Pharmacokinetics, Bioavailability and Serum Levels of Cardiac Glycosides

Home , Deslanoside, Ouabain

View metadata, citation and similar papers at core.ac.uk brought to you by CORE JACCVol. 5, No.5 provided by Elsevier - Publisher43A Connector May 1985:43A-50A

THOMAS W. SMITH, MD, FACC Boston. Massachusetts

Digoxin, the cardiac glycoside most frequently used in bioavailability of digoxin is appreciably less than that of clinical practice in the United States, can be givenorally digitoxin, averaging about two-thirds to three-fourths of or intravenously and has an excretory half-life of 36 to the equivalent dose given intravenously in the case of 48 hours in patients with serum creatinine and blood currently available tablet formulations. Recent studies urea nitrogen values in the normal range. Sincethe drug have shown that gut ftora of about 10% of patients re is excreted predominantly by the kidney, the half-life is duce digoxin to a less bioactive dihydro derivative. This prolonged progressivelywithdiminishingrenal function, process is sensitiveto antibiotic administration, creating reaching about 5 days on average in patients who are the potential for important interactions among drugs. essentially anephric. Serum protein binding of digoxin Serum or plasma concentrations of digitalis glycosides is only about 20%, and differs markedly in this regard can be measured by radioimmunoassay methods that are from that of digitoxin, which is 97% bound by serum nowwidelyavailable, but knowledgeofserum levelsdoes albumin at usual therapeutic levels. Digitoxin is nearly not substitute for a sound working knowledge of the completely absorbed from the normal gastrointestinal clinical pharmacology of the preparation used and care tract and has a half-lifeaveraging 5 to 6 days in patients ful patient follow-up. receiving usual doses irrespective of renal function. The (J Am Coil Cordial 1985;S:43A-SOA)

Pharmacokinetics Digoxin Preparations ofcardiac glycosides used clinically are de Routes of administration. Digoxin is the form of dig rived from the leaves and seeds of plants from the genera italis most frequently used in the United States, both in Digitalis and Strophanthus. Other flora contain cardiac gly hospital and office practice. This is probably due in large cosides that may be of importance with regard to toxicity, part to its flexible routes of administration, its intermediate but are not used clinically. These include Convallaria ma duration of action and the availability of convenient methods jalis (lily of the valley) and Thevetia neriifolia (yellow for assaying serum digoxin concentrations. Although dig oleander) . oxin may be administered orally , intravenously or intra All of the cardiac glycosides consist of combinations of muscularly , the intramuscular route is not recommended an aglycone, or genin , with one to four attached sugar res because it is painful and produces localized muscle damage idues. The characteristic pharmacologic properties reside in reflected in increased serum creatine kinase levels . The clas the genin, while factors influencing pharmacokinetic prop sic studies of the pharmacokinetics of digoxin carried out erties , including water solubility, are influenced by the na by Doherty et al. (2) established the patterns observed after ture of the sugars (glycosides) attached to the genin. use of these three routes of administration in volunteers with Figure I outlines the derivation of the cardiac glycosides normal renal function (Fig. 2). As expected, the serum currently in clinical use in the United States. Leaves from concentration rises most rapidly after intravenous admin the Digitalis lanata plant contain precursor glycosides that istration , followed by oral and, still later, by intramuscular yield digitoxin and digoxin after minor chemical derivation administration . (1) . The seeds of Strophanthus gratus contain ouabain . Excretion. Excretion of digoxin occurs at a rate inde pendent of the route of administration. Digoxin is excreted From the Cardiovascular Division, Brigham and Women 's Hospital in an exponential manner, with an average half-life of 36 and the Departments of Medicine, Brigham and Women 's Hospital and hours in healthy individual s with normal renal function (3). Harvard Medical School, Boston, Massachusetts. Approximately 37% of total body stores of digoxin are ex Address for reprints: Thomas W. Smith , MD, Cardiovascular Division, Brigham and Women's Hospital, 75 Francis Street, Boston, Massachusetts creted daily under these circumstances. Older patients usu 02115. ally have a reduced glomerular filtration rate even if the

~Duocerrldi9iI0"id A

0_ purpurlo

DIGITALIS ( leal) Lanatoside A - --ACltrldIQitolin -IOIGITO Xl!il Figure 1, Derivation of clinically relevant digitalis .( 46%) preparations. D. lonola .• Lanatoside 8 ( 17%)

• Lanoloside C ~ACI'rld iQo'in --IDIGOXIN I (37%1 / <( Cedilonid I CUilonid 0 ( Dulonoside )

