Drug Interference in Laboratory Testing - ACS Symposium Series

Jul 23, 2009 - Drug Interference in Laboratory Testing. DONALD T. FORMAN. Department of Pathology and Laboratory Medicine, Evanston Hospital, ...
0 downloads 0 Views 1MB Size
10 Drug Interference in Laboratory Testing

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

DONALD T. FORMAN Department of Pathology and Laboratory Medicine, Evanston Hospital, Evanston, Ill. 60201 DONALD S. YOUNG Clinical Center, National Institutes of Health, Bethesda, Md. 20014

Over the past decades, the practice of medicine has undergone considerable sophistication. Potent new drugs in p r a c t i c a l l y every therapeutic category are available. Whenever a patient receives a drug there is a p o s s i b i l i t y of drug-test i n t e r f e r ­ ences. When more than one drug is administered there may also be drug-drug interactions. These developments have created new problems for the laboratory s c i e n t i s t . A characteristic problem is the interpretation of laboratory data when a drug or its metabolite may interfere with a laboratory procedure. One drug may affect the action of another given at the same time, thus changing the value of a test r e s u l t . Numerous attempts have been made by a number of compilations and a r t i c l e s to document these effects. In 1962, Caraway (2) published a comprehensive article on the chemical and diagnostic s p e c i f i c i t y of laboratory tests. In 1964, Borushek and Gold (1) published an article con­ taining a list of drugs interfering with routine endocrine pro­ cedures. Wirth and Thompson (19) in 1965 published a list of substances and conditions which affect the results of laboratory procedures. They included such factors as temperature, l i g h t and drugs. Other review a r t i c l e s (8,11,14,17) have been of great value in helping to define this problem. One of the most ambitious approaches to t h i s problem has been the development of a 9000-entry computerfile based on a compilation of 1030 litera­ ture references (20). This computerized compilation of effects of drugs on laboratory tests has been used to assist in the interpretation of unusual test results in the clinical chemistry laboratories of the National Institutes of Health and the Evanston Hospital of Northwestern University Medical Center, as well as in some other centers. A major problem facing users of the various collations of laboratory test interferences is the relationship between drug dosage and diagnostic test interference and the possible syner­ gistic effect of several drugs. The metabolic and clinical states 271

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

272

CLINICAL CHEMISTRY

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

of the patient also affect the degree of laboratory interference. If an individual has normal kidney and l i v e r function, he w i l l respond d i f f e r e n t l y to a drug dosage than the patient who i s unable to detoxify or excrete the compound. In order to cope with the problem of drug induced modifi­ cations of laboratory test values, laboratory s c i e n t i s t s should possess an understanding of the mechanisms involved i n these interferences. The purpose of t h i s a r t i c l e i s to consider these mechanisms and discuss the role of the computerized drug-test interference f i l e i n assisting i n the prediction as w e l l as interpretation of apparent test r e s u l t s . Mechanisms of Interference There are four d i s t i n c t mechanisms of interference i n laboratory testing. Physical, chemical, pharmacological and drug-drug interactions are of special interest to the laboratory s c i e n t i s t . The mechanisms of the interferences w i l l be b r i e f l y reviewed and wherever possible, solutions suggested. Physical Effects Drugs may interfere with colorimetric, photometric or fluorometric analyses by imparting a characteristic color to the specimen. Table I describes some unusually colored urines and their relationship to drug administration or metabolic disorders. This type of interference can be detected v i s u a l l y by laboratory personnel and drug interferences may be circumvented by c o l l e c t ­ ing the specimen after an appropriate length of time during which the drug i n question has been discontinued. There are many v a r i e t i e s of drugs and foods that can cause t h i s type of interference. These should be considered f i r s t as possible causes of any unusually colored urine. Some drugs act as indicators (e.g. phenolphthalein, vegetable laxatives) and affect tests carried out at a particular pH. The presence of sulfobromophthalein dye (BSP) i n serum w i l l interfere with serum protein determined by the biuret method. The dye i s colorless at the pH of blood, but purple when made alkaline during the assay. Chemical Effects Drugs may be measured as analytes, either because of t h e i r s i m i l a r i t y i n chemical structure or because they contain a com­ ponent that i s an analyte. Patients who have received radio­ graphic contrast media containing iodine w i l l have an altered protein-bound iodine, although thyroxine i s not affected when determined by protein binding methods. The most serious and long-lasting interferences are those caused by intravenously administered iodinated radiopaque agents (e.g. acetrizoate and

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

10.

