CARES on Febuxostat 3/3 Application and My Views

I started reading of the red flags over the cardiovascular safety of febuxostat while preparing for a talk to be given during UP Medicine’s Grand Scientific Symposium (2014?). I included data from the meta of Tayar JH (2012) to spark interest in monitoring developments. I belatedly came across the article of WB White in the American Heart Journal (2012) describing the study design for CARES. It’s quite fortunate that 4 years later I’m finally reading of the results. Interestingly, a trial using a PROBE design (allegedly it approximates real world experience better than RCTs) in determining CV safety of febuxostat (vs. allopurinol) had been described in BMJ Open (2014) by MacDonald TM, et al. – The Febuxostat versus Allopurinol Streamlined Trial (FAST).

Let me differentiate the application of CARES study in different patient scenarios.

In patients with known CV disease (prior MI, stroke/ TIA, peripheral vascular disease, unstable angina and diabetes with vascular complications), I’d choose allopurinol as my primary urate lowering drug. Since pegloticase is not available in the Philippines, I’d offer febuxostat as an alternative when the patient develops severe AEs or cannot tolerate allopurinol.  And it’s use should be with full disclosure of the increased risks for CV deaths. As of now, it is not yet known whether aggressive risk modification (ASA, NSAID avoidance, statins, etc) would off-set the increase in sudden cardiac deaths. Right now, I’m already thinking of reviewing my febuxostat treated patients and will probably shift to allopurinol if without contraindications.

In patients without CV disease, the use of febuxostat in treating hyperuricemia may remain unchanged. There are still no studies to suggest a change in practice. But a little caution should be exercised probably in handling patients with higher CV risk (and yes, we’d probably need to start using CV risk calculators in the clinic for this). I wouldn’t hesitate using febuxostat in gout patients with low CV risks (some NSAID trials define this as an estimated 10 year risks of CV mortality <10%). However, for those with high CV risk (estimated 10 year risk of CV mortality >10%), I’d probably be more careful in using febuxostat but the drug still remains an option. Of course, it follows that aggressive risk modification should be implemented. I’d still use febuxostat unless the FDA issues a black box warning for the drug.

Some of you may be wondering why I still prefer placing patients on febuxostat rather than completely shifting all my patients to allopurinol. One word. COMPLIANCE. Regardless of dose, Febuxostat is given as a once a day pill. And if you have a patient being treated for multiple conditions (think HPN, HF, dyslipidemia), then having a once-a-day pill can mean a big difference in compliance.

Notice that I didn’t mention efficacy. If you back read my posts, it was only the febuxostat 80mg which was better than allopurinol 300mg in bringing down urate levels to target. In fact, as a personal rule of thumb, allopurinol 300mg may be as effective as febuxostat 40mg in bringing down SUA <6mg/dl. (Therefore, allopurinol 600 mg/d may be as effective as febuxostat 80mg/d). Maximizing allopurinol use (yes, it can be used to 800mg/d!) can bring about similar outcomes in reducing urate levels. Likewise, combining allopurinol with other drugs with uricosuric properties (losartan, fenofibrate) may help achieve those targets. If you bothered to check the secondary endpoint of the CARES study, the proportion of patients who experienced a treatment needing flare and those who achieved SUA <6mg/dl were almost comparable between groups. Allopurinol remains an effective urate lowering drug, we just have to learn to maximize it. And, please, no more Hande protocol!

The CARES study just complicated an already challenging disease to treat. It’s challenging because gout is a very treatable disease and successful management of hyperuricemia can avert complications and repeated attacks. Now aside from NSAIDs, we’d have to be wary of the impact our ULT has on heart health. Hopefully, we are clarified of this issue once the results of FAST come out (and hopefully, a meta-analysis can come out).


CARES on Febuxostat 2/3 The Appraisal

For my mentor (HMR) who’s fond of how investigators name their trials, the full name of the study was “CArdiovascular Safety of Febuxostat and Allopurinol in Patients with Gout and Cardiovascular ComoRbiditiES“. The article discussing the results was published in the NEJM just last 12 March 2018. A link to the article can be found here.

PATIENTS. The CARES study was a multi-center (from the US, Canada and Mexico), randomized, active-control, non inferiority trial among patients with gout and cardiovascular disease.