•I DiQilonid • composite 01 lonotosidee A, 8, C )

blood urea nitrogen and creatinine levels remain in the nor daily losses of digoxin through renal and other pathways mal range, and a half-life of 48 hours represents a more match the daily intake of the drug. reasonable estimate. Renal excretion of digoxin tends to be Absorptioh. A useful clinical principle is that when daily independent of the rate of urinary flow in patients with maintenance therapy is begun in patients not previously reasonably normal renal function (4,5). There is, however, receivingdigoxin, steady state plateau concentrations exist some dependence of renal tubular reabsorption of digoxin after four to five half-livesor about 7 days in patients with on the urinary flow rate, and hence digoxin reabsorption normal renal function (9). In the presence of renal impair ruay increase at very low rates of urinary flow and con mentresulting in prolongation of the half-lifeof elimination, versely may be decreased by acute vasodilator therapy in the time required to reach steady state on a daily mainte patients with congestive heart failure (6,7). Digoxinexcre nance regimen is correspondingly prolonged. When rapid tion by way of the kidney is predominantly in an unchanged onset of digoxin effect is needed, a loading dose is admin to.rn, but a subset of about 10% of patients excrete appre istered. followed by daily maintenance therapy. In obese ciablequantitiesof relativelyinactivemetabolites (8). Equi subjects, digoxin pharmacokinetics change little after the librium or steady state conditions occur at the point when loss of large amounts of adipose tissue, indicating that lean body mass should be used when dosages are calculated (10,11). Figure 2. Comparison of pharmacokinetics of digoxin given by the intravenous (IV), oral and intramuscular (1M) routes. As ex Nomograms for calculationof the total oral loadingdose pected, serum concentration rises most rapidly by the intravenous anddaily maintenance doseof digoxinbasedon bodyweight form of administration. T'I2 = half-life of elimination. (Repro and renal function are available (12). Such nomograms are duced from Doherty et al. [2] withpermission of the authors and helpful for the initial estimation of digoxin dosage, but do Progress in Cardiovascular Diseases.) not change the need for careful follow-up by the clinician.

1 AT Ea CR[ToG \0 "". Gastrointestinal absorption ofdigitalis glycosides is pas ' 00 0 ~( R "T, _ , tOOl sive . The rate and completeness of absorption are greatest ,n ... ,s IV 92 - with the least polar molecules and decrease with increasing cardiac glycoside polarity (13). Patients with normal gas '8"' " -,". • 63"- trointestinal function shouldabsorbdigoxinadministered by l 00 J .a Pu ," o A. 62' the oral routeat about85%of the administered dose in elixir ~ • ., form and 60 to 75% in tablet formulations that meet current <.." Food and Drug Administration and United States Pharma '" ~ copeia guidelines (14) (see subsequent comments on

... ' 0 bioavailability). -.... _-- - - -::-: '"' -..... -- ... Protein binding of digoxin averages 20 to 25% under usualcircumstances (15,16). Only the free fraction is phar macologically active. In the caseof digoxin, proteinbinding is a minor factor in determining the pharmacokinetics be 0 ' cause glomerularfiltration of the drug should be about 80% o ,'2 of the serum concentration. JACC Vol. 5. No.5 SMITH 45A May 1985:43A-50A CARDIACGLYCOSIDES

