A 52yr old white male with history of heart failure presents to the Emergency Dept. complaining of nausea, vomiting, and a decreased level of consciousness.
Although nearly every imaginable cardiac dysrhythmia has been linked to digitalis poisoning, junctional tachycardia remains uniquely suspicious for this toxidrome. In order to understand the cellular mechanisms connecting digoxin with this and other highly suggestive EKG signatures, the enzyme-level pharmacodynamics must be appreciated.
The direct cardiotonic effects of digitalis arise from inhibition of the transmembrane ion exchange protein, Na+/K+ATPase. Through the exchange of two extracellular potassium ions for three intracellular sodium ions, the phosphorylation of this complex creates a disequilibrium of monovalent cations necessary to the maintenance of the cell’s 80-90mv resting membrane potential.
Fig 1. Conformational shifts of Na+/K+ATPase relative to ion and phosphate binding. Note that the net result of this cycle is the establishment of intracellular hyponatremia.
When digitalis binds to the extracellular surface of the Na+/K+ATPase, a local deformation of the protein’s tertiary structure cripples the ion transport function of the complex . Na+ export is halted and the intracellular environment becomes relatively hypernatremic. This, in turn, exerts a critical effect on yet another ion exchange system—the Na+/Ca2+ antiporter. Normally, the steep Na+ ion concentration gradient across the cell membrane provides an osmotic motive for the Na+/Ca2+ antiporter to drive excess Ca2+ out of the cell. With the Na+/K+ATPase inhibited, however, intra and extracellular sodium concentrations equilibrate. Intracellular calcium levels rise and the sarcoplasmic reticulum becomes over-saturated; cellular depolarization thus triggers a heaver tide of Ca2+ ions and the contractile apparatus responds with greater force.
Fig 2. Inhibition of the Na+/K+ATPase results in elevated intracellular Na+; this stymies the Na+/ Ca2+ exchanger and causes intracellular Ca2+ to rise. Pro-contractile inotropic effects ultimately result.
Yet inhibition of the Na+/K+ATPase has its detriments. With the suppression of normative ion exchange comes a reduction in the ability of the conduction tissues to maintain their 80-90mv membrane resting potential. The gradual influx of cations due to natural membrane permeability cannot be opposed by active transport, and the resting voltage of the intracellular space becomes increasingly positive. As the voltage difference across the cell membrane approaches the trigger threshold of the action potential, the excitability of the conduction tissues rises proportionally. Graphically this can be appreciated below—the slope of the baseline intracellular voltage (phase 4) is seen to rise as cations “leak” into the cell making it less negative. Ultimately, the trigger threshold is reached and the cell depolarizes automatically.
Fig 3. Unipolar recording of a transmembrane action potential from a Purkinje fiber. Control conditions are traced with a solid line, digitalized tissue with dashed line. Note the morphological similarities between the experimental digitalis-altered ST segment presented here and the ST segments seen in the precordial distribution of the case study EKG above.
Due to this effect, digitalis enhances the automaticity of the conduction tissues, encouraging the independent activity of ectopic pacemakers. Depolarization occurs not only more automatically, but also more readily due to the heightened excitability of the cells. This results in a shorter R-T interval and a net positive chronotropic influence. Often in digitalis toxicity the rate of ectopic pacemaker depolarization is accelerated beyond the typical upper limit of the cellular tissue, as seen in the title EKG. Perhaps not ironically given Paracelsus’ adage, “only the dose,” the adverse effects of digitalis toward increased automaticity and excitability, therefore, stem from the same mechanistic activity by which the drug confers its beneficial inotropic influence.
Having thus looked more closely at the direct enzyme-level mechanistics of the cardiac glycosides, it is not surprising to find that the greater portion of dysrhythmias arising from digitalis toxicity consist in ectopic tachycardias such as multifocal atrial tachycardia, junctional tachycardia, and (often bifocal) ventricular tachycardia– as seen below.