FORMAN

A N D YOUNG

Drug Interference in Laboratory Testing

273

TABLE I Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

COLORED URINES DUE TO DRUGS OR METABOLIC DISORDERS Color of Urine

Cause

Disorder

Orange-Red

Phenazopyridine Phenindione Phensuximide

Drug Related

Red

Porphyrins

Congenital Porphyria Acute Intermittent Porphyria

Red-Brown

Hemoglobin and Derivatives

Crush Syndrome

Red-Brown

Urobilin

Hemolytic Anemia

Yellow

Chloroquine

Drug Related

Green-Yellow

B i l e Pigments

Regurgitative Janudice

Blue

Methylene Blue Triamterene

Drug Related

Blue

Indigo Compounds

Intestinal Disease (Indicanemia)

Blue-Green

Amitriptylene

Drug Related

Brown-Black

Melanin Homogentisic Acid

Malignant Melanoma Alkaptonuria

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

274

CLINICAL CHEMISTRY

adipiodone)· The drugs orabilex and d i o n s i l have a negligible effect on the analysis of protein bound iodine at a concentration no greater than 100 yg/dl, but seriously affect the procedure at 1000 yg/dl. Serum potassium concentration i s increased by the concurrent administration of intravenous potassium p e n i c i l l i n G. The p e n i c i l l i n preparation contains 1.7 mmol of potassium per m i l l i o n units. Thus, a patient receiving 10 m i l l i o n units of the a n t i ­ b i o t i c receives 17 mmol (mEq.) of potassium. A chemical interference may also result when the substance being determined reacts i n vivo with a drug administered p r i o r to the c o l l e c t i o n of blood for analysis. For example, serum calcium determinations may be affected by the administration of the chelating agent ethylenediaminetetracetate (EDTA) (3). In certain conditions such as lead poisoning and collagen disease, EDTA has been used therapeutically to reduce the lead and calcium concentrations, respectively. Less w e l l recognized i s the fact that various chelating agents are also used as preser­ vatives to prevent oxidation reactions i n drug preparations. In patients on EDTA therapy, calcium cannot be determined by the indirect colorimetrie or fluorometric methods based on the chelation of a calcium - EDTA complex. However, i n calcium determinations by atomic-absorption spectroscopy, the complexing agent i s destroyed i n the flame and the direct concentration of calcium can be determined. Drugs can also interfere with laboratory results by negating certain nonspecific oxidation and reduction reactions essential for the chemical assay. P e n i c i l l i n , streptomycin and ascorbic acid are known to react with cupric ion; thus, false p o s i t i v e results for glucose may occur i f a copper reduction method i s used. I f the s p e c i f i c enzymatic glucose-oxidase method i s employed, ascorbic acid can cause a f a l s e negative result by preventing the oxidation of a specific chromogen i n the reaction. Increased transaminase a c t i v i t y has been observed when the diazocolorimetric method i s used, unusually increased a c t i v i t y has been reported i n patients with ketosis and also i n patients receiving erythromycin (13) or p-aminosalicylic acid (9). These test interferences can be obviated by employing ultraviolet k i n e t i c procedures. Colorimetric procedures used i n steroid assays are often subject to drug interference. In the determination of 17-Ketosteroids by the Zimmerman reaction, drugs with the 17-Keto basic structure such as ascorbic acid, morphine and reserpine w i l l cause increased values. In the determination of 17,21 dihydroxysteroids by the Porter-Silber reaction the dihydroxyacetone chain i s the reactive unit. Drugs l i k e meprobamate, chloral hydrate, chloropromazine and potassium iodide w i l l i n t e r ­ fere with t h i s reaction and cause elevated values. In the colorimetric determination of vanillylmandelic acid (VMA) by a diazo reaction, drugs l i k e methocarbamol and methyl dopa cause

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

10.

FORMAN

AND YOUNG

Drug Interference in Laboratory Testing

275

f a l s e l y elevated r e s u l t s . Where specific drugs have been demonstrated to interfere with chemical reactions, patients should be maintained free of these drugs for at least 72 hours before c o l l e c t i n g the specimen. Other a n a l y t i c a l techniques, e.g. column chromatography and radioimmunoassay procedures, can also be substituted for an affected method.