Time course of digoxin distribution in the body. The binds avidly to serum proteins. This results in very limited time course is well characterized by a two compartment renal excretion, and the half-life of elimination of digitoxin model (2). Distribution from the central (vascular) com accordingly varies littie from the average of 4 to 6 days, partment to peripheral sites occurs with a half-time of about irrespective of renal status (15). Because of the longer half 30 minutes. Although the concentration of digoxin in skel life, initiation of daily maintenance doses of digitoxin will etal muscle is relatively low compared with that in the kid not produce steady state levels until about 3 to 4 weeks. ney or heart, the total digoxin content of skeletal muscle The bioavailability of digitoxin when administered orally represents the largest single pool in the body because of the approaches 100% (25). large mass of this component of body composition. As stated, digitoxin binds avidly to human serum albumin Available studies (17) demonstrate a fairly constant ratio such that about 97% of the drug circulates in a bound state between the digoxin bound to heart muscle and the serum under usual clinical circumstances. A number of drugs are concentration. It is of importance, however, that electrolyte reported (26) to displace digitoxin from serum albumin bind derangements can exert clinically significant effects on myo ing sites, but it is not clear if these interactions are of clinical cardial uptake and distribution of cardiac glycosides. Hy importance. It may be that severe depletion of serum al perkalemia and hyponatremia reduce digoxin binding to the bumin (for example, in nephrotic syndrome) may effectively myocardium (4,18). Important interactions among drugs af increase the effect of digitoxin at any given total serum fect the distribution of digoxin in the body, as discussed concentration of the drug. elsewhere in this symposium (19). Mode of elimination. The major route of elimination Absorption and excretion in infants. Infants and chil of digitoxin is metabolism in the liver. Approximately 8% dren usually absorb and excrete digoxin in a manner similar of the body digitoxin store undergoes biotransformation to to that of adults. Nevertheless, digoxin doses in infants metabolites, some of which are cardioactive and some of characteristically are larger than those conventionally used which are not (27,28). Enterohepatic cycling of digitoxin in adults when calculated on the basis of mg/kg of body averages approximately 6 to 7%/day (29). This cycle can weight or per unit of body surface area. These higher doses be interrupted at least in part by nonabsorbable resins, such tend to produce higher serum digoxin concentrations, which as cholestyramine, that bind digitoxin in the gut (30). This are usually well tolerated (20-22). Digoxin crosses the pla tends to enhance the clearance rate, although it is not Clear cental barrier, as documented by the observation (20) that that the effect is sufficient to be decisive in the management fetal umbilical cord digoxin concentrations are comparable of overt digitoxin toxicity. with those found in the venous blood of the mother at term. Metabolism of digoxin. Although renal excretion of di Other Cardiac Glycosides goxin is the dominant route of elimination, digoxin is me Deslanoside (Cedilanid-D), This glycoside is structur tabolized to some degree to reduced compounds such as di ally similar to digoxin, but has an additional glucose residue hydrodigoxin and dihydrodigoxigenin (2). In certain attached to the terminal sugar. This results in markedly individuals (about 10% of most patient groups), such me decreased gastrointestinal absorption, and deslanoside is used tabolism accounts for up to 30 to 40% of the total urinary very little in the United States at the present time. Its ex excretion of digoxin and its metabolites (8,23). The work cretion kinetics are very similar to those of digoxin (2). of Lindenbaum et al. (24) established that the metabolism Ouabain. This is the most polar of the clinically avail of digoxin to reduce products varies both with the bio able cardiac glycosides, and is only used intravenously for availability of the drug and with the nature of the bacterial rapid digitalization. It is excreted by exponential or first flora in the gastrointestinal tract. Consequently, antibiotic order kinetics, predominantly through the kidney. Patients therapy that changes the flora reduces the metabolism of with normal renal function have a serum ouabain half-life digoxin to inactive products and can produce an apparent averaging 21 hours. As with digoxin, impaired renal func increase in the effective amount of drug in the body when tion will slow the excretion of ouabain (31). the intake of digoxin remains constant. Particular care is Information on the pharmacokinetics of other glycosides needed in the surveillance of patients receiving digoxin when not commonly used in the United States is discussed in antibiotics are added to the regimen. Data regarding the appropriate review articles (17). pharmacokinetics of digoxin and other clinically relevant glycosides are summarized in Table I. Bioavailability Digitoxin Factors affecting digoxin absorption. Observations in Half-life of elimination. Although it is the second most the 1970s led to the conclusion that some patients had low frequently prescribed cardiac glycoside in the United States, serum digoxin concentrations and responded poorly to the digitoxin is far less commonly used than digoxin. It is less drug despite relatively large oral doses (32). This prompted polar than any other glycoside in current use and, therefore, the investigation of the bioavailability of several digoxin 46A SMITH lACC Vol. 5, No.5 CARDIAC GLYCOSIDES May 1985:43A-50A

Table 1. Cardiac Glycoside Preparations

Principal Average Digitalizing Metabolic Usual Daily Dose (mg) Gastroin- Onset of Peak Route Oral Mainte testinal Action* Effect Average (excretory Intra- nance Dose Agent Absorption (min) (h) Half-Lifet pathway) Oralj venous§ (rngj]

Ouabain Unreliable 5 to 10 1/2 to 2 21 hours Renal; some 0.30 to gastrointestinal 0.50 excretion Deslanoside Unreliable 10 to 30 I to 2 33 hours Renal 0.80 Digoxin 55 to 75%** 15 to 30 IV2 to 5 36 to 48 Renal; some 1.25 to 0.75 to 0.25 to (Lanoxicaps hours gastrointestinal 1.50 1.00 0.50** 90 to 100%) excretion Digitoxin 90 to 100% 25 to 120 4 to 12 4 to 6 Hepatictt: renal 0.70 to 1.00 0.10 days excretion of 1.20 metabolites Digitalis About 40% 4 to 6 Similar to 0.80 to 0.10 g leaf days digitoxin 1.20 g Lanatoside C 10 to 40% Similar to Renal 10 0.5 to digoxin 1.5 Gitalintf 4 to 6 Similar to 6 0.25 to days digitoxin 1.25 Acetyl About 70% 20 to 30 8 to 10 Similar to Similar to 2.0 to 1.4 to 0.1 to digitoxin digitoxin digitoxin 3.0 1.6 0.2