Fig 4. Bidirectional ventricular tachycardia. Note the alternating QRS axes and right bundle-branch block type morphology. This occurred in the setting of a supratherapeutic serum level of digoxin as a consequence of acute renal failure.
Yet the mainstay of digitalis pharmacotherapy in the modern era lies in controlling rather than encouraging tachydysrthymias; the treatment of atrial fibrillation, for example, remains central to the role of this drug in the contemporary pharmacopoeia. To explain this seeming contradiction, we must appreciate the scope of the indirect influence in digitalis therapy.
Although less well understood, the anti-chronotropic power of the cardiac glycosides appears to be largely mediated through vagomimetic mechanisms. Increases in efferent vagal impulses, decreases in sympathetic tone, modifications of nerve fiber excitability, and sensitization of arterial baroreceptors have all been described as contributors towards this effect . In supratherapeutic concentrations, the vagal activity of digitalis becomes pathological, giving rise to the bradycardic dysrhythmias—sinus bradycardia and various forms of AV block.
Ultimately what we encounter in digitalis toxicity is a pharmacodynamic system capable of inciting almost any imaginable dysrhythmia and imitating any electrophysiologic pathology. The prevalence of junctional tachycardia in this context may be understood as a logical result of excessive supraventricular vagotonia coupled with enhanced automaticity and excitability of distal ectopic pacemakers.
In the case presented here, laboratory assays returned a substantially elevated serum Digoxin level securing the diagnosis of cardiac glycoside poisoning. This pt. received conservative treatment and was discharged on the third hospital day without incident.
The title of this post is quoted from Louis N. Katz, widely known for his work in electrocardiography. In addition to many articles, he is author of Introduction to the Interpretation of the Electrocardiogram (1952), and Electrocardiography Including an Atlas of Electrocardiograms (1946).
Fig. 1: Graphic on loan via http://www.angelfire.com/sc3/toxchick/celmolbio/celmolbio12.html
Fig. 2: Graphic on loan via http://www.cvpharmacology.com.
Fig. 3: A. Goodman Gilman. The Pharmacological Basis of Therapeutics. Pergamon Press, NY 1990. p. 819.
Fig. 4 and explanatory subtext. Joseph L. Kummer. Bidirectional Ventricular Tachycardia Caused by Digitalis Toxicity. Circulation. 2006; 113:p156-156.
 In depth discussion of the molecular mechanistics involved with glycoside / ATPase binding can be explored via H. Ogawa et al. Crystal Structure of the Sodium-Potassium Pump with Bound Potassium and Ouabain. Proceedings of the National Academy of Science, Vol. 106, No. 33, 2009, pp.13742-13747. See also, S.M. Keenan et al. Elucidation of the NaKATPase Digitalis Binding Site. Journal of Molecular Graphics and Modeling, Vol. 23, 1995, pp.465-475.
 A. Goodman Gilman, Pp. 814-829.
As a final note, due to the nature of this material the account presented here is necessarily simplified and incomplete in many respects; further inquiry can be well satisfied via A. Goodman Gilman, Lipman-Massie Clinical Electrocardiography, Goldfrank’s Toxicologic Emergencies, as well as other more detailed resources.
A 68yr long-term care inmate presented to nursing with an altered level of consciousness, chest pain, and bradycardia. Paramedic services were called to the scene for transport and found the nursing staff encouraging the pt to walk back and forth across the exam room to, “help bring his pulse up.” The following EKG was recorded. Note that voltage enhancement has been maximized in the rhythm strip to 2cm/mv, while the 12-lead is displayed with the standard gain of 10mm/mv.
