American Association for Clinical Chemistry
Better health through laboratory medicine
April 2008 Clinical Laboratory News: Therapeutic Drug Monitoring

CLN Banner Logo

April 2008: Volume 34, Number 4


Therapeutic Drug Monitoring
Recognizing the Sources of Interferences in Immunoassays
Amitava Dasgupta, PhD, DABCC

A number of effective but highly toxic drugs exhibit a narrow therapeutic index and marked interpatient pharmacokinetic variability. Individualized therapy with such drugs requires therapeutic drug monitoring (TDM) to safely obtain the desired clinical effects. Today, labs rely chiefly on competition immunoassays to establish the relationship between plasma drug concentration and therapeutic response and/or toxicity. This immunoassay format depends on single, analyte-specific antibodies and is well suited to analytes with small molecular weights, such as therapeutic drugs.

But antibody specificity is a major concern for TDM competition immunoassays. Many drug metabolites have structures similar to the parent drug and can interfere with the immunoassay results, falsely elevating or lowering the true value of the drug concentration. In addition, other molecules unrelated to the drug may have a comparable recognition motif. Both these types of molecules are referred to as “cross reactants” and may originate endogenously or exogenously.

Usually cross-reactants have no effect on more specific analytical techniques for measuring drug concentrations. For example, HPLC coupled with UV detection or mass spectrometric detection are unaffected by cross-reactants. In addition, a few relatively volatile drugs, such as benzodiazepines and pentobarbital, can also be analyzed using gas chromatography/mass spectrometry (GC/MS), another very specific analytical technique that is unaffected by cross-reactants. High concentrations of bilirubin and triglycerides, as well as the presence of heterophilic antibodies in the specimen, also can interfere with TDM immunoassay results but have no effect on HPLC or GC/MS analysis.

This article describes some of the well-known interferences in TDM immunoassays and how they affect assay results. Laboratorians need to be aware of the sources of these interferences in order to provide clinicians with accurate drug levels for monitoring therapy.

Interferences in Digoxin Immunoassays

Digoxin is a cardiac glycoside that is commonly used for the treatment of congestive heart failure and cardiac arrhythmia. Digitoxin, which is structurally similar, has the same effects as digoxin. Unlike digoxin, which is eliminated from the body via the kidneys, digitoxin is eliminated via the liver, so it is sometimes given to patients with poor or erratic kidney function.

TDM for digoxin was introduced more than 30 years ago and resulted in a marked reduction in the incidence of digoxin toxicity. Despite the longevity of this assay, false-positive and false-negative interferences are common and arise from a number of different sources.

Digoxin-like immunoreactive substances.

Researchers have hypothesized that an endogenous equivalent of a cardiac glycoside is responsible for interference in digoxin immunoassays. In TDM assays, the anti-digoxin antibody may detect the presence of these substances, which have been labeled digoxin-like immunoreactive substances (DLIS). Patients with certain conditions exhibit a greater degree of interference from these in TDM assays ( 1). DLIS have been found in various human body fluids including cord blood, amniotic fluid, bile, cerebrospinal fluid and saliva, as well as placenta and meconium.

Table 1
DLIS Interference in Digoxin TDM*

Uremia

Transplant recipients

Liver disease

Premature babies/newborns

Essential hypertension

Cardiomyopathy

Volume expansion/hypertension

Intensive care unit patients

Congestive heart failure

Diabetics

*Types of patients that exhibit above average interference from DLIS.

Because digoxin has a narrow therapeutic range (0.8–1.9 ng/mL), DLIS interference can be particularly troublesome. Elevated concentrations of DLIS in serum can falsely increase or decrease serum digoxin concentrations. Normally, the concentration of free DLIS is very small and undetectable and boiling plasma releases it from protein. In general, fluorescence polarization immunoassays (FPIA) that use polyclonal antibodies to digitoxin have greater cross-reactivity with DLIS, compared to monoclonal antibody-based immunoassays. DLIS binds strongly to serum proteins, but is virtually absent in the protein-free ultrafiltrate. Therefore, labs may want to consider monitoring the free or unbound digoxin concentration to eliminate DLIS interference.