* For intravenous dose; t for normal subjects (prolonged by renal impairment with digoxin, ouabain and deslanoside and probably by severe hepatic disease with digitoxin and digitalis leaf); t divided doses over 12 to 24 hours at intervals of 6 to 8 hours; * given in increments for initial subcomplete digitalization, to be supplemented by further small increments as necessary; II average for adult patients without renal or hepatic impairment; varies widely among individual patients and requires close medical supervision; ** for tablet form of administration (may be less in malabsorption syndromes and in formulations with poor bioavailability). A recently marketed preparation (Lanoxicaps, Burroughs Wellcome Co.) contains digoxin solution in capsules and has a bioavailability of 90 to 100%. If such a preparation is used, the daily oral maintenance should be reduced by about 20% (for example, Lanoxin tablets of 0.25 mg and 0.125 mg strengths are approximately equivalent to Lanoxicaps of 0.2 and 0.1 mg strengths, respectively); tt enterohepatic cycle exists; H Gitalin is a mixture of cardiac glycosides, the principal one of which is digitoxin; *§ approximately 20% lower maintenance doses are required if gel solution in capsules (Lanoxicaps) is used. (Modified from Smith TW. Drug therapy: digitalis glycosides. N Engl J Med 1973;288:719.)

preparations. Although various pharmaceutical formulations margin between therapeutic and toxic digoxin doses and may contain chemically equivalent amounts of digoxin, studies serum levels. Current FDA and USP bioavailability guide (32) have shown a wide range of dissolution rates of avail lines include defined permissible ranges of tablet dissolution able marketed digoxin tablets from several suppliers. This, rates, and require the demonstration in human subjects that together with individual patient variation, can result in sub such dissolution rate data translate into clinically reliable stantial variability in digoxin bioavailability. Malabsorption bioavailability (37). Current FDA guidelines require that all syndromes may also produce poor and erratic digoxin ab digoxin tablets marketed in the United States must have an sorption (33), although maldigestion caused by pancreatic in vitro dissolution rate of at least 65% in I hour, but no insufficiency usually results in normal absorption of the greater than 90% in 15 minutes. Digoxin elixir is more drug. Digoxin ingestion along with or shortly after meals bioavailable (70 to 85% of the intravenous dose) than the tends to decrease the peak serum concentration, but total usual tablet form (60 to 80% of the intravenous dose) (38,39). absorption is not appreciably reduced in patients with normal A recently marketed digoxin gel solution in capsules (Lan gastrointestinal function (34). Digoxin absorption tends to oxicaps, Burroughs-Wellcome Co.) has enhanced bioavail be enhanced by drugs that reduce the motility of the gas ability (90 to 100%) compared with tablets, elixir or intra trointestinal tract, and conversely can be decreased by drugs muscular preparations. The clinician should be aware that that increase motility (35). Certain nonabsorbable sub early high serum digoxin concentrations occur after inges stances, including cholestyramine, colestipol, kaolin-pectin tion of the encapsulated gel formulation. Because of the and nonabsorbable antacids, can interfere with gastrointes greater bioavailability, Lanoxicaps of 0.1 and 0.2 mg strength tinal absorption of digoxin when taken concurrently (17). are approximately equivalent to tablets with a strength of The antibiotic neomycin has also been shown to interfere 0.125 and 0.25 mg, respectively. with digoxin absorption (36). As emphasized elsewhere in this symposium (19), cli Clinical implications. Bioavailability considerations have nicians must be alert to alterations in the total medication important clinical implications given the relatively narrow program of the patient because addition or discontinuation lACC Vol. 5, No.5 SMITH 47A May 1985:43A-50A CARDIAC GLYCOSIDES

