Is Co-trimoxazole plus RAAS Inhibition a Deadly Combination?

Could a commonly prescribed antibiotic, when combined with a renin-angiotensin system inhibitor, lead to sudden death?  Angiotensin converting enzyme inhibitors (ACE-I), angiotensin receptor blockers (ARB), and spironolactone are frequently used in older adults for a variety of indications.  All can potentially cause hyperkalemia.^1-2 Co-trimoxazole (trimethoprim-sulfamethoxazole) is commonly prescribed for a wide range of infections.  One study found that when co-trimoxazole is used concurrently with an ACE-I or ARB, there was nearly a seven-fold increase in the risk for hospital admission due to hyperkalemia.³ Elderly patients with cardiac and renal dysfunction are particularly prone to developing hyperkalemia — which in turn has been associated with increased all cause mortality.^4,5

A recent study found that elderly patients prescribed co-trimoxazole and renin-angiotensin system (RAS) blockers (ACE-I or ARB) were more likely to suddenly die.⁶  While these findings are alarming, this study has a number of limitations that demand further discussion.

In this case-control study which used the Ontario drug benefit and the Canadian Institute for Health Information databases to identify cases and controls, patients aged 66 years or older who received an ACE-I or ARB between 1994 and 2012 were included.⁶  Cases (n=1027) were patients who died suddenly after receiving an outpatient antibiotic prescription within the previous seven days, including co-trimoxazole, amoxicillin, ciprofloxacin, norfloxacin, or nitrofurantoin, and were matched to controls (n=3733) who did not have sudden death.  Amoxicillin was used as the reference exposure because it is not known to cause hyperkalemia or QT prolongation.  Relative to amoxicillin, both co-trimoxazole and ciprofloxacin were associated with an increased risk of sudden death at 7 days following receipt of the antibiotic (adjusted OR 1.38, 95% CI 1.09-1.76 and adjusted OR 1.29, 95% CI 1.03-1.62, respectively), while nitrofurantoin was associated with a lower risk of sudden death (adjusted OR 0.64, 95% CI 0.46-0.88).  Co-trimoxazole was also associated with an increased risk for sudden death at 14 days after receipt of the antibiotic (adjusted OR 1.54, 95% CI 1.29-1.84), while ciprofloxacin was not.

Approximately two additional cases of sudden deaths were observed per 1000 co-trimoxazole prescriptions dispensed when compared with amoxicillin.⁶  The authors propose that an increased risk of sudden death with co-trimaxazole is secondary to hyperkalemia, while ciprofloxacin may be attributable to QT prolongation. The authors recommend increased monitoring when prescribing co-trimoxazole or ciprofloxacin and also suggest limiting the dose and duration as well as using alternative antimicrobial therapies.

There are several limitations and confounders in this analysis that need to be pointed out. One important limitation — which the authors acknowledge — is the retrospective case-control study design.  Such study designs often find many interesting associations but cannot prove causality.  Hyperkalemia caused by ACE-Is, ARBs, and co-trimoxazole is a well documented dose-dependent phenomena, seen most often in patients with renal dysfunction.^1-3,7-9  However, the authors did not have access to serum potassium concentrations or serum creatinine values.  Nor did they provide data regarding the antimicrobial, ACE-I, or ARB doses.

These results are also confounded by treatment indication.  While the authors matched patients for a number of variables (age, sex, and presence of chronic kidney disease and diabetes), they did not match patients based on the antibiotic indication. Co-trimoxazole has a wide spectrum of activity and range of indications.  It is commonly used to treat and prevent serious/life threatening infections including opportunistic pathogens in transplant and HIV/AIDS patients (e.g. Pneumocystis, Nocardia).  Conversely, amoxicillin, norfloxacin, and nitrofurantoin have relatively narrow spectrums of activity and are typically used to treat less serious infections.¹ Indeed, nitrofurantoin is most often prescribed to treat uncomplicated cystitis in healthy patients with normal renal function. Co-trimoxazole users in the study may simply have been a sicker cohort of patients when compared to amoxicillin users.

Further, there were intrinsic differences in the baseline characteristics among groups in this study.   The cases had higher severity of illness scores (p<0.001); rates of renal disease (p<0.001) and heart failure (p<0.001); higher use of potassium sparing diuretics (p<0.001) and a trend towards higher use of potassium supplements (p=0.08) when compared to controls.  Again, these differences suggest a sicker cohort of patients.⁶

While this study has identified a potentially significant interaction with co-trimoxazole and RAAS blockers, the retrospective design, a number of confounding variables (lack of data regarding treatment indication, potassium, serum creatinine, QT interval, and dose), and differences at baseline between groups (severity of illness, renal disease, heart failure, and potassium sparing diuretics) make the findings questionable. It may be just a red herring.  What do you think?