Antidigoxin Fab Fragments

Another source of interference in digoxin immunoassays comes from treatment of potentially life-threatening digoxin intoxication. Clinicians administer the Fab fragment of the antidigoxin antibody, commercially available as Digibind (GlaxoSmithKline) and DigiFab (Protherics, Inc.), to patients who experience such complications. The Fab binds free digoxin in serum and effectively removes pharmacologically active, free drug.

Not surprisingly, these Fab preparations interfere with serum digoxin immunoassay measurements to some degree, depending on the assay design and the specificity of the antibody used. Microparticle enzyme immunoassay (MEIA) formats produce values that are higher than free digoxin concentrations in the presence of the Fab fragment.

In one study, researchers investigated the effect of Digibind and DigiFab on 13 different digoxin immunoassays. Positive interference in the presence and absence of digoxin was observed with Digibind and DigiFab, although the magnitude of interference was somewhat less with DigiFab (1). Overall, the magnitude of interference varied significantly for all the immunoassay methods tested. Because the molecular weights of these antibody fragments—46 kDa for DigiFab and 46.2 kDa for Digibind—are much higher than the cutoff of the filters used to prepare protein-free ultrafiltrates, labs should filter specimens from patients who have received these preparations before TDM analysis.

Spironolactone, Canrenone, and Ootassium Canrenoate

Spironolactone, a competitive aldosterone antagonist, has been used for many years in both hypertension and congestive heart failure therapy. The body rapidly and extensively metabolizes the drug to the active form, canrenone. Spironolactone and canrenone are structurally similar to digoxin. Potassium canrenoate, which is not in the U.S. drug formulary but is used in other countries, is metabolized to canrenone. In some cases, patients on digoxin therapy may also be taking spironolactone, which interferes with the results of digoxin TDM.

Both canrenone and spironolactone cause falsely low digoxin values due to negative interference in digoxin MEIAs (2). Consequently, the assay could produce misleading, sub-therapeutic values, which could lead a physician to increase a patient’s dose to dangerous levels.

Researchers examined nine assays from various manufacturers: Abbott (AxSYM MEIA II, IMx MEIA II, TDx FPIA); Dade Behring (Emit 2000, aca; Dimension); and Roche (Tian-Quant Elecsys) for interference from spironolactone, canrenone, and three metabolites (2). They observed lower than expected values in the presence of 3125 µg/L canrenone for the AxSYM MEIA II (42% of expected value), IMx MEIA II (51%), and Dimension assays (78%). A positive bias was observed for the aca assay (0.7 µg/L), the TDx FPIA (0.62 µg/L), and the Tian-Quant assay (>0.58 µg/L).

Conversely, spironolactone and its metabolite canrenone can falsely elevate serum digoxin levels when measured by FPIA (positive interference) and falsely lower digoxin levels when measured by MEIA (negative interference). The magnitude of interference is more significant with potassium canrenoate, because the concentration of its metabolite, canrenone, can be significantly higher (2).

Complementary and Alternative Medicines

Complementary or alternative medicines have become increasingly popular in recent years. These include a range of traditional medical practices that originated in China. Two such substances that have origins in Chinese traditional medicine are Chan Su and Lu-Shen-Wan, both of which interfere with serum digoxin measurements. Chan Su is prepared from the dried white secretion of the auricular and skin glands of Chinese toads. This compound is also a major component of the traditional Chinese medicines, Lu-Shen-Wan and Kyushin. Chan Su’s cardiotonic effect comes from the bufadienolides compounds it contains, primarily bufalin, which is structurally similar to digoxin.

In one study, researchers found that a patient who had ingested the recommended dose of Chan Su had an apparent digoxin concentration of 0.9 ng/mL, although the individual was never exposed to digoxin. In another report involving a case of a fatal Chan Su overdose, apparent digoxin concentration as measured by the FPIA method was 4.9 ng/mL, well above the 1.9 ng/mL therapeutic upper limit (4). In general, assays using polyclonal antibodies against digoxin, such as the FPIA and MEIA marketed by Abbott, are more susceptible to interference by Chinese medicine than assays that use monoclonal antibodies, such as the particle-enhanced turbidimetric inhibition immunoassay method (Beckman Coulter Synchron LX system) and the Roche Tian-Quant digoxin assay.