of a variety of agents, particularly nonabsorbable resins, of Klein et al. (45), in which serum digoxin levels were antacids or antibiotics, can alter digoxin bioavailability or correlated with acetylstrophanthidin tolerance, demonstrated the degree of conversion of digoxin to inactive metabolites, that severe pulmonary, coronary and aortic valve disease, or both. as well as old age, predisposed patients to cardiac glycoside sensitivity. Other variables influencing individual patient Serum Digoxin and Digitoxin Concentrations: sensitivity to cardiac glycosides are discussed in detail else where in this symposium (46). Use and Misuse Clinical correlations of serum digitalis levels. The problem of the narrow margin between therapeutic Approximately 50 studies correlating serum digitalis levels and toxic doses of digitalis has resulted in the development with clinical state have been published and have been re of various methods to determine serum glycoside levels. viewed recently (17). These are summarized briefly in Ta Serum concentration measurements do have clinical value, bles 2 and 3. The findings from these studies, based on data but inappropriate use or interpretation of such data limits from more than 1,000 patients, indicate that the serum di their usefulness and can, on occasion, lead to suboptimal goxin level in patients receiving therapeutically appropriate management decisions. doses averages about 1.4 ng/ml (1.8 nmol/liter), while mean Assay methods. Assay methods currently available have levels in patients with drug toxicity, as defined by the pres been reviewed elsewhere (40). Each method has its specific ence of typical arrhythmias, are generally higher by a factor advantages and disadvantages. The radioimmunoassay tech of 2 to 3. Although this difference is statistically significant nique (41) is now used in most clinical laboratories . Digoxin in the vast majority of studies, overlap clearly does exist and digitoxin concentrations in serum can be determined between serum digoxin levels in groups of patients with and separately given appropriate attention to technical details. without toxicity. Such overlap is generally more pronounced However, radioimmunoassay techniques in general have a in prospective blind study protocols than in unblinded re variety of pitfalls, and the clinical usefulness of serum con trospective studies (47). centrations is limited by the accuracy of the reported values. Extensive data are also available for patients receiving Proper selection of the antibody population, accuracy of digitoxin (Table 3). Digitoxin levels in serum or plasma are standards, purity of tracer glycosides and care in the use of higher than those of digoxin by a factor of about 10 because counting methods are all requisite to obtaining satisfactory of substantial serum protein binding of digitoxin as just results. The presence in serum of diagnostic radioisotopes discussed. The results of studies of clinical correlations are used in nuclear medicine scanning techniques can be mis otherwise analogous to those for digoxin in that mean levels leading. In addition, the clinician must avoid drawing blood observed for groups of patients considered to be optimally samples for cardiac glycoside analysis shortly after doses digitalized are significantly less than values for patients with so that full equilibration of the drug between the plasma overt toxicity. Again, substantial overlap exists between and the heart and other body compartments is complete. groups with and without toxicity. Sampling of serum should generally be done at least 5 to 6 In terms of the management of individual patients, I hours after either oral or intravenous doses. would emphasize that no specific serum concentration exists Rationale for serum digitalis measurements. Several that can be used to define a clear boundary between the lines of evidence bear on the relation between serum gly presence and absence of toxicity. A particular problem in coside levels and the corresponding pharmacologic effect. this regard is the definition of digitalis toxicity because Both therapeutic and toxic effects of cardiac glycosides are abnormalities of cardiac impulse formation or conduction dose-related. All studies demonstrate that serum glycoside resulting from digitalis excess also occur as a consequence levels increase with increasing dosage, and a statistical cor of intrinsic heart disease, even in patients without prior relation between serum level and clinical state would, there digitalis exposure. Consequently, serum digitalis levels should fore, be anticipated. Experimental and clinical studies (42,43) be considered together with all of the additional clinical data demonstrate a relatively constant ratio of serum to myo available before a clinical management decision is reached. cardial digoxin concentrations after full equilibration be Optimal serum glycoside levels. It is no easy matter to tween vascular and peripheral compartments. Since the car establish optimal serum cardiac glycoside levels, even in diac glycoside binding site of the receptor Na-K adenosine patients with no evidence of toxicity. Chamberlain et al. triphosphatase (ATPase) faces the outer cell surface, a basis (48) observed a correlation, within broad limits, between exists for the translation of serum level to effect on the serum digoxin concentrations and the ventricular rate in the myocardial cell. Perhaps most importantly, experimental presence of atrial fibrillation, at least in patients who had studies (44) have shown a predictable relation between serum relatively rapid ventricular rates before administration of digoxin concentration and cardiac electrophysiologic effects. digitalis. It is commonly found that "therapeutic" digoxin Although true in a general sense, these considerations levels are inadequate to control the ventricular rate in acutely leave room for substantial variability in the response of the ill patients with problems such as hypoxia, infection or individual patient to a given serum digitalis level. The study recent surgery. In the study by Goldman et al. (49), serum 48A SMITH JACC Vol. 5, No.5 CARDIAC GLYCOSIDES May 1985:43A-50A

Table 2. Serumor PlasmaDigoxin Concentrations in Patients Table 3. Serumor PlasmaDigitoxin Concentrations in Patients With and Without Toxicity* With and Without Toxicity*

