A combination of two common drugs – one an antidepressant, the other used to lower blood cholesterol – may put people at risk of developing diabetes. This unexpected finding shows the benefits of data mining to discover hidden hazards lurking in our medicine cabinets.
Interactions between drugs can be hazardous or even fatal. When a particular combination of drugs is known to be dangerous a warning appears on the drugs' labelling, but there are many potential hazards that are not yet known about.
To look for drug combinations that might trigger diabetes, Nicholas Tatonetti and Russ Altman of Stanford University in California turned to a database called the Adverse Event Reporting System (AERS), run by the US Food and Drug Administration (FDA).
AERS contains reports from doctors whose patients have experienced drug side effects, but some researchers dismiss it as being too "noisy" to yield useful insights. One problem is that many adverse drug reactions never get reported. "There's a pretty big threshold for a physician to decide to submit a report to the FDA," Tatonetti explains.
And when a drug's hazards become well known, the database gets flooded with "me too" reports – for instance, AERS contains more than 70,000 reports about Vioxx, an anti-inflammatory painkiller that was withdrawn from sale in 2004 after being linked to deaths from heart attacks and stroke.
Blood sugar rise
When the team looked directly for drug combinations that caused blood sugar to rise – a symptom commonly associated with the onset of diabetes – they drew a blank, so instead they decided to combine a list of drugs known to affect pathways involved in diabetes, and then mined AERS for side effects associated with these drugs. Then they set their algorithms loose on AERS to find combinations of drugs that produced the same constellation of side effects, thinking that these might also affect pathways involved in diabetes.
One combination stood out: the antidepressant paroxetine and pravastatin, used to lower blood cholesterol.
Next, the Stanford researchers teamed up with colleagues at Harvard University and Vanderbilt University in Nashville, Tennessee, to study the electronic medical records of patients given both drugs. For patients treated at hospitals with ties to each of the three universities, the combination of paroxetine and pravastatin caused blood sugar to rise – especially in patients who were already diabetic.
Finally, the researchers showed that mice made "pre-diabetic" by giving them a diet laden with fat and sugar showed a similar spike in blood sugar when given the drug combination. Neither drug caused blood sugar to rise when given alone.
Tatonetti and Altman estimate that as many as 715,000 Americans were given paroxetine together with pravastatin in 2009. For those with diabetes, the combination's effect on blood sugar may mean that they need to take higher doses of drugs to control the condition; for others it could push them over the threshold into becoming diabetic.
Exactly why the drug combination should trigger a rise in blood sugar is unclear. The problem doesn't occur when you combine other selective serotonin reuptake inhibitors with statins, the classes of drugs to which paroxetine and pravastatin belong. "We were very surprised that it was so specific to these two drugs," Tatonetti says.
The researchers are optimistic that a similar approach will uncover drug combinations that trigger high blood pressure, elevated blood cholesterol and depression.
Nigam Shah, a specialist in biomedical informatics at Stanford who was not involved in the research, now wants to try the approach for himself. "After reading the paper, I will probably try out their method on the roughly one million patients' worth of electronic medical records that my group is analysing," he says.
Photo: Zen my Fitness
Technorati tags: Diabetes, Type 2, Prevention, Antidepressants, Cholesterol, Blood Sugar.
Blogalaxia tags: Diabetes, Type 2, Prevention, Antidepressants, Cholesterol, Blood Sugar.