Table 2 provides a list of interferences from more Chinese herbal medicines and how they affect digoxin TDM assays (5).

Table 2
Effect of Herbal Supplements on Digoxin TDM

Herb

Protein Binding

Free Digoxin Eliminates Interference?

Chan Su1

80–90%

Yes

Lu-Shen-Wan1

75–85%

Yes

Oleander1

80–95%

Yes

Dan Shen2

70%

Yes

Asian Ginseng2

35–45%

No

Siberian Ginseng2

30–40%

No

Indian Ginseng2

25–35%

No

1 Interferes with both polyclonal antibody-based digoxin immunoassay
and monoclonal antibody-based digoxin immunoassays.

2 Interferes with FPIA only; monoclonal antibody-based digoxin assays
are not affected.

Reference 5.

Oleander Poisoning

Another source of interference in digoxin TDM comes from cardiac glycosides contained in two types of oleander plants, nerium and yellow oleander. These plants are considered extremely poisonous and contain numerous toxic compounds, including oleandrin, nerin, digitoxigenin, and olinerin, with oleandrin being the principal toxin. Accidental poisoning can occur by ingestion, inhalation of smoke from burning oleander, or from the use of medical preparations from the leaves of oleander that have been used as treatments for malaria, leprosy, venereal diseases, and to induce abortions. Oleander is also used as rat poison. Poisoning with oleander or ingestion of oleander containing herbal products causes significant digoxin-like immunoreactivity in patients who are not taking digoxin.

After exposure to oleander, patients may have an apparent digoxin concentration of 1.2 ng/mL. But in the case of a fatal overdose, the value may be 2.1 ng/mL and higher. FPIA has the highest cross-reactivity with oleander extract while the Beckman Coulter digoxin assay (Synchron LX analyzer) and the Siemens turbidimetric assay (ADVIA 1650 analyzer) also show significant interference with oleander (6). The magnitude of interference was approximately 65% less with these assays compared to the FPIA method.

Heterophilic Antibodies

Heterophilic antibodies are poorly defined poly-reactive human antibodies that recognize IgG from different species. These antibodies are non-specific, with no clearly identifiable immunogen, and they commonly bind to the Fc region antibodies in immunoassays, causing interference. While interference of heterophilic antibodies in TDM immunoassays is infrequent, there is one report of a falsely elevated serum digoxin (4.2 ng/mL) in an asymptomatic patient (7). Due to high molecular weight of such antibodies, possible interference can be eliminated by assaying samples after an ultrafiltration step that eliminates large proteins.

Interference in Immunosuppressant Drug Immunoassays

Immunoassays for many commonly used immunosuppressants also suffer from cross-reactivity with metabolites of the parent drug (Table 3). Below are some of the most prominent sources of interference in assays for these drugs.

Cyclosporine Immunoassays

Cyclosporine TDM is essential for management of transplant patients; however, interference from cyclosporine metabolites with immunoassay measurements of the drug is a serious problem. Consequently, HPLC with UV detection or HPLC coupled with mass spectrometry is considered the gold standard for determination of cyclosporine concentrations in whole blood.

Of the commercially available immunoassays, Dade Behring’s antibody-conjugated magnetic immunoassay (ACMIA) has the least overall metabolite cross-reactivity. This fully automated assay requires no specimen pre-treatment, while other immunoassays on the market require a whole-blood pretreatment stage.

FPIAs for cyclosporine have the highest cross-reactivity with the cyclosporine metabolites, AM1 and AM9. The magnitude of metabolite cross-reactivity contributes to the degree of overestimation of cyclosporine concentrations when immunoassay results are compared with values obtained by HPLC. Mean cyclosporine concentrations are higher than HPLC values, but vary with the method used: 12% for ACMIA (Dade Behring); 13% for Syva EMIT (Dade Behring); 17% for Microgenics CEDIA Plus; 22% for FPIA (Abbott TDx analyzer); and 40% for FPIA (Abbott AxSYM analyzer) (4).