Mean Digoxin Mean Digitoxin Concentration (ng/ml) Concentration (ng/ml) Patients Patients Patients Patients Source Without Toxicity With Toxicity Source Without Toxicity With Toxicity Beller et al. t 20.0 34.0 Aronson et al. 1.60 2.60 Bentley et al.f 23.0 39.0 Beller et al. 1.00 2.30 Brooker and Jelliffe' 31.8 48.8 Bernabei et al. 1.()O 2.90 Chiche et al.t 25.4 57.0 Bertler and Redfors] 0.90 2.40 Dessainrt 26.8 96.0 Bertler et al. t 1.40 3.10 Hillestad et aLII .. 16.8 28.3 Brooker and Jelliffet 1.40 3.10 Lukas and Peterson 20.0 43.0 to 67.0 Burnett and Conklin 1.20 5.70 (range) Ca~thers et aI.' 1.21 2.76 Morrison and Killipt 25.0 53.0 Chamberlain et al. 1.40 3.10 (0.1 mg/day) Doering et al. 1.02 3.07 Peters et al. t 28.8 56.4 Evered and Chapman 1.38 3.36 Rasmussen et aLII 16.6 48.7 Fogelman et aLII 1.40 1.70 Ritzmann et aLII 20.5tt 37.0tt Follath et al. 1.20 3.20 Smithj 17.0 34.0 Grahame-Smith and Everestt 2.40 5.70 *Full references to the papers cited are found in reference 17; tra Hayes et al. dioimmunoassay; tadenosim; triphosphatase inhibition; 'enzymatic dis Infants 2.80 4.40 placement; Ilrubidium-86 uptake; **double isotope dilution derivative; Children 1.30 3.40 ttmedian concentration. Hoeschen and Proveda 0.80 to 1.30 2.80 Howard et aLII 0.97 0.91 Huffman et al. 1.49 3.32 Iisalo et al. 1.20 3.10 Johnston et al. 1.00 3.15 digoxin levels of +.5 ng/ml or greater were required to Krasula et al. control the ventricular response in 15 of 39 patients. The Infants 1,70 3.60 current availability of beta-adrenergic blocking agents and Children 1.10 2.90 verapamil usually obviates the need for the use of unusually Lader et al. 1.10 2.20 large doses of digitalis to control the ventricular response Lehmann et al. Normokalemic patients NS** 2.73 in the presence of supraventricular tachyarrhythmias. Hypokalemic patients NS** 1.76 A particularly difficult question relates to the definition Lichey et al. 1.20 2.50 of optimal serum glycoside concentrations ill patients with Loes et al. 1.10 4.90 congestive heart failure and normal sinus rhythm. Although McCredie et al. the available data are somewhat limited, recent studies (17) Infants 3.45 Children 1.41 3.81 suggest that the favorable hemodynamic effects of digoxin Morrison et aUt 0.76 3.35 may tend ~o plateau at serum levels in the range of 1.0 to Oliver et al. 1.60 3.00 1.5 ng/rnl. It remains to be shown whether higher, serum Park et al. 1.10 3.80 levels will produce further enhancement of the contractile Ritzmann et al. t 1.20 5.50H state. Thus, it may well be that the optimal risk-benefit ratio Shapiro Normokalemic patients NS** 3.68 is achieved at serum levels in the 1.0 to 1.5 range. There Hypokalemic patients NS** 1.13 is certainly no doubt that higher levels expose the patient Shermann and Bourdon 1.37 4.58 to an increased risk of cardiac glycoside-induced rhythm Singh et al. 2.91 4.79 disturbances. Smith et al. 1.30 3.30 Cost-effectiveness. The cost-effectiveclinical use of serum Smith and Haber 1.40 3.70 Suzuki and Ogawa 1.20 3.20 glycoside measurements, as with all laboratory tests, re Waldorff and Buch 1.00 2.30 quires considerable sophistication on the part of the clini Weissel et al. 1.38 2.97 cian. Evaluation of timing and magnitude of prior digoxin Whiting et al. 1.40 3.50 doses, renal function and body mass allows a first approx 4.40 Zeegers et al. 1.60 imation of the body stores of the drug. When patients re *Full references to the papers cited are found in reference 17; radioim ceiving maintenance digitalis develop symptoms such as munoassay used in all instances except as noted: trubidium-86 uptake; fatigue, visual changes, anorexia, nausea, vomiting or car tenzymatic displacement; 'adenosine triphosphatase inhibition; 'differ diac arrhythmias, digitalis intoxication should be suspected ences in mean concentration were statistically significant (p < 0.05) in all series except as noted; **not stated; ttstatistical significance not stated; and serum level data are likely to be of use. Serum digoxin Hmedian concentration. levels within the range usually considered as "therapeutic" JACCVol. 5. No.5 SMITH 49A May 1985:43A-50A CARDIACGLYCOSIDES