Cardiovascular disease was defined as any of the following:

  • Myocardial infarction (but not in the 2 months preceding the randomization)
  • Hospitalized unstable angina
  • Stroke (but not in the 2 months preceding the randomization)
  • Hospitalized transient ischemic attack
  • Coronary or cerebral revascularization procedure
  • Peripheral vascular disease
  • Diabetes mellitus with evidence of micro- or macro-vascular disease (retinopathy, nephropathy, neuropathy or small vessel disease)

Gout was diagnosed using the older ARA Criteria (Wallace) and not the newer 2015 ACR EULAR Criteria.

INTERVENTION/ COMPARISONS. Patients were to receive either febuxostat or allopurinol. Randomization was done based on renal function.

Patients randomized to febuxostat received 40mg initially regardless of renal function. The dose was increased to 80 mg after 2 weeks if SUA was not <6mg/dl. Patients with GFR >60ml/min and randomized to allopurinol received 300mg initially and the dose was titrated by 100mg until a max dose 600mg/d or the target SUA was achieved. Those with GFR 30-60 ml/min, received allopurinol 200mg initially then titrated up by 100mg until a max dose 400mg/d or the target urate levels were reached.

All patients received colchicine 0.6 mg/d for the first six months. If colchicine couldn’t be tolerated, either naproxen 250 mg BID + lansoprazole 15 mg OD (if eGFR>50ml/min) or prednisone <10mg/d (if eGFR<50ml/min) were given. There was no specific duration of treatment as termination of the trial hinged on the accrual of 624 Major Adverse Coronary Events.

OUTCOME. The primary endpoint was the occurrence of Major Adverse Coronary Events (MACE) which includes: CV death, non-fatal MI, non-fatal stroke and unstable angina with urgent coronary revascularization.

Secondary endpoints were the: (1) mean number of flares from the end of the first to the end of the fifth year; (2) percentage of patients with tophi resolution by the end of the fifth year; (3) Anti-Platelet Trialist Collaboration (APTC) endpoint – which is basically MACE excluding unstable angina; (4) Pre-defined CV endpoints such as hospitalized CHF, arrhythmias, venous or arterial peripheral thromboembolic events and TIAs; (5) Proportion of patients achieving SUA <6mg/dl at end of years 1-5.

Were patients randomized to treatments? Yes. via IVRS.

Were the groups similar at the start of the trial? Yes. See Table 1 of the article.

Were groups treated equally? Yes. The protocol (available as a supplement) provided details on this.

Were all patients accounted for? Yes. And there was a big number of patients who discontinued trial treatment (57.3% vs. 55.9%) and who did not complete the trial visits (45% vs 44.9%) in both the febuxostat and allopurinol arms, respectively. (If you had doubts about the PRECISION trial over the high dropout rates, then …)

Were patients analyzed in the groups they were randomized? Yes. The modified ITT just did not include the patients who never received treatment. A total of 6190 patients were included in the analysis. Only 8 patients were excluded from the mITT.

Were patients and clinicians blinded to treatment? Yes. And it appears the outcome assessors were too.

What were the results? The outcome of interest was the MACE. You may read the article if you’re interested in other outcomes.


There were significantly more CV deaths and Death from any cause in the febuxostat-treated patients. That’s 34% more CV related deaths and 22% more deaths from any cause in the febuxostat group. Hence, the grave concern over the results of this study.

Now, here’s heading to the study supplement and looking at the breakdown of causes of CV deaths.

Adjudicated Causes of CV Deaths (CARES Study

We see that most of the CV related deaths were from sudden cardiac death (defined as death that occurs unexpectedly in a previously stable patient and includes the following deaths: (1) witnessed and instantaneous without new or worsening symptoms; (2) witnessed within 60 minutes of the onset of new or worsening cardiac symptoms; (3) witnessed and attributed to an identified arrhythmia; (4) subjects unsuccessfully resuscitated from cardiac arrest or successfully resuscitated from cardiac arrest but who die within 24 hours without identification of a non-cardiac etiology; and, (5) unwitnessed death or other causes of death.) And a review of all-cause mortality revealed that the increase was brought about by these CV deaths – specifically SCDs.

