An 88 year old woman was admitted to the hospital for atrial fibrillation with a rapid ventricular response. She complained only of palpitations. Her only medication was digoxin, of which she ran out a week earlier. After blood sampling for a digoxin level, initial therapy was begun. Intravenous digoxin, 1 mg was given in divided doses over three hours. The patients heart rate slowed (repeat cardiogram revealed an otherwise "normal" atrial fibrillation) and she was admitted to the CCU for further care. Her pre-digitalization serum digoxin level was zero; the remainder of her serum chemistry analysis was normal. A digoxin level drawn one hour after administration of the final dose of digoxin was 5 ng/mL. To be sure that this level was falling, a repeat level was drawn several hours later and was reported as > 100 ng/mL.
Is the initial post infusion level of 5 ng/mL concerning?
Proper interpretation of a digoxin level is predicated on proper sampling techniques. This requires blood sampling to occur 4-6 hours after intravenous administration of the digoxin. This allows redistribution of the medication from the central compartment (blood and highly perfused organs) to the peripheral compartment (skeletal and cardiac muscle in the case of digoxin). If sampled before distribution is complete, the digoxin level will be elevated. However, only the equilibrated digoxin level is useful in clinical practice.
Decision making with regard to toxicity combines both the clinical condition of the patient, the electrocardiogram and the serum digoxin level. In this woman, with no complaints, a normal physical examination, and an essentially normal cardiogram (baseline), only a level sure to produce toxicity upon distribution would be concerning. This level is generally taken to be 15-20 ng/mL. The management strategy for this patient should be to observe, make sure the serum potassium is adequate (hypokalemia is associated with enhanced toxicity), and repeat the cardiogram. The potassium is additionally helpful in predicting outcome of overdosed patients. It was documented twenty years ago that a potassium greater than 5.0 mEq/L is associated with a very high fatality rate (about 50%) and if greater than 5.5 mEq/L a 100% mortality. This study (Bismuth 1973) was performed in the pre-immunotherapy days (see below) so while the mortality is now lower, use of the serum potassium level as an indication for therapy remains.
The use of digoxin to control atrial fibrillation is based on the drug's ability to prolong the effective refractory period of the AV node (increases AV block). It is therefore able to slow the progression of atrial impulses to the ventricles, slowing the heart rate. Digoxin is not associated with an increased incidence of conversion to sinus rhythm from atrial fibrillation. If such conversion is later deemed necessary, caution needs to be exercised as both commonly used techniques are risky in patients on digoxin. Quinidine reduces tissue and protein binding, resulting in an elevated digoxin level, and also reduces the drug's renal elimination. Electrical cardioversion should be used with extreme caution in digitalized patients as conduction abnormalities and dysrhythmias frequently result. There does not appear to be an interaction with adenosine (other than enhanced duration of AV block); if such is felt necessary to differentiate atrial fibrillation or flutter from a re-entrant supraventricular tachycardia it can be used cautiously.
What caused the level on repeat analysis to be so high?
The first question to answer is whether this level is real. Assuming that the patient truly received only 1 mg of digoxin, the maximum achievable serum digoxin level can be calculated (knowing the volume of distribution [Vd] and weight of the patient in kg) by the following:
Concentration = dose (Vd x weight)
= 1 mg (5L/kg x 50 kg)
= 4 ng/mL (after converting units)
Therefore, a level of greater than 100 is not attainable after a 1 mg dose of digoxin. It is possible that the patient was taking cardiac glycosides from another source (many plants and certain animals). However, her predigitalization serum level was zero. This was presumably measured by the same method as the later concentrations.
Certain subgroups of patients develop elevated levels of an endogenous substance that cross reacts with the digoxin assay. The identity of the digoxin like immunoreactive factor (DLIF) is still largely a mystery. Depending on the assay used, levels may be non-detectable to as high as 3 ng/mL (not even close to 100!). However, this patient is not in a group known to have measurable DLIF, which includes neonates, pregnant women, and patients with renal or hepatic failure.
A common reason that astronomical digoxin levels are obtained is the administration of digoxin specific Fab (Digibind ®). Fab is the antigen recognizing portion of the intact antibody, and is therefore very large in comparison to the digoxin molecule. While the digoxin molecule has the ability to leave the blood and enter the tissue (thus the large volume of distribution), the Fab are too big to diffuse from the intravascular space. Upon administration of Fab, the extravascular digoxin is drawn back into the intravascular space. Depending on the assay used to detect the digoxin, the post-Fab serum digoxin concentration may be zero (or very low) or very high. If the assay measures only free digoxin, the level will be low, since if enough Fab was administered, no free digoxin will be available. If the assay measures total digoxin, however, the post Fab level typically may be elevated up to 20 times the predosing level. In addition, the simple administration of digoxin specific Fab to many assays results in falsely elevated serum digoxin levels with several common assays.
What was going on with this patient?
Upon receiving the serum digoxin level, the house officer was concerned that the patient was poisoned by digoxin. Digoxin specific Fab was administered and the repeat level was analyzed with an assay not specific for free digoxin. Fortunately, more Fab was not administered when the digoxin concentration was > 100. She would have needed 50 vials calculated roughly as:
# of vials of Digibind ® = [serum digoxin concentration (ng/mL) x weight (kg)] 100
Digibind costs approximately $200 per vial.
Further readings:
Duhme DW, Greenblatt DJ, Koch-Weser J. Reduction of digoxin toxicity associated with measurement of serum levels. Ann Intern Med 1974;80:516-519.
Rainey PM. Effects of digoxin immune Fab (ovine ) on digoxin immunoassays. Am J Clin Path 1989;92:779-786.
Jiang F, Wilhite TR, Smith CH, Landt M. A new digoxin immunoassay substantially free of interference by digoxin immunoreactive factor. Ther Drug Monit 1995;17:184-188
Gilman AG, Rall TW, Nies AS, Taylor P eds. Goodman and Gillman's The Pharmacologic Basis of Therapeutics. Ed. 8. Pergamon Press, New York, 1990.
Bismuth C, Gaultier M, Conso F, Efthymiou ML. Hyperkalemia in acute digitalis poisoning: prognostic significance and therapeutic implications. Clin Tox 1973;6(2):153-162