Tacrolimus Immunoassays

HPLC combined with UV detection is not suitable for measurement of tacrolimus, because the molecule does not have a chromophore. HPLC/MS is the gold standard method for this drug; however, TDM immunoassays are commercially available. Various metabolites, including M-I (13-O-demethyl), M-II (31- O-demethyl), M-III (15- O-demethyl) and M-V (15, 31, di- O-demethyl), cross-react with immunoassays for tacrolimus. In general the MEIA-II assay (Abbott) produces tacrolimus results that are 15–20% higher than those for the HPLC/MS assay, while Syva EMIT and CEDIA results are 17% and 19% higher than HPLC/MS values, respectively.

Low hematocrit values also have been reported to skew tacrolimus values. Recent studies have reported false-positive tacrolimus concentrations in patients with low hematocrit values and high imprecision at tacrolimus values < 9 ng/mL using the MEIA assay, but the EMIT assay was not affected (10). The researchers reported false-positive results in 63% of the specimens with MEIA for patients who did not receive tacrolimus, while only 2.2% of the specimens analyzed with the EMIT assay had false-positive results. In patients who did not receive tacrolimus, false-positive values ranged from 0.0 –3.7 ng/mL for the MEIA assay and 0.0–1.3 ng/mL for the EMIT assay.

Interference in Antidepressant TDM Immunoassays

Interference is also common in TDM assays used to monitor patients taking certain antidepressants. Metabolites of the antidepressant are responsible for the major interferences (Table 3).

Table 3
Common Interferences in TDM Immunoassays

Drug /Drug Metabolite

Drug Immunoassay Affected

Spironolactone*

Digoxin

Potassium Canrenoate*

Digoxin

Canrenone (metabolite of spironolactone and potassium canrenoate)*

Digoxin

Fosphenytoin

Phenytoin

5-p-hydroxyphenyl 5-phenylhydantoin (HPPH) and its glucuronide conjugate

Phenytoin

Hydroxyzine/Cetirizine

Carbamazepine (Dade Behring—Dimension)

Hydroxyzine/Cetirizine

FPIA for Total Tricyclic (Abbott)

Carbamazepine

FPIA for Total Tricyclic (Abbott)

Carbamazepine 10, 11-Epoxide

Carbamazepine

Cyclosporine metabolites

Cyclosporine

Tacrolimus metabolites

Tacrolimus

*Interference can be positive (falsely elevated digoxin values) or negative (falsely lower digoxin values), depending on the digoxin immunoassay.

Carbamazepine Immunoassays

An anticonvulsant and mood stabilizing drug, carbamazepine is used primarily in the treatment of epilepsy and bipolar disorder. In the steady state, the drug’s active metabolite, carbamazepine 10, 11-epoxide, makes up 15–20% of the total carbamazepine concentration. In overdose cases or in patients with renal failure, however, the concentration is significantly higher. The cross-reactivity of carbamazepine 10, 11-epoxide with different immunoassays for carbamazepine varies from 0 % (Ortho Clinical Diagnostics Vitros) to 94 % (Dade Behring Dimension) (11).

Another source of interference in carbamazepine assays comes from hydroxyzine and cetirizine. Hydroxyzine is a commonly prescribed first-generation antihistamine with sedative properties. It is metabolized to cetirizine, which has antihistamine properties but is devoid of sedative effect. Cetirizine is also used as a second-generation H1-antagonist.

Hydroxyzine interferes with the carbamazepine particle-enhanced turbidimetric inhibition immunoassay (PENTINA) (Dade-Behring Dimension analyzer). For example, a 22-year-old female with a hydroxyzine concentration of 1.8 µg/mL and cetirizine concentration of 2.1 µg/mL showed an apparent carbamazepine level of 5.3 µg/mL. Another patient with a hydroxyzine level of 520 ng/mL and cetirizine level of 2.2 µg/mL demonstrated a carbamazepine level of 25.4 µg/mL. The carbamzepine EMIT 2000 assay, however, showed no cross-reactivity with these molecules (12).