As this is a third party case, little direct clinical or situational information is available to contextualize this EKG or the surrounding events. Objectively speaking, a markedly bradycardic junctional rhythm can be appreciated with retrograde conduction of p-waves, seen inverted, buried 160ms into the QRS complex. Net positive QRS deflections in I-III, avL and avF, and negative in avR indicate an axis in the lower left quadrant. Close examination reveals a 0.1mv electrical alternans, perhaps most evident in the limb leads, but also apparent (~0.05mv) in V5 and V6. Explicitly pathological features include subtle precordial T-wave inversions in V1-3 and conspicuous low voltage QRS amplitude in all leads.
Regarding this latter subject, numerous criteria have been suggested as to what constitutes abnormally low voltage; a consensus approach would consider either the sum of the QRS voltages in all 12 leads as necessarily less than 12mv, or a combined judgment requiring the average of QRS voltages in the limb leads as less than 5mm and that in the precordial leads less than 10mm.
The typical differential diagnosis associated with low voltage QRS includes etiologies of increased impedance (such as obesity, hyperinflative lung disease, and pericardial/pleural effusion), etiologies of infiltrative disease (such as hemochromatosis, amyloidosis, and neoplasm), and metabolic or toxicological causes (such as hypothyroidism and alcoholism). Low voltage has also been associated with both chronic and acute ischemic heart disease. An exhaustive review of the DDx can be found here.
While neither the clinical nor the electrocardiographic features of this case are sufficiently specific to seal any one diagnostic verdict, there are nonetheless some possibilities here worthy of note. Exogenous toxicological etiologies should be ruled out; hypotension with a slow junctional escape could be linked to digitalis, beta and calcium channel blockers, or other readily available pharmaceuticals. Of particular interest, the possibility of RCA associated ischemia must also be entertained. The pt’s clinical picture, low voltage QRS amplitude, and junctional bradycardia are strongly suggestive in this direction. Similar presentations with more explicit pathological substrates can be seen on this site in case nos. 4A- 4D, particularly the slow junctional STEMI of no. 4D.
Lastly, the subtle finding of electrical alternans forces a compelling consideration of pericardial effusion. Were the heart indeed spatially shifting within the pericardium from beat to beat, one would anticipate a greater shift of axis in the frontal, limb-lead plane than the transverse plane of the precordial leads, just as is present on this tracing. Alternating junctional foci or an artifact of physical positioning could produce a similar bigeminal effect, yet when this alternans is seen in the context of low voltage, the finding commands greater attention.
Paramedic services successfully temporized this pt’s status with atropine and supportive care until he reached the emergency department; there, after 20 minutes, he receded into semi-consciousness. No follow-up could be done.
A 52 yr old man, transfered from an outside hospital, pain free at this time and resting comfortably.
Included for completeness, this ECG demonstrates no AV block but is a distractor from the previous series. Individual P-waves cannot be identified; the rhythm is irregularly irregular. In the context of inferior wall MI, this pt is experiencing slow ventricular response a-fib with an associated digitalis-type ST-segment morphology.
Again, unfortunately, no follow up was possible regarding this pt’s outcome.
A 19yr old white male, s/p cardiac arrest, now with transtentorial herniation.
As with this case, ECG patterns in the context of acute CNS disease have been primarily associated with ventricular repolarization, i.e. the morphology of the QT segment and T-wave, and the presence of prominent U-waves (not present here). Although little sensitivity or specificity has been accorded to this connection, the phenomenon raises interesting questions of nerocardiac interrelation. Most explicitly, bradycardia as a result of hypervagotonia is often noted in the setting increased intracranial pressure. Yet more difficult to explain are the deep, symmetrical T-wave inversions and prolonged Q-T frequently described as more specific indicators of intracranial pathology. It has been hypothesized that these effects are due to an autonomicaly mediated catocholamine surge causing transient coronary vasospasm and subsequent myocardial ischemia.
The ST segments in this case are of a somewhat novel morphology, perhaps even reminiscent of the scooped out troughs seen as a common Digitalis effect.
This pt. was taken to the OR for withdrawal of ventilatory support and organ donation later in the night.