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

Pharmacological

Effects

This type of interference usually results from the pharma­ cological or toxic a c t i v i t y of drugs but may also a r i s e from therapeutic quantities of drugs. An example of this type i s the increase i n prothrombin time which results from the administra­ tion of a therapeutic amount of sodium warfarin. However, unexpected interferences r e s u l t i n g from side effects or adverse reactions may create problems i n the interpretation of laboratory data. Disturbances i n f l u i d and e l e c t r o l y t e balance and hypopotassemia are common. Thiazide d i u r e t i c s and large doses of glucocorticosteroids may severely reduce the concentration of plasma potassium. Acetazolamide and ethoxazolamide may cause metabolic acidosis, a raised serum u r i c acid, and a lowered urinary excretion of u r i c acid. A r i s e i n serum u r i c acid has also been reported following the administration of levodopa (4). The thiazides can also have a hyperglycemic a c t i v i t y and diabetics may require more i n s u l i n or o r a l antidiabetic drugs. Patients with latent diabetes occasionally develop an abnormal glucose tolerance curve. Salicylates have also been reported to reduce hyperglycemia i n diabetics and produce hypoglycemia i n normal children. Tables I I and I I I l i s t various drugs capable of increasing or decreasing blood glucose i n non-diabetic subjects. Hydantoin anticonvulsants may cause the appearance of positive L.E. c e l l s and methemoglobinemia. Chloropromazine has been implicated i n causing a syndrome l i k e systemic lupus erythematosus with accompanying positive tests for L.E. c e l l s and antinuclear antibodies (6)· In many instances administered drugs a l t e r a metabolic pathway and d i r e c t l y produce changes i n biochemical values. Insulin administration results i n changes i n blood glucose, potassium, phosphorus and f a t t y acids. A l l o p u r i n o l reduces u r i c acid by i n h i b i t i n g the enzyme, xanthine oxidase; other drugs e.g. azoserine and methotrexate do so by preventing i t s bio­ synthesis. Hydrochlorothiazide increases u r i c acid retention by decreasing i t s tubular excretion. Oral contraceptives have a marked effect on plasma glucose, f a t t y acids and growth hormone levels (16). Estrogen-progestin o r a l contraceptives produce increases i n serum iron, iron-binding capacity, transferrin, ceruloplasmin and copper (18). These effects on endocrine functions are quite marked as evidenced by measurement of the related biochemical constituents (Table IV) (12).

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

CLINICAL CHEMISTRY

TABLE I I

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

DRUGS CAPABLE OF INCREASING BLOOD GLUCOSE Dextrothyroxine

Nicotinic Acid

Diazoxide

Phenothiazines

Diphenylhydantoin

Steroids Adrenocorticosteroids Estrogens

Diuretics

Sympathomimetic Amines

TABLE I I I DRUGS CAPABLE OF DECREASING BLOOD GLUCOSE Alcohol

Propoxyphene

Asparaginase

Propranolol

Caffeine

Anabolic Agents

Haloperidol Monoamine Oxidase Inhibitors

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

10.