Phenytoin Immunoassays

Phenytoin sodium is a commonly used antiepileptic for treatment of seizures. A phosphate ester product of phenytoin, fosphenytoin, provides improved efficacy and safety when given intravenously or intramuscularly. After systemic administration, the phosphate moiety is rapidly cleaved to form an unstable intermediate that breaks down to yield phenytoin and formaldehyde. Fosphenytoin, a phosphate ester prodrug of phenytoin that provides improved efficacy and safety when given intravenously or intramuscularly, cross-reacts in phenytoin immunoassays. After systemic administration, the phosphate moiety is rapidly cleaved to form an unstable intermediate that breaks down to yield phenytoin and formaldehyde.

Roberts et al. analyzed samples from seven hospitalized, renal-failure patients receiving fosphenytoin intravenously (13). HPLC analysis of the samples did not detect fosphenytoin. However, in comparison to HPLC pheytoin values, all phenytoin immunoassays tested showed higher concentrations of total and free phenytoin. For example, one patient had 5.3 mg/mL phenytoin by HPLC analysis and the immunoassay values were 22.0 mg/mL (Bayer ACS:180), 12.7 mg/mL (Abbott AxSYM), and 28.0 mg/mL (Abbott TDxII). Based on these findings, the researchers proposed the presence of a novel metabolite of fosphenytoin that has a very high cross-reactivity with antibodies used phenytoin immunoassays.

An unusual fosphenytoin metabolite, oxymethylglucuronide, also has been observed in renal failure patients on fosphenytoin therapy. This metabolite cross-reacts with many phenytoin immunoassays and falsely elevates phenytoin concentrations by 2–3-fold.

Researchers have also found that a phenytoin metabolite, 5-p-hydroxyphenyl 5-phenylhydantoin (HPPH), and its glucuronide conjugate may accumulate in patients with uremia. HPPH has 16% cross-reactivity and HPPH-glucuronide has 1.6% cross-reactivity with the Bayer ADVIA phenytoin immunoassay (14).

Interaction of St. John’s Wort with Western Drugs

St. John’s wort is a plant with yellow flowers that has been used for centuries for health purposes, including depression and anxiety. Although components of St. John’s wort do not interfere in most TDM immunoassays, the pharmacokinetic interactions of many Western drugs with St. John’s wort causes significant reductions in drug concentrations (Table 4). Because most patients do not report the use of complementary and alternative medicine to their clinicians, such results may be the first clue that the patient is taking St. John’s wort.

Table 4
Common Western Drugs Affected by Interaction with St. John’s Wort*

Cyclosporine

Alprazolam

Ritonavir

Tacrolimus

Midazolam

Atazanavir

Digoxin

Methadone

Lopinavir

Theophylline

Verapamil

Indinavir

Oral Contraceptives

Simvastatin

Amitriptyline

Omeprazole

Irinotecan

Coumadin

*In all cases, the drug concentration is lowered.

Lower than expected levels of certain drugs have been attributed to two active components in St. John’s wort, hyperforin, and hypericin. Hyperforin has been shown to induce CYP3A4 and CYP2B6 probably through activation of nuclear/pregnane and xenobiotic receptors, and hypericin modulates P-glycoprotein. Therefore, St. John’s wort can increase metabolism of many Western drugs that are metabolized via cytochrome P-450 or non–liver pathways (15).

Miscellaneous Sources of Interference

Other sources of miscellaneous interference in TDM assays also exist. Nonspecific interference from high concentrations of bilirubin (>20 mg/dL) and triglycerides (>1000 mg/dL) may occur with certain TDM immunoassays (16). High protein levels, especially from paraprotein, an abnormal protein in the urine or blood most often associated with benign monoclonal gammopathy of undetermined significance, and multiple myeloma may also interfere with immunoassays of drugs, but the incidence of this is relatively low.

Recognizing Interferences

Both endogenous and exogenous factors interfere with TDM immunoassays and can produce either falsely elevated or falsely lowered values. While HPLC produces more precise analytical values, it is more time-consuming than analysis on automated immunoassay instruments.