An interesting association with the study was the impact of some patient characteristics on CV deaths. The relative risks for the following factors wereas follows:

  • NSAIDs use: RR 2.17 (95% confidence interval 1.28, 3.68)
  • Colchicine as prophylaxis: RR 1.38 (95% confidence interval 1.05, 1.81)
  • Non-colchicine as prophylaxis: RR 1.09 (95% confidence interval 0.53, 2.23)
  • Non-use of low dose ASA: RR 1.99 (95% confidence interval 1.39, 2.85)
  • Use of low dose ASA: RR 0.85 (95% confidence interval 0.58, 1.24)
  • Non-use of insulin among DM: RR 1.46 (95% confidence interval 1.06, 2.00)
  • Use of insulin among DM: RR 1.13 (95% confidence interval 0.86, 1.66)

These would suggest that agents other than colchicine be used for gout prophylaxis (in the study, most patients intolerant to colchicine were given naproxen followed by steroids). We do not know if limiting the duration of colchicine use would reduce the flares.

Now some may be wondering why using NSAIDs for prophylaxis didn’t result in increased CV deaths but NSAID use was. Well, I suspect 2 reasons: 1. There is a difference in the dose used for prophylaxis vs treatment of flares. We use higher doses when treating flares and there is a dose-dependent relationship between NSAIDs and CV events. Even in the CNT meta-analysis, we see that rates of CV events increased with NSAID dose (at least, for most agents). And, (2) the NSAID used for prophylaxis was naproxen which has comparable COX1: COX2 activity as ASA (only reversible). Naproxen has long been “believed” to be the most heart-friendly among the NSAIDs. Unfortunately, it’s notorious for causing GI upset.

And if patients need glycemic control, insulin is the preferred agent.

CONCLUSIONS. While rates of major adverse cardiovascular events were similar between groups, all cause mortality was significantly higher (22% more) among febuxostat treated patients – which were driven mostly by an imbalance in CV-related deaths, particularly SCDs. Rates for other components of the MACE (non fatal MI/ stroke, UA needing revascularization) and other predefined CV safety outcomes (arrhythmia, peripheral vascular disease and HF) were not statistically different. The impact of using some drugs to minimize risks of CV related deaths warrants a separate study.

THESE ARE TROUBLING RESULTS. Do we now stop using febuxostat because over these concerns? Will febuxostat go the way of rofecoxib?

Next: Applying the Results to Patient Care and My Take of the Results

Reference article: White WB, Saag KG, Becker MA, et al. Cardiovascular Safety of Febuxostat or Allopurinol in Patients with Gout. NEJM 2018; 378: 1200-1210.

CARES on Febuxostat 1/3 The Scenario

It was welcome news when febuxostat was introduced to the Philippine market. Pinoys have long been identified in rheumatology texts as one of the groups with the highest prevalence of gout. A local study found that, among the 25% of the population with hyperuricemia, 1.6% actually had gouty arthritis (1). That’s 1.6 million people needing treatment for the condition. Sadly, 50% of these patients have complications (i.e. tophi or kidney stones) by the time they seek medical consult (2). And the lure of using febuxostat was made more attractive when the local distributor introduced it at a very competitive price – PhP26 or ~US$0.50 – compared to the Western markets (I heard it was US$10).

Prior to febuxostat, allopurinol was the go to drug for treating hyperuricemia. Aside from uricosurics being more expensive and often unavailable (as of now, both probenecid and sulfinpyrazone are no longer available in the Philippines), many patients preferred the convenient once a day dosing of allopurinol. Unfortunately, many physicians limit their use to either 100 or 300mg. On checking local drugs references (MIMS, PPD), 300 mg was cited as the maximum allowed dose. This is in contrast to the 800mg allowed in the US or 900mg in the EU. You would infrequently encounter a patient with >300mg allopurinol. And, if ever you do, the dose would be reduced (instead of maintained) as soon as urate levels normalize.

Strong concerns over SJS, TENS and allopurinol hypersensitivity syndrome made going beyond 300mg very unattractive. And as if that were not enough, concerns over the impact of renal impairment (on allopurinol-related adverse events) lead many to abide by the Hande protocol when starting the drug. So really, Pinoys on allopurinol were not really maximizing the use of the drug.

