rates in the placebo arms of the parallel ACEi and ARB trials1.
Although head to head trials of ACEi vs. ARB have the potential to provide further insight into the “ARB MI Paradox”, there is a paucity of data. In the ONTARGET trial (n=17,118, follow up 56 months)8 – the largest head to head trial – telmisartan was not statistically equivalent to ramipril for the combined primary cardiac endpoint, rather was statistically “non-inferior”, with ONTARGET concluding that “telmisartan retained at least half the effect of ramipril”. The Food and Drug Administration and the Health Protection Bureau (Canada) provided telmisartan a second line indication – not a first line - in high risk patients who are ACEi intolerant. A meta-analysis of head-to-head trials (8 trials, n=22,543)1, adds little to the discussion as 99% of those deaths are from ONTARGET.
In conclusion, the Hicks’ paper has failed to adequately consider / adjust for two important explanations for their unexpected observations (1) survivorship bias (patients on ACEi living longer and hence more patient years over which a diagnosis of lung cancer might be made) (2) ascertainment bias (patients with ACEi more likely to undergo cancer tests as a result of cough)9. As discussed above, it is vital to note that multiple analyses of randomized controlled trials – both individually and together – have demonstrated significant survival benefits with ACEi that are entirely absent for patients treated with ARB4. Such data benefits from having the robust methodology of randomization to ensure removal of bias relating to case-mix and also placebo control with parallel design to ensure only true drug effects are observed.
We look forward to Hicks sharing their data on all-cause mortality.
1. Bangalore S, Fakheri R, Toklu B, Ogedegbe G, Weintraub H, et al. Angiotensin-Converting Enzyme Inhibitors or Angiotensin Receptor Blockers in Patients Without Heart Failure? Insights From 254,301 Patients From Randomized Trials. Mayo Clin Proc. 2016;91:51-60.
2. Strauss MH and Hall AS. Angiotensin receptor blockers may increase risk of myocardial infarction: unraveling the ARB-MI paradox. Circulation. 2006;114:838-854.
3. Verma S and Strauss M. Angiotensin receptor blockers and myocardial infarction. BMJ. 2004;329:1248-1249.
4. Strauss MH and Hall AS. The Divergent Cardiovascular Effects of Angiotensin-Converting Enzyme Inhibitors and Angiotensin II Type 1 Receptor Blockers in Adult Patients With Type 2 Diabetes Mellitus. Can J Diabetes. 2018;42:124-129.
5. Thomopoulos C, Parati G and Zanchetti A. Effects of blood pressure lowering on outcome incidence in hypertension: 4. Effects of various classes of antihypertensive drugs--overview and meta-analyses. J Hypertens. 2015;33:195-211.
6. Savarese G, Costanzo P, Cleland JG, Vassallo E, Ruggiero D, et al. A meta-analysis reporting effects of Angiotensin-converting enzyme inhibitors and Angiotensin receptor blockers in patients without heart failure. J Am Coll Cardiol. 2013;61:131-142.
7. Cheng J, Zhang W, Zhang X, Han F, Li X, et al. Effect of Angiotensin-Converting Enzyme Inhibitors and Angiotensin II Receptor Blockers on All-Cause Mortality, Cardiovascular Deaths, and Cardiovascular Events in Patients With Diabetes Mellitus: A Meta-analysis. JAMA Intern Med. 2014.
8. Yusuf S, Teo KK, Pogue J, Dyal L, Copland I, et al. Telmisartan, ramipril, or both in patients at high risk for vascular events. N Engl J Med. 2008;358:1547-1559.
9. Delgado-Rodriguez M and Llorca J. Bias. J Epidemiol Community Health. 2004;58:635-41.
Competing interests:
both authors received speaking honorarium from Servier Pharma
14 November 2018
Martin H Strauss
cardiologist
Alistair S Hall, Leeds MRC Medical Bioinformatics Centre, University of Leeds, West Yorkshire, United Kingdom
Department of Medicine, University of Toronto, North York General Hospital, Toronto
Rapid Response:
Re: Angiotensin converting enzyme inhibitors and risk of lung cancer: population based cohort study
In the prospective registry by Hicks et al. of new prescriptions of ACEi and ARB for hypertension, the “absolute” risk of lung cancer over 6.4 years is relatively small at 0.16% vs. 0.12% respectively, with ACEi having a 0.04% absolute greater risk of lung cancer than ARB, which translate into a 14% increase in relative risk (HR 1.14; 95% CI 1.01-1.29). Both Hicks and the accompanying editorial outline the limitations of this registry which makes clear that there can be no conclusion for causality of lung cancer by ACEi – a hypothesis that would need to be tested in a prospective randomized controlled trial. Any potential increase risk of lung cancer with ACEi as compared to an ARB – albeit small – reminds us that drug safety is paramount, and therefore must also be assessed in the context of the most inclusive and robust endpoint for safety, as well as drug efficacy - all-cause mortality. All-cause mortality was not reported by Hicks although patients were “censored” at the time of death. In the randomized ACEi/ARB trials, all-cause mortality is 50 times more common than the incidence of lung cancer reported in the registry (all-cause mortality rates in placebo arms of parallel meta-analysis of ACEi/ARB trials - 7.8 and 9%, respectively1). This is relevant as there is an ongoing controversy – the “ARB MI Paradox”2 - that proposes that ARB, despite good tolerability and effective blood pressure lowering, do not reduce the risk of MI or death as do ACEi3. If this hypothesis is valid4, then the Hick registry must be viewed in the context of their yet to be reported all-cause mortality.
