DILANTIN TOXICITY IN THE ICU
Sanjeev Patel, T.J. Gallagher, A.J. Layon
Department of Anesthesiology, University of Florida College of Medicine
Case Report: A 76-year-old man involved in a motor vehicle
accident sustained bilateral subdural hematomas. He was admitted
to the surgical intensive care unit (SICU) for close
neurological assessment and monitoring. Medications included
phenytoin 100 mg every 8 hr for seizure prophylaxis. He became
increasingly lethargic, and subsequently underwent a left
craniotomy for evacuation of hematoma. Postoperatively, he was
again awake and alert with intact speech and language, and
without focal neurologic deficits.
On postoperative day (POD), he became increasingly somnolent and
febrile. He was empirically started on broad-spectrum
antibiotics. Intravenous fosphenytoin 500 mg was administered
for a plasma concentration of 4.8 mcg/ml, and his next phenytoin
dose was increased by 100-mg. On POD 5 he was afebrile, but
found to be increasingly lethargic, intermittently following
simple commands. There was no improvement in his neurological
status over the subsequent three days. He remained NPO during
his hospital course. On POD 8, his mental state deteriorated to
the extent that he ceased to follow commands, and had ‘myoclonic
jerks of his extremities’. He was intubated for airway
protection. All cultures remained negative. A CT scan of his
head showed no new pathologic findings, but EEG monitoring
showed frequent epileptiform discharges. He received two
additional boluses of fosphenytoin for a level of 5.0 mcg/cc. On
POD10, he received four additional boluses of fosphenytoin for a
level of 10-13 mcg/cc, and his regular dose was doubled to 200
mg three times daily. In addition, he was started on valproic
acid and levetiracetam, and EEG monitoring discontinued. His
albumin on POD 10 was 1.9 mg/ml. On PO day 12, an unbound ‘free’
phenytoin level was sent, because of his unchanged neurological
status and low albumin concentration. This was found to be 4.6
mcg/ml (normal 1-2 mcg/ml), with a normal total level of 13
mcg/cc (normal 10-20 mcg/ml). Phenytoin was discontinued. Over
the next few days, as his free phenytoin levels normalized, he
began to show marked neurological improvement. He was extubated
after a successful T-piece trial. At the time of discharge from
the SICU, he was alert and orientated, with some residual
weakness of his right upper extremity.
His deterioration in mental state was attributed to phenytoin
toxicity, because the clinical improvement upon stopping the
drug coincided with a normalization of his free phenytoin
levels.
Discussion: In certain situations, as in our patient’s case,
reliance on total phenytoin levels can be misleading.
Hypoalbuminemia or concomitant therapy with highly protein-bound
drugs, such as ceftriaxone, may reduce the number of binding
sites and thus increase in the fraction of unbound drug, causing
toxicity despite normal total drug levels. The unbound, ‘free’
fraction, which is usually less than 10% (1-2 mcg/cc) of the
total plasma concentration, may increase to 20-30% of the total
plasma concentration. In addition, patients with acute traumatic
injury may have altered protein binding, with a decreased
affinity of albumin to phenytoin. Signs of toxicity include the
classic findings of cerebellar signs, nystagmus, diplopia, and
vomiting. Less well-known manifestations, as in our patient’s
case, include a more gradual decrease in the level of
consciousness and a paradoxical increase in seizure activity.
The latter is most often seen with very high or rapidly rising
toxic concentrations. Direct measurement of the unbound ‘free’
fraction of the drug as a guide to therapy can avoid this
complication.
