By Laurence Kinsella, MD, FAAN, Saint Louis University, Saint Louis, MO
In 1984, an 18-year-old college freshman died in New York Hospital. Few events have had as great an impact on the training of medical residents. Libby Zion was admitted for agitation, confusion, and muscular twitching. She had a history of depression and was taking phenelzine, an MAO inhibitor. The house officers assigned to her care sedated her with meperidine and haloperidol and placed restraints to prevent self-harm. By the following morning, she had a fever of 107 and died from cardiac arrest. Her father, Sidney Zion, a prominent journalist for the New York Daily Post, brought charges against the hospital and the physicians, indicting the medical training system for excessive work hours and poor supervision that, he argued, contributed to poor judgment and medical negligence.1
In 1995, the jury in the Zion v. New York Hospital trialreturned a mixed verdict, finding that the doctors were partially responsible for Libby's death, but that Libby was also responsible based on autopsy samples positive for cocaine metabolites.2
As a result of Libby's death—and her father's considerable influence—the Bell Commission was convened in New York to address the issue of residency work hours.3 In 2003, the Accreditation Council for Graduate Medical Education (ACGME) adopted most of the Bell Commission's recommendations, restricting residency work hours at all US training programs to their current levels of 80 hours per week.4
But would the current work hour restrictions have saved Libby Zion? Would a well-rested, post-Bell Commission resident have recognized the signs and symptoms of serotonin syndrome that Libby Zion exhibited on admission, namely confusion, agitation, and muscular hyperactivity? Would s/he have known that meperidine is associated with significant drug interactions that might worsen serotonin syndrome?
The lesson to be learned from Libby Zion's death is that drug-drug interactions (DDI) and adverse drug events (ADE) are under-recognized, and many are unaware of the potential for harm in many commonly prescribed medications.
A number of clinical syndromes relevant to neurologic practice result from drug toxicity (Table 1). Several are associated with acute confusional states, a common reason for neurologic consultation.
| Syndrome | Medication Examples |
| Serotonin Syndrome | SSRI, meperidine, MAOI |
| Neuroleptic Malignant Syndrome | haloperidol, chlorpromazine |
| Akinetic-Rigid Syndromes | metoclopramide, neuroleptics |
| Acute Confusional States | baclofen, topiramate, many others |
| Anticholinergic Syndrome | TCA, trihexiphenidyl |
| Orthostatic Myoclonus/Asterixis | gabapentin |
| Stevens Johnson Syndrome | carbamazepine, lamotrigine |
| Drug-Induced Seizure | buproprion, theophylline, TCA |
The term "serotonin syndrome" was coined by Sternbach in 1991.5 It is an increasingly common complication of serotonergic drugs, especially when used in combination.6 The incidence is unclear, since most cases are probably unrecognized. Over 200 cases have been published, most since Sternbach's review.7 Toxicity occurs in 27% and deaths in 0.3% of patients overdosing on SSRIs.8 For nefazadone alone, the incidence is 0.4 cases per 1,000 patient-months.9 Patients present with a clinical triad of mental status changes, autonomic instability, and motor hyperexcitability, usually within 24 hours of a change in medication.10 Most patients recover within one to two days after drug withdrawal; however, some may develop respiratory failure, seizures, rhabdomyolysis, and cardiac arrest. Most of the deaths reported due to serotonin syndrome occurred in patients taking an MAO inhibitor.11
Adverse drug events (ADEs) are common. Nearly seven percent of all hospitalized patients experience ADEs, with a fatality rate of 0.32%.12 Drug-drug interactions (DDIs) account for a quarter of all adverse drug events. Some suggest that ADEs are the fifth leading cause of death in hospitalized patients.17
Risk factors for ADEs and DDIs include: age greater than 65, multiple medications and over-the-counter medications (OTCs), genetic variability in drug metabolism, and medical comorbidity. Sixty-four percent of outpatient visits result in a prescription.13 In 2000, there were 2.8 billion prescriptions written, 10 for every person in the US. ADEs rise with number of prescribed medications, exceeding 50% likelihood over four medications.
The P450 enzyme system is an important determinant of drug interactions. Formerly, protein binding was considered an important cause, yet it is rarely clinically relevant.14 Even highly protein-bound drugs such as phenytoin and warfarin quickly alter binding ratios when co-administered and achieve a new steady state. Less well appreciated by practicing physicians is the importance of the few P450 enzymes that metabolize 90% of all medications, and that many medications may either inhibit or induce the metabolism of other drugs. Further, these enzymes are subject to genetic variation, making some patients susceptible to toxicity at low doses.15,16
Drugs undergo phase 1 and phase 2 metabolism. Phase 1 is carried out by the P450 system, primarily in the liver, and includes oxidation, hydroxylation, and methylation. Monamine oxidation, also phase 1, is not part of the P450 system.
Phase 2 prepares the drug for elimination. Glucuronidation makes the drug water soluble for elimination in the urine or stool. Further discussion of phase 2 metabolism may be found in Sirot et al.17
These enzymes are located primarily in the liver, kidney, intestine, lungs, and brain. Six enzymes metabolize over 95% of all medications.16,17 They are CYP1A2, 2B6, 2C9, 2C19, 2D6, and 3A4. CYP2D6, 2C19, and 2C9 are especially prone to genetic variability. Depending upon the number of copies of a particular allele, patients may be poor metabolizers (no functioning alleles), intermediate metabolizers (one copy), extensive metabolizers (two copies—the wild type), or ultra-metabolizers (three to thirteen copies).
|
Polymorphism |
Frequency |
Interacting Drugs |
CYP2C9 poor metabolizer |
1–10% of Caucasians |
Phenytoin, warfarin, glipizide |
CYP2C19 poor metabolizer |
13–23% of Asians |
Barbiturates, benzodiazepines |
CYP2D6 poor metabolizer |
5–7% of Caucasians |
Amiodarone, TCA, neuroleptics |
CYP2D6 ultra-metabolizer |
5–20% of Turks, Southern Europeans, Saudi Arabians, Ethiopians |
Opioid intoxication with codeine |
Pharmacogenomics is the study of the marked variability in drug metabolism in individuals. Screening tests are available from a number of laboratories to assess the likelihood of genetic susceptibility to a drug interaction (cf. Genelex and others). Phillips reviewed 18 studies of ADEs related to genetic variations of CYP enzymes. Of 27 drugs identified, 59% were metabolized by a polymorphic enzyme, including cardiac, psychiatric, antibiotic, and analgesic classes.18
Although 2D6 polymorphisms are present in 7% of the population, 14% of hospitalized psychiatric patients have 2D6 variants, suggesting a far greater risk of adverse drug events requiring hospitalization.19
1. Take a medication history (mnemonic—AVOID Mistakes)
Allergies?
Vitamins and dietary supplements
Old drugs and OTC?
Interactions risk?
Dependence?
Mendel: any family history of drug sensitivity?
2. Identify high risk patients
> 3 medications
red-flag drugs—anticonvulsants, antibiotics, digoxin,
warfarin, amiodarone
3. Check pocket reference card
4. Consult pharmacist/ drug specialist
5. Check computer programs
Epocrates
Medical letter drug interaction program
Source: Preventable Adverse Drug Reactions: A Focus on Drug Interactions
Within the past 24 months, Dr Kinsella has received honoraria for speaking for American Medical Seminars, Medical Education Resources, and CME LLC. He has also received honoraria for consultation with Cross Country Education and Therapath Laboratories. Additionally, he held stock in Passnet Air Ambulance, serving Native American communities in South Dakota. Dr Kinsella has also received personal compensation for case reviews and serving as an expert witness.