Clearly, TDM has improved patient care, but many studies suggest that it could do even better. Until superior, less time- consuming methods are available, laboratorians need to be proactive in educating clinicians about the limitations of assays for certain therapeutic drugs. Comprehensive TDM programs can improve the probability that drug concentrations will be within the therapeutic range, that samples will be collected appropriately, and that results will be used appropriately. Implementing such programs, however, requires a major commitment of resources from the lab. To be successful, laboratorians need to provide clinicians with ongoing education, such as bulletins and newsletters, to keep them current with information about the assays.

REFERENCES

McMillin GA, Qwen W, Lambert TL, De B et al. Comparable effects of DIGIBIND and DigiFab in thirteen digoxin immunoassays. Clin Chem 2002;48: 1580–1583.

Steimer W, Muller C, Eber B. Digoxin assays: frequent substantial and potential dangerous interference by spironolactone, canrenone and other steroids. Clin Chem 2002;48:507–516.

Panesar NS. Bufalin and unidentified substances in traditional Chinese medicine cross-react in commercial digoxin assay. Clin Chem 1992; 38: 2155–2156.

Ko R, Greenwald M, Loscutoff S, Au A et al. Lethal ingestion of Chinese tea containing Chan Su. West J Med 1996;164:71–75.

Dasgupta A, Bernard DW. Complementary and alternative medicines: Effects on clinical laboratory tests. Arch Pathol Lab Med 2006;130:521–528.

Dasgupta A, Datta P. Rapid detection of oleander poisoning by using digoxin immunoassays: comparison of five assays. Ther Drug Monit 2004; 26: 658–663.

Liendo C, Ghali JK, Graves SW. A new interference in some digoxin assays: anti-murine heterophilic antibodies. Clin Pharmacol Ther 1996; 60: 593–598.

Schütz E, Svinarov D, Shipkova M, Niedmann PD, Armstrong VW, Wieland E, Oellerich M. Cyclosporin whole blood immunoassays (AxSYM, CEDIA, and Emit): a critical overview of performance characteristics and comparison with HPLC. Clin Chem 1998;44:2158–2164.

Butch A. Introduction to Immunosuppressive Drug Monitoring. In: Hammett-Stabler C and Dasgupta A, eds. Therapeutic Drug Monitoring Data: A Concise Guide. Washington, DC: AACC Press, 2007:159–160.

Armedariz Y, Garcia S, Lopez R et al. Hematocrit influences immunoassay performance for the measurement of tacrolimus in whole blood. Ther Drug Monit 2005; 27:766–769.

Hermida J, Tutor JC. How suitable are currently used carbamazepine immunoassays for quantifying carbamazepine 10, 11-epoxide in serum samples? Ther Drug Monit 2003;25:384–388.

Parant F, Moulsma M, Gagnieu MC, Lardet G. Hydroxyzine and metabolite as a source of interference in carbamazepine particle-enhanced turbidimetric inhibition immunoassay (PENTINA). Ther Drug Monit 2005;27:457–462.

Roberts W, Rainey P. Phenytoin overview: metabolite interference in some immunoassays could be clinically important. Arch Pathol Lab Med 2004;128:734.

Roberts W, De B, Coleman J, Annesley T. Falsely increased immunoassay measurement of total and unbound phenytoin in critically ill uremic patients receiving fosphenytoin. Clin Chem 1999;45:829–837.

Venkataraman R, Komoroski B, Strom S. In vitro and in vivo assessment of herb-drug interactions. Life Sci 2006; 78: 2105-2115.

Dasgupta A. Effect of bilirubin, lipemia, hemolysis, paraproteins, and heterophilic antibodies on immunoassays for therapeutic drug monitoring. In: Hammett-Stabler C and Dasgupta A, eds. Therapeutic drug monitoring data: a concise guide. Washington, DC: AACC Press, 2007:27–33.

Dasgupta

Amitava Dasgupta, PhD, DABCC, is Professor of Pathology and Laboratory Medicine at the University of Texas Health Sciences Center at Houston. He is also the Director of Clinical Chemistry, Toxicology and Point of Care Testing at the Laboratory Services of Memorial-Hermann Hospital at Houston.