TAMPA – Why would two patients respond so differently the same medication? Although the reasons are several, recent advances support a prominent role for genetic differences in drug metabolism, according to a panel at the AMDA – Dedicated to Long Term Care Medicine annual meeting.
The good news is that this is not theoretical science. Some variations in patient response can be predicted using commercially available genetic assays, said the genetics and long-term care experts.
One strategy is to test patients for variations in genes for cytochrome P450 enzymes. Before prescribing an antidepressant, for example, you can test a patient for variations in their 2D6, 2C9, and 2C19 enzymes, which metabolize psychoactive medications and many others.
Depending on a patient’s genetics, in other words, he or she might be a poor, intermediate, or ultrarapid metabolizer of a particular medication, which can dramatically increase or decrease efficacy and side effects. Slow or poor metabolizers can experience a build-up of particular drug and extra side effects, whereas a rapid metabolizer might not get a full therapeutic benefit because of a drug’s quick clearance.
"If I have someone who is a 2D6-poor metabolizer, Paxil [paroxetine] is going to be at sky-high level, so instead I might try something like Luvox [fluvoxamine] or Celexa [citalopram]," said Dr. Joel Winner, medical director of Assure Rx Health, a laboratory service offering pharmacogenomics. "I am a clinician as well. I see patients and I use this technology," said Dr. Winner, who also is a private practice psychiatrist in Boulder, Colo.
He estimated that 60% of his patients undergo some form of genetic testing to help him make treatment decisions. "Fifty percent of depressed patients prescribed drugs are nonresponders," Dr. Winner said.
The ability to predict response versus waiting up to 8 weeks to judge a drug’s efficacy is an advantage, said Dr. Amita R. Patel, a geriatric psychiatrist at Wright State University, Dayton, Ohio.
Minimizing side effects is another goal of genetic testing, she added. "Drug-induced side effects, unfortunately, are very, very common and lead to noncompliance," Dr. Patel said. As an example, she described a woman not responding well to duloxetine [Cymbalta]. The patient felt sad, unmotivated, restless, anxious, and worried and reported sleep disturbances and daytime fatigue. Her Patient Health Questionnaire (PHQ-9) depression score was 15. Eight weeks after she switched to desvenlafaxine [Pristiq], her PHG-9 improved to a 5. (Desvenlafaxine and clonazepam are her current regimen.)
"She was not responding well to Cymbalta ... because her serum levels were too low. She was an ultrarapid metabolizer," Dr. Patel said. The patient also was taking insulin and omeprazole, two medications that were lowering her response to antidepressants.
"We are continuing to look at genes that predict weight gain, metabolic syndrome, or other issues relevant to psychiatry," Dr. Winner said.
A 60-year-old patient with treatment-resistant depression is another example. History revealed that the patient was unresponsive to escitalopram and budeprion XL. A switch to nortriptyline (Pamelor) did not significantly improve symptoms. When low-dose fluoxetine (Prozac) was added, the patient developed sedation and experienced a fall.
Subsequent genetic testing showed that the patient had impaired metabolism affected by 2D6, which metabolizes nortriptyline and fluoxetine. "So pharmacogenomic testing is very useful in our practice," Dr. Patel said. "We really need to look at a patient’s 2D6 whenever it is possible."
"Sometimes when we cannot figure out why a patient is oversedated at a low dose, [genetic testing] is something that can answer that question," said Manju T. Beier, PharmD, who is a senior partner for Geriatric Consultant Resources and on the pharmacy faculty at the University of Michigan, Ann Arbor.
Dr. Beier explained that metabolism of prodrugs works counterintuitively. "You need an enzyme to metabolize it to active compound, so you will get higher adverse events if you are a rapid metabolizer and inefficacy if you are a poor metabolizer."
A 79-year-old woman hospitalized with pneumonia was treated with antibiotics and codeine for her cough, for example. She was found unresponsive on day 2 of admission. She was transferred to critical care, intubated, put on a respirator, and given supportive care.
She recovered fully, but genetic testing revealed that the patient "was an ultrarapid metabolizer ... and she converted that small amount of codeine [a prodrug] to large amounts of morphine rapidly," Dr. Beier said.
If a depressed patient also takes tamoxifen and is a poor metabolizer through 2D6, he or she will not convert as much tamoxifen to its active metabolite endoxifen, Dr. Patel said. When choosing an antidepressant for this type of patient, consider mirtazapine or venlafaxine (which do not strongly inhibit 2D6) vs. paroxetine or fluoxetine, she added.