The clinical trial evidence that supports the “ARB MI Paradox” is extensive and cogent. In a parallel meta-analysis of ACEi and ARB trials in high-risk patients2, ACEi vs. all comparators (39 trials, n=150,943) reduce the relative risk of all-cause mortality 9% (p<0.0001) and MI 14% (p<0.0001)2. In contrast, the parallel meta-analysis of ARB vs. all comparators (11 trials, n=55,050) demonstrates no reduction in mortality (OR 1.01; 95% CI 0.96-1.06, p=0.8) and suggested that the risk of MI actually increased 8% (OR 1.08; 95% CI 1.01-1.16; p=0.03). When systolic blood pressure reductions within ACEi and ARB trials were regressed against the risk of MI and death (CHD –coronary heart disease) (N=137,356) by the Blood Pressure Lowering Treatment Trialists Collaboration, ACEi reduce the risk of CHD an additional 9% above and “independent” of the effects of blood pressure lowering, with that risk reduction apparent even in the absence of any blood pressure reductions (p=0.002) – ARB did not. Furthermore, for any given blood pressure reduction, ACEi reduce the risk of MI and death an additional 15% (p=0.001) above that of an ARB.
Drug safety as well as drug efficacy, is best evaluated when the comparator is a placebo – the gold standard. There are 4 large parallel meta-analyses of ACEi and ARB vs. placebo randomized controlled trials that include up to 130,000 patients1! ACEi demonstrate robust risk reductions of 9 to 11% in all-cause mortality, 13-17% in CV mortality, and 17-21% in MI (all p<0.05; Thomopoulos5 – hypertension; Savarese6 - high risk; Messerli1 - high risk; Cheng7 - diabetes mellitus)4. In the parallel meta-analyses of ARB vs. placebo-controlled trials, ARB had a complete lack of risk reduction in both all-cause and CV mortality (HR 1.01-1.03, HR 1.02-1.21; p=ns, respectively). ARB absolute lack of cardiovascular protection was apparent despite a robust average blood pressure reduction of 3.7/2.0 mmHg, compared to placebo in the hypertension trials5 and similar event rates in the placebo arms of the parallel ACEi and ARB trials1.
Although head to head trials of ACEi vs. ARB have the potential to provide further insight into the “ARB MI Paradox”, there is a paucity of data. In the ONTARGET trial (n=17,118, follow up 56 months)8 – the largest head to head trial – telmisartan was not statistically equivalent to ramipril for the combined primary cardiac endpoint, rather was statistically “non-inferior”, with ONTARGET concluding that “telmisartan retained at least half the effect of ramipril”. The Food and Drug Administration and the Health Protection Bureau (Canada) provided telmisartan a second line indication – not a first line - in high risk patients who are ACEi intolerant. A meta-analysis of head-to-head trials (8 trials, n=22,543)1, adds little to the discussion as 99% of those deaths are from ONTARGET.
In conclusion, the Hicks’ paper has failed to adequately consider / adjust for two important explanations for their unexpected observations (1) survivorship bias (patients on ACEi living longer and hence more patient years over which a diagnosis of lung cancer might be made) (2) ascertainment bias (patients with ACEi more likely to undergo cancer tests as a result of cough)9. As discussed above, it is vital to note that multiple analyses of randomized controlled trials – both individually and together – have demonstrated significant survival benefits with ACEi that are entirely absent for patients treated with ARB4. Such data benefits from having the robust methodology of randomization to ensure removal of bias relating to case-mix and also placebo control with parallel design to ensure only true drug effects are observed.
We look forward to Hicks sharing their data on all-cause mortality.
1. Bangalore S, Fakheri R, Toklu B, Ogedegbe G, Weintraub H, et al. Angiotensin-Converting Enzyme Inhibitors or Angiotensin Receptor Blockers in Patients Without Heart Failure? Insights From 254,301 Patients From Randomized Trials. Mayo Clin Proc. 2016;91:51-60.
2. Strauss MH and Hall AS. Angiotensin receptor blockers may increase risk of myocardial infarction: unraveling the ARB-MI paradox. Circulation. 2006;114:838-854.
3. Verma S and Strauss M. Angiotensin receptor blockers and myocardial infarction. BMJ. 2004;329:1248-1249.
4. Strauss MH and Hall AS. The Divergent Cardiovascular Effects of Angiotensin-Converting Enzyme Inhibitors and Angiotensin II Type 1 Receptor Blockers in Adult Patients With Type 2 Diabetes Mellitus. Can J Diabetes. 2018;42:124-129.
5. Thomopoulos C, Parati G and Zanchetti A. Effects of blood pressure lowering on outcome incidence in hypertension: 4. Effects of various classes of antihypertensive drugs--overview and meta-analyses. J Hypertens. 2015;33:195-211.
6. Savarese G, Costanzo P, Cleland JG, Vassallo E, Ruggiero D, et al. A meta-analysis reporting effects of Angiotensin-converting enzyme inhibitors and Angiotensin receptor blockers in patients without heart failure. J Am Coll Cardiol. 2013;61:131-142.
7. Cheng J, Zhang W, Zhang X, Han F, Li X, et al. Effect of Angiotensin-Converting Enzyme Inhibitors and Angiotensin II Receptor Blockers on All-Cause Mortality, Cardiovascular Deaths, and Cardiovascular Events in Patients With Diabetes Mellitus: A Meta-analysis. JAMA Intern Med. 2014.
8. Yusuf S, Teo KK, Pogue J, Dyal L, Copland I, et al. Telmisartan, ramipril, or both in patients at high risk for vascular events. N Engl J Med. 2008;358:1547-1559.
9. Delgado-Rodriguez M and Llorca J. Bias. J Epidemiol Community Health. 2004;58:635-41.
Competing interests: both authors received speaking honorarium from Servier Pharma