should not be considered to excludetoxicity in view of the 8. Peters U, Falk LC. Kalman SM. Digoxin metabolism in patients. Arch marked variability in individual patient responses. The na Intern Med 1978;138:1074-6. ture and severity of underlying heartdisease are particularly 9. Marcus FL, Burkhalter L. Cuccia C, et al. Administration of tritiated digoxin with and without a loading dose: a metabolic study. Circulation important variables. 1966;34:1165-74. Failure to achieve adequate therapeutic response. 10. Ewy GA, Groves BM, Ball MF. et al. Digoxin metabolism in obesity. Another common clinical problem is the failure to achieve Circulation 1971;44:810-4. an adequate therapeutic response in a patient receiving con II. Abernethy DR, Greenblatt DJ. Smith TW. Digoxin disposition in ventional digoxin doses. It must be determined whether the obesity: clinical pharmacokinetic investigation. Am Heart J 1981;102:740-4. dose is inadequate (for example, as a result of impaired 12. Jelliffe RW. Brooker G. A nomogram for digoxin therapy. Am J Med gastrointestinal absorption or noncompliance with a pre 1974;57:63-8. scribed regimen) or whether there are reasons why the pa 13. Haass A, Lullmann H. Peters T. Absorption rates of some cardiac tient may be relatively resistant to usual digoxin doses and glycosides and portal blood flow. Eur J Pharmacol 1972;19:366-70. serum levels (for example, occult thyrotoxicosis or mitral 14. lisaJo E. The clinical pharmacokinetics of digoxin. Clin Pharmacol stenosis). If the patient is taking the prescribed dose and 1977;2:1-16. the serumdigoxin level remains low, this information may 15. Lukas DS. De Martino AG. Binding of digitoxin and some related provide the clue to other clinical disorders or drug inter cardenolides to human plasma proteins. J Clin Invest 1%9;48:1041-53. actions. Hyperthyroidism tends to produce relatively low 16. Storstein L. Studies on digitalis. V. The influence of impaired renal function, hemodialysis, and drug interaction on serum protein binding serum digoxin concentrations in addition to true resistance of digitoxin and digoxin. Clin Pharmacol Ther 1976;20:6-13. to control of the ventricular response to atrial fibrillation 17. Smith TW. Antman EA, Friedman PL, Blatt CM, Marsh 10. Digitalis typical of this condition. Malabsorption syndromes andpoorly glycosides: mechanisms and manifestations of toxicity. Prog Car bioavailable digoxin preparations will result in low serum diovasc Dis 1984;26:413-41; 26:495-523; 27:21-56. digoxin concentrations and clinical signsof inadequate dig 18. Goldman RH, Deutscher RN, Schweizer E, et al. Effect of a phar macologic dose of digoxin on inotropy in hyper- and normokalemic italization. Drug interactions, as discussed elsewhere (19), dogs. Am J Physiol 1972;223:1438-43. can produce substantial variations in serum levels. 19. Marcus FI. Pharmacokinetic interactions between digoxin and other Lastly, skepticism is appropriate when laboratory results drugs. J Am Coil Cardiol 1985;5(suppl):82A-90A. conflict with clinical judgment. Underno circumstances can 20. Rogers Me. Willerson JT. Goldblatt A, et al. Serum digoxin con serum digitalis levels replace sound clinical judgment. In centrations in the human fetus. neonate and infant. N Engl J Med dividual laboratory values should never be used as the sole 1972;287:1010-3. 21. Krasula RW. Pellegrino PA, Hastreiter AR, et al. Serum levels of basis for determining the presence of drug toxicity or effi digoxin in infants and children. J Pediatr 1972;1l1:566--9. cacy. I do not advocate the periodic routine measurement 22. lisalo E, Dahl M. Sundqvist H. Serum digoxin in adults and children. of serumdigoxin concentrations in patients with a satisfac Int J Clin Pharmacol 1973;7:219-22 tory therapeutic response to a properly chosen dosage reg 23. Luchi RJ. Gruber JW. Unusually large digitalis requirements: a study imen. The best assurance of an optimal risk-benefit ratio of altered digoxin metabolism. Am J Med 1968;45:322-8. will always be the careful follow-up of the patient by a 24. Lindenbaum J, Rund DO, Butler VP, et al. Inactivation of digoxin vigilant clinician. by the gut flora: reversal by antibiotic therapy. N Engl J Med 1981;305:789-94. 25. Gold H, Cattell M. Modell W, et al. Clinical studies on digitoxin (Digitaline nativelle): with further observations on its use in the single average full dose method of digitalization. J Pharmacol Exp Ther References 1944;82:187-95. 26. Solomon HM. Abrams WB. Interactions between digitoxin and other I. Moe GK, Farah AE Digitalis and allied cardiac glycosides. In: Good drugs in man. Am Heart J 1972;83:277-80. man LS, Gilman A, eds. The Pharmacological Basis of Therapeutics. New York: Macmillan, 1970:677-708. 27. Pettier D, Mayerson M. Marcus FI. Clinical pharmacokinetics of digitoxin. Clin Pharmacol 1977;2:292-311. 2. Doherty JE deSoyza N. Kane JI. et al. Clinical pharmacokinetics of digitalis glycosides. Prog Cardiovasc Dis 1978;21:141-58. 28. Lukas DS. The pharmacokinetics and metabolism of digitoxin in man. In: Storstein 0, ed. Symposium on Digitalis. Oslo: Glydendal Norsk 3. Smith TW. Drug therapy: digitalis glycosides. N Engl J Med Forlag, 1973:84-102. 1973;288:719-22. 29. Storstein L. Studies on digitalis. III. Biliary excretion and enterohe 4. Marcus Fl. Metabolic factors determining digitalis dosage in man. In: patic circulation of digitoxin and its cardioactive metabolites. Clin Marks BH, Weissler AM, eds. Basic and Clinical Pharmacology of Pharmacol Ther 1975;17:313-20. Digitalis. Springfield. IL: Charles C Thomas, 1972:243-59. 30. Caldwell JH, Bush CA. Greenberger NJ. Interruption of the entero 5. Bissett JK, Doherty JE. Flannigan WJ. et al. Tritiated digoxin. XIX. hepatic circulation of digitoxin by cholestyramine. II. Effect on met Turnover studies in diabetes insipidus. Am J Cardio11973;31 :127-30. abolic disposition of tritium-labeled digitoxin and cardiac systolic time 6. Steiness E. Renal tubular secretion of digoxin. Circulation intervals in man J Clin Invest 1971;50:2638-44. 1974;50:103-7. 31. Selden R, Smith TW. Ouabain pharmacokinetics in dog and man: 7. Cogan JJ, Humphreys MH, Carlson CJ. et al. Acute vasodilator ther determination by radioimmunoassay. Circulation 1972;45:1176--82. apy increases renal clearance of digoxin in patients with congestive 32. Greenblatt DJ. Smith TW, Koch-Weser J. Bioavailability of drugs: heart failure. Circulation 1981;64:973-6. the digoxin dilemma. Clin Pharmacokinet 1976;1:36--51. 50A SMITH lACC Vol. 5, No.5 CARDIAC GLYCOSIDES May 1985:43A-50A

