NPC Archive Item: Observational study suggests candesartan may be preferable to losartan in heart failure

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review NPC Logo

20th January 2011

A Swedish observational study has suggested that patients with heart failure have improved survival when they are treated with candesartan compared with losartan. This study has limitations but highlights the possibility that there may be some differences between individual angiotensin 2 receptor antagonists when they are used in people with heart failure.

Level of evidence:
Level 2 (limited quality patient-oriented evidence) according to the SORT criteria.

NICE guidance
recommends that an ACE inhibitor is the first choice renin-angiotensin system (RAS) drug in heart failure. An angiotensin 2 receptor antagonist (A2RA, also known as an ARB or sartan) licensed for heart failure can be considered if the patient develops intolerable side effects with an ACE inhibitor; or used in combination with an ACE inhibitor and beta blocker in certain patients on specialist advice. This study provides no information about the comparative effects of losartan or candesartan in hypertension or other indications, but A2RAs are not recommended by NICE as first choice RAS drugs for any indication. Prescribing managers should review local prescribing trends for RAS drugs as suggested in the document Key therapeutic topics 2010/11 – Medicines management options for local implementation, produced by the NPC as part of the QIPP programme. Despite this study’s limitations, pending further data it would seem appropriate to exercise caution when considering, after a careful medication review, whether to change from candesartan to losartan in patients with heart failure.

What is the background to this?
ACE inhibitors are the first choice RAS drugs in heart failure, according to NICE guidance. A meta-analysis found that A2RAs as a class reduce hospitalisations for heart failure compared to placebo, although the observed benefit on all-cause mortality was of borderline statistical significance.  However, individual A2RAs have not been compared head to head in randomised controlled trials (RCTs). The authors of this study conducted a retrospective cohort study to compare the effects of losartan and candesartan on all-cause mortality in a large cohort of unselected people with heart failure. This was taken from the Swedish Heart Failure Registry, which included 5,139 people who had taken either candesartan or losartan, and who entered the register between 2000 and 2009.

What does this study claim?
The authors conclude “Our findings suggest that candesartan is associated with less all-cause mortality than losartan”. The adjusted hazard ratio (HR) for all-cause mortality with losartan compared to candesartan was 1.43 (95% confidence interval [CI] 1.23 to 1.65; P <0.001).

The 5-year survival was 61% in the group of patients receiving candesartan and 44% in patients receiving losartan.  However, the limitations of this observational study (see below) should be borne in mind when considering these figures.

So what?
Although the study raises concerns about the effectiveness of losartan in heart failure compared to candesartan, it has some limitations. First of all, there was no control arm, perhaps of treatment with an ACE-inhibitor of proven efficacy, against which to measure both losartan and candesartan. A second limitation was the criteria for entry into the cohort – this was on the basis of clinician judgement, rather than strict diagnostic criteria as would be the case in the ideal RCT. Nevertheless, 70 variables were recorded for each patient at discharge or after clinic visit and these data were incorporated into the adjustments for confounding (see below). In contrast, it is often stated that a strength of observational studies is their pragmatism, which may be more representative of normal clinical practice.

However, a single observational study can prove only association not causation and observational studies are prone to confounding. Unlike in the setting of an RCT, in ‘real life’, treatment plans are chosen, changed, or actively not chosen in the light of individual patients’ risk factors, preferences and tolerability or response to other drugs. Thus observed differences in outcomes may be due to differences among the patients, not only the different treatments. In this study there were a number of imbalances between the two groups at baseline and it is likely that these imbalances overall would have favoured the candesartan group. Patients receiving candesartan were overall healthier – for example 61% of candesartan patients had NYHA class I or II heart failure compared to 49% of losartan patients, 51% of candesartan patients had ischaemic heart disease compared to 61% of losartan patients, and more candesartan patients were receiving specialist cardiologist care (55% vs 51%).  Some significant imbalances did favour losaratan – for example, candesartan patients were less likely to reach target doses (defined as 32mg/day for candesartan and 50mg/day for losartan).

The authors of the study used propensity score matching to reduce the likelihood of bias arising from the choice of treatment for each patient. In brief, this considers and adjusts for the likelihood of a patient receiving one treatment rather than the other based on a number of pre-treatment factors. They also attempted to adjust for confounders which may have affected the outcomes for patients, such as whether the patient was receiving a beta-blocker, and the dose of ARB prescribed. This included adjusting for an (unlicensed) dose of 150mg/day losartan, which the HEAAL study suggested may have some benefits over the 50mg/day dose in patients with heart failure. Despite these attempts to minimise biases it is possible that the model used did not fully account for the imbalances, and also that other confounders existed which were not known and therefore not adjusted for.

NICE guidance on management of heart failure recommends that ACE inhibitors and beta blockers licensed for heart failure are first line drug treatment for all patients with heart failure due to left ventricular systolic dysfunction. An A2RA licensed for heart failure could be considered as an alternative in patients who develop intolerable side effects to an ACE inhibitor, such as cough. However, health professionals should not assume that a cough is inevitably caused by an ACE-inhibitor: other possible causes include respiratory tract infections and worsening of heart failure. If patients remain symptomatic despite optimal therapy with a beta-blocker and an ACE inhibitor, adding an A2RA is one of several options, but should be done only on specialist advice. Candesartan and losartan are both licensed for use in heart failure. Generic formulations of losartan are now available and, at the doses recommended for heart failure, losartan is less than one tenth the cost of candesartan (£16.13 per month vs £1.54 per month). This has led some to suggest therapeutic substitution of candesartan (and other A2RAs) with losartan to reduce NHS prescribing costs. However, this study suggests caution should be exercised before assuming a class effect for A2RAs in heart failure, and therefore caution is needed before changing, after a careful medication review, from candesartan to losartan in patients with heart failure.

Study details
Eklind-Cervenka M et al. Association of candesartan vs losartan with all-cause mortality in patients with heart failure JAMA 2011;305:175-82

Design Cohort study

Patients 5,139 patients in the Swedish Heart Failure Registry registered on the basis of clinician-judged HF, registered between 2000 and 2009.

Intervention and comparison Candesartan (n=2,639) or losartan (n=2,500)

Outcomes and results One-year survival was 90% (95%CI 89% to 91%) for patients receiving candesartan and 83% (95%CI 81% to 84%) for patients receiving losartan, and 5-year survival was 61% (95%CI 54% to 68%) and 44% (95%CI 41% to 48%), respectively (log-rank P< 0.001). In multivariate analysis with adjustment for propensity scores, the hazard ratio for all-cause mortality for losartan compared with candesartan was 1.43 (95%CI 1.23 to 1.65; P<0.001).

Sponsorship This study received no specific funding. However, establishment of the Swedish Heart Failure Registry was supported by the Swedish National Board of Health and Welfare, the Swedish Association of Local Authorities and Regions, the Swedish Society of Cardiology, AstraZeneca, Roche, Merck, Orion Pharma, and ResMed.

More information can be found on the heart failure section of NPC

Please comment on this blog in the NPC discussion rooms, or using our feedback form.

Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you