Cardiovascular Disease and Uric Acid

Is The Not-so-innocent Bystander Becoming a True Culprit and Does the Us Black Box Warning for Febuxostat Indicate That Not All Uric Acid Lowering Is Beneficial?

Franz H. Messerli; Michel Burnier


Eur Heart J. 2019;40(22):1787-1789. 

Ever since the observation of Cannon et al.[1] more than half a century ago that almost half of all hypertensive patients have hyperuricaemia, there has been increasing recognition of the link between uric acid, cardiovascular (CV) disease (CVD),[2] and chronic kidney disease (CKD).[3,4] Numerous observational studies have shown that the risk of CV mortality is higher in people with gout and, conversely, that CVD/CKD patients often have elevated serum uric acid levels and are at increased risk of gout (Figure 1).

Figure 1.

Renal and cardiovascular consequences of hyperuricaemia.

Disentangling association from causality is exceedingly difficult. However, defining the evidence regarding the pathogenic impact of uric acid levels is seminal because of its potential implications regarding the prevention and treatment of CVD and CKD.

Some studies have shown that lowering uric acid levels with allopurinol may improve CV conditions, such as hypertension,[5] heart failure,[6] and CKD,[7] but the level of evidence for causality remains low. The recent CARES (Cardiovascular Safety of Febuxostat and Allopurinol in Patients with Gout and Cardiovascular Morbidities) trial[8] demonstrated that treatment with febuxostat resulted in overall rates of major CV events (a composite of non-fatal myocardial infarction, non-fatal stroke, urgent revascularization for unstable angina, and death because of CV causes) comparable to those of allopurinol in patients with high CV-risk gout, although CV death and deaths from any cause were higher on febuxostat. Among a cohort of 99 744 older Medicare patients with gout, Zhang and colleagues[9] reported no difference in the risk of myocardial infarction, stroke, new-onset heart failure, coronary revascularization, or all-cause mortality between patients initiating febuxostat compared with allopurinol. With exposure exceeding 3 years, mortality for febuxostat vs. allopurinol also may have increased. However, neither the observational study by Zhang et al. nor the CARES trial had a placebo control group. Therefore, neither trial allows us to conclude that there was an increase in mortality due to febuxostat per se. As a consequence of the CARES data, a black box warning was mandated by the Food and Drug Administration (FDA) in February 2019, with the recommendation to use febuxostat only in patients who have failed on or do not tolerate allopurinol.[10]

In the present issue of the European Heart Journal, Kojima et al. report a multicentre, prospective, randomized open-label, blinded endpoint study in which a total of 1070 elderly patients with hyperuricaemia (serum uric acid >7.0 to ≤9.0 mg/dL), and a high risk of CV and renal events, were randomized to febuxostat and non-febuxostat arms (Febuxostat for Cerebral and CaRdiorenovascular Events PrEvEntion StuDy: FREED).[11] Cerebral, CV, and renal events, and all deaths, were defined as the primary composite event. Patients with a recent episode of gout were excluded. After a follow-up of 36 months, the primary composite event rate was lower in the febuxostat arm than in the non-febuxostat arm [hazard ratio (HR): 0.750; 95% confidence interval (CI): 0.592–0.950; P = 0.017). Of note, the composite endpoint was almost exclusively driven by renal events, which amounted to 16.2% in the febuxostat group and 20.5% in the non-febuxostat group (HR: 0.745; 95% CI: 0.562–0.987; P = 0.041). The renal benefits were characterized by less development of microalbuminuria or progression towards macroalbuminuria. However, there was no difference between the two groups regarding the harder endpoints such as doubling of serum creatinine or progression towards end-stage renal disease. In the absence of any cerebrovascular or cardiac benefits, the authors finally concluded that 'febuxostat lowers uric acid and delays the progression of renal dysfunction'.

Before we accept the authors' conclusion, we need to evaluate whether changes in uric acid are 'causal' or are merely a marker of increased risk. In 1965, Sir Austin Bradford Hill, an English epidemiologist, established what is now known as Hill's criteria for causation.[12,13] The below section scrutinizes the association between uric acid and outcomes with regard to Hill's criteria.

  1. Strength of association. A number of studies have shown strong associations between uric acid and the risk of adverse CV outcomes and death. In the current study, a febuxostat-induced decrease in uric acid was associated with a significant delay in the progression of renal dysfunction, thus showing a modest effect size for this outcome, but no impact on cardiac and cerebral outcomes.

  2. Consistency (repeated observation). Most studies have shown a consistent increase in the risk of outcomes with increased uric acid, an observation seen in different cohorts, by different people, across different countries. However, there could be a potential publication bias, as negative studies are unlikely to be published.

  3. Specificity. The criterion of specificity is also fulfilled based on published studies, as increased uric acid has been linked with worse CV outcomes and death. However, a reduction of uric acid has not been consistently associated with an improvement in CV outcomes, except perhaps for renal endpoints.

  4. Temporality (exposure precedes effect). Sensitivity analyses in the current study and other have shown a similar association of worse outcomes when uric acid parameters are estimated for a period of time, and also for outcomes evaluated after this period of time, suggesting that the effect does indeed occur after the exposure.

  5. Biological gradient (dose-response relationship). In several epidemiological studies, there was a graded and linear relationship between increased uric acid and worse outcomes, suggesting a dose-response relationship. In the present study, there was no clear dose-response between serum uric acid levels at 12 weeks and the cumulative primary composite event rate (see figure S6 in the paper).

  6. Plausibility. Although we and others have suggested plausible mechanisms of increased risk with increased uric acid, none have been proven to date, except for tissue uric acid deposits in patients with severe gout.

  7. Coherence. Coherence between epidemiological and laboratory findings increases the likelihood of an effect. However, there are no such laboratory findings (for example in animal models) where such increased uric acid increases the risk of adverse CV outcomes.

  8. Experiment. No randomized placebo-controlled trial has tested the effect of increased/decreased uric acid on outcomes. Indeed, it may not be possible to design such randomized trials. The FREED trial, presented in this issue of the European Heart Journal, is one attempt to fill this gap with a prospective randomized trial in a population of patients with hypertension who do not have a clear indication for treatment. Unfortunately, this trial was open-labelled, with one-quarter of the patients in the control group treated with allopurinol. However, the results of this study do not show any CV or cerebral benefits of lowering uric acid levels in this population.

  9. Analogy with other causal relationships. The relationship seen between uric acid and outcomes may be analogous to that seen with glycaemia. However, the 'causal' relationship between glycaemia and outcomes itself remains inconsistent.

Although the relationship between uric acid and outcomes does satisfy many of Hill's criteria, substantial gaps and a number of unanswered questions remain. Whether lowering of uric acid leads to CV benefits and not merely CKD benefits is still unclear. The recent FDA decision to put a black box warning on febuxostat argues against such a sequence of events and indicates that, with regard to CV and CKD benefits, not all uric acid-lowering drugs are created equal. However, the present results of the FREED trial are at least reassuring in that CV and cerebral events were not increased in patients with hyperuricaemia and a high CVD risk. Thus, uric acid has moved closer to being a true culprit for CKD, but for CVD it seems to be mired in the status of being a not-so-innocent bystander.

For practical purposes, clinicians may want to remember that:

  1. Uric acid-lowering drugs are not indicated for the prevention or treatment of hypertension, CVD, or CKD.

  2. With regard to uric acid levels, not all CV drugs are created equal.

  3. In the many patients at risk for hyperuricaemia, preference should be given to CV drugs that are either neutral or are able to lower uric acid levels.[14–16]