33. Heizer WD, Smith TW, Goldfinger SE. Absorption of digoxin in 42. Hartel G, Lyllonen K, Merikallio E, et aI. Human serum and myo patients with malabsorptionsyndromes. N Engl J Moo 1971;285:257-9. cardium digoxin. Clin Pharmacol Ther 1976;19:153-7. 34. White RJ, Chamberlain DA, Howard M, et aI. Plasma concentrations 43. Jogestrand T. Digoxin concentration in right atrial myocardium, skel of digoxin after oral administration in the fasting and post-prandial etal muscle and serum in man: influence of atrial rhythm. Eur J Clin state. Br Med J 1971;1:380-2. Pharmacol 1980;17:243-50. 35. Manninen V, Apajalahti A, Melin J, et aI. Altered absorption of 44. Barr I, Smith TW, Klein MD, Hagemeijer F, Lown B. Correlation digoxin in patients given propantheline and metoclopramide. Lancet of the electrophysiologic action of digoxin with serum digoxin con 1973;1:398-400. centration. J Pharmacol Exp Ther 1972;180:710-22. 36. Lindenbaum J, Maulitz RM, Saha JR, et aI. Impairment of digoxin 45. Klein MD, Lown B, Barr I, et aI. Comparison of serum digoxin level absorption by neomycin (abstr). Clin Res 1970;20:410. measurement with acetyl strophanthidin tolerance testing. Circulation 37. Harter JG, Skelly JP, Steers AW. Digoxin-the regulatory viewpoint 1974;49:1053-62. (editorial). Circulation 1974;49:395. 46. Surawicz B. Factors affecting tolerance to digitalis. J Am Coli Cardiol 38. Greenblatt DJ, Duhme DW, Koch-Weser J, et aI. Bioavailability of 1985;5(suppl):69A-8IA. digoxin: what the clinician needs to know. Conn Med 1974;38:220-32. 47. Beller GA, Smith TW, Abelmann WH, et aI. Digitalis intoxication: 39. Huffman DH, Azarnoff DL. Absorption of orally given digoxin prep prospective clinical study with serum level correlations. N Engl J Med arations. JAMA 1972;222:957-60. 1971;284:989-97. 40. Smith TW, Curfman GD. Radioimmunoassay of cardiac glycosides. In: Strauss W, Pitt B, eds. Cardiovascular Nuclear Medicine. 2nd ed. 48. Chamberlain DA, White RJ, Howard MR, et aI. Plasma digoxin con St. Louis: CV Mosby, 1979:394. centrations in patients with atrial fibrillation. Br Med J 1970;3:429-32. 41. Smith TW, Butler BP Jr, Haber E. Determination of therapeutic and 49. Goldman S, Probst P, Selzer A, et aI. Inefficacy of "therapeutic" toxic serum digoxin concentrations by radioimmunoassay. N Engl J serum levels of digoxin in controlling the ventricular rate in atrial Med 1969;281:1212-6. fibrillation. Am J Cardiol 1975;35:651-6.