The first two years of industry-sponsored talks focused on 80mg febuxostat being more effective than 300mg allopurinol in bringing down serum urate levels to <6mg/dl. The safety profile of febuxostat 40mg was comparable to allopurinol 300mg. And only GI adverse events (transaminitis, abdominal discomfort) increased with higher doses of febuxostat. The incidence of skin reactions were similar between the two drugs (although, personally, I have yet to encounter a patient who developed them on febuxostat). Since the two drugs were structurally different, febuxostat was initially indicated as an alternative to allopurinol when a patient developed a hypersensitivity reaction. However, when the target for tophaceous gout was pegged at <5mg/dl, many looked past this indication and preferred starting patients on febuxostat. And soon enough, even practice guidelines (3) suggested choosing between the alllopurinol and febuxostat as first line drugs for hyperuricemia in gout. No longer was febuxostat an alternative, it became an option. And, for the first time in the Philippines, costs wasn’t much of a factor (compared to our Western counterparts) in choosing one drug over the other.

Towards the third year, we started hearing about possible alarms over the CV safety of febuxostat. Doing a meta-analysis of febuxostat RCTs (4) showed no significant difference in cardiovascular events. In fact, the meta-analysis showed a lower risk of CV events with febuxostat (40/ 80mg) compared to allopurinol (200/ 300mg). However, the 95% confidence intervals suggested NO statistical difference. But we have to realize that none of the first RCTs were actually designed to detect significant differences in these adverse events. And doing a network meta was just a way to find evidence on a concern not yet properly addressed by any existing RCT. (Note to self: Ask Drs Tony and Inday Dans which is better between the two – a network meta of secondary endpoints or an RCT having it as a primary endpoint.)

Here is where the Cardiovascular Safety of Febuxostat or Allopurinol in Patients with Gout Study or CARES Study comes in. It’s the first RCT to directly address this concern over the CV safety of febuxostat. And in contrast to FACT, APEX, FOCUS, EXCEL and CONFIRMS, this had CV safety as the primary endpoint while efficacy was checked as a secondary endpoint. You can find a link to the article here.

Next: The Appraisal of CARES Study


  1. Salido EO, et al. Phil J Int Med 2008; 46: 273-6.
  2. Roberto LC, et al. Poster Presentation. PRA Annual Meeting 2017.
  3. Khanna D, et al. 2012 American College of Rheumatology Guidelines for Management of Gout. Part 1. Arth Care & Res 2012; 64 (10): 1431-46.
  4. Tayar JH, et al. Febuxostat for treating chronic gout. Cochrane Database of Sys Rev 2012; 11: Art No CD008653.

Vitamin B for Aches and Pains

Even before you can properly assess any complaint of joint pain, paresthesias, tingling, numbness or ngalay, the typical Filipino patient already has an expectation that he’ll be prescribed vitamin B. And if you do decide NOT to give it, the Pinoys will make sure to ask you about it. Some physicians give in, thinking there’s no harm to it. Others take the time to explain why they didn’t include it. But what is the state of evidence surrounding the use of vitamin B in various painful conditions?

I had the chance to review it during the 12th Post Graduate Course of the East Avenue Medical Center Department of Internal Medicine. Here were my slides for that presentation last 16 September 2016.


ACR 2015 Guidelines on Rheumatoid Arthritis Part 2: Conflicts

There’s this expectation that guidelines put perspective into the scheme of things as new evidence is obtained on the use of both novel and established agents. Personally, a well made algorithm is the core of guidelines as it organizes the tools available to a physician after weighing in concerns such as efficacy, safety, tolerability and, hopefully, costs (biologics and targeted synthetic DMARDs are far from being cheap!).

The following are my opinions on parts of the ACR 2015 Guidelines on Rheumatoid arthritis  that I feel more time should have been devoted to in clarifying issues. Others may call these missed opportunities (see  the editorial by van Vollenhoven R.  Nature Rev Rheum 2016; 12. doi: 10.1038/nrrheum.2015.181). I’d refer to them as conflicts as these added more confusion rather than organization to how we manage the disease. Feel free to discuss them with me as I’d appreciate more ideas on these matters. Continue reading “ACR 2015 Guidelines on Rheumatoid Arthritis Part 2: Conflicts”