FORMAN AND

YOUNG

Drug Interference in Laboratory Testing

277

Some drugs induce unexpected side effects unrelated to their therapeutic actions. Inheritance of hemoglobins d i f f e r i n g i n structure from normal hemoglobin A i s w e l l described. A l l types of hemoglobin are oxidized to methemoglobin by a wide range of agents and about 1% of hemoglobin present i n the blood of normal people i s i n t h i s form. Many drugs s l i g h t l y increase the amount of methemoglobin, even i n normal people, but induce a s t r i k i n g and sometimes f a t a l methemoglobinemia i n susceptible patients. These drugs include chlorates, sulfonamides, n i t r i t e s , quinones, acetanilid and acetophenetidin. A similar phenomenon occurs with primaquine s e n s i t i v i t y . Susceptible patients have an inherited defect i n red c e l l glucose 6-phosphate dehydrogenase (G6PD) and develop a severe hemolytic anemia when exposed to primaquine. Another abnormal drug response i s porphyria. Acute porphyria on exposure to drugs can be caused by sudden massive induction of deltaamino l e v u l i n i c acid synthetase i n hepatic mitochondria; which results i n uncontrolled formation of porpho­ bilinogen and i t s metabolites. Susceptibility i s inherited as an autosomal dominant t r a i t and occurs even i n hétérozygotes. The nature of the molecular defect i s unclear and presumably l i e s i n the repression mechanism for the gene controlling forma­ tion of the enzyme protein. Exposure to any of the drugs l i s t e d i n Table V results i n further marked de-repression of enzyme synthesis and severe porphyria. Table VI l i s t s several drugs inducing hepatic microsomal enzymes (5)· These enzymes can metabolize the drug as well as other substrates. Barbiturates, griseofulvin, and glutethimide induce enzymes which metabolize coumarin and phenindione derivatives and thus reduce their anticoagulant a c t i v i t y . Diphenylhydantoin and phenylbutazone stimulate C o r t i s o l hydroxy­ l a s e a c t i v i t y and i n c r e a s e the urinary excretion of B-hydroxy C o r t i s o l and decrease the concentration of C o r t i s o l i n the plasma. Adverse side reactions may also occur with pharmacologically active drugs. Many drugs have been associated with inducing abnormal l i v e r , renal and pulmonary function (10). These drugtest effects are unexpected and should be recognized. Drugs i n this group include methyltestosterone, reserpine, phenacemide, oxyphenylbutazone, anesthetics and a variety of cancer chemotherapeutic agents. Drug-Drug Interactions Drugs which independently have l i t t l e effect on laboratory tests may interact to produce a significant alteration of test values. Table VII presents several drug interactions i n diabetic human subjects which can affect an apparent test result (15). The administration of c l o f i b r a t e to a patient t a k i n g ~ a r f a r i n w i l l potentiate the anticoagulant effect of warfarin by displacing i t from i t s protein binding s i t e (7). This interaction w i l l cause

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

278

CLINICAL CHEMISTRY

TABLE IV HORMONE LEVELS IN WOMEN ON ORAL CONTRACEPTIVES* Hormone

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

Plasma Cortisol (8 AM) Serum Thyroxine

Estrogens, Total (Urine)

*

Level i n Women on Oral Contraceptives

15.0 ± 4.0 yg/dl

40.0+11 yg/dl

± 1.4 yg/dl

9.4 t 1.7 yg/dl

4.6 ± 1.1 mg/d

2.6 ± 0.9 mg/d

G.6

Urinary 17-OH Corticosteroids

Vanilylmandelic Acid (Urine)

Level i n Menstruating Women

40.0 ± 18 yg/d 3.5 ± 1.6 mg/d

26.0 ± 7.4 yg/d 3.3 ± 1.6 mg/d

from Lucis, O.J. and Lucis, R.; B u l l . W.H.O. (1972), 46, 443-450.

TABLE V DRUGS INDUCING MARKED DE-REPRESSION OF ENZYME SYNTHESIS Chloroquine Aminopyrine Sulfonamides Diallylbarbiturate Allylisopropylacetylurea Hexachlorobenzene

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

FORMAN

AND YOUNG

Drug Interference in Laboratory Testing

TABLE VI

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

DRUGS INDUCING HEPATIC MICROSOMAL ENZYMES Phenylbutazone

Barbiturates

Cortisone

Glutethimide

Aminopyrine

Chlordiazepoxide

Diphenylhydantoin

Testosterone

Carbutamide

Norethynodrel

TABLE V I I DRUG INTERACTIONS REPORTED IN DIABETIC SUBJECTS Hypoglycemic Drug

Reaction

Insulin Sulfonylurea

Inhibition of Gluconeogenesis

Phenformin

Lactic Acidosis

Bishydroxycoumarin

Chlorpropamide

Decrease i n Excretion or Metabolism of Hypoglycemic Agent

Guanethidine

Insulin

Decrease i n Insulin Dosage Needed to Control Patient

Oxytetracycline

Insulin

Decrease i n Insulin Dosage Needed to Control Patient

Phenothiazine

Insulin

Increase i n Insulin Dosage Needed to Control Subject

Propranolol

Insulin

Decrease i n Insulin Dosage Needed to Control Subject

Drug Interaction Alcohol

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

280

CLINICAL CHEMISTRY

a s i g n i f i c a n t increase i n the prothrombin time test used to regu­ late warfarin administration. Displacement of drugs from their binding sites i s a common mechanism by which drug-drug i n t e r ­ actions result i n unexpected laboratory data. After absorption, about 50 to 80 percent of s a l i c y l a t e i s bound to serum albumin, from which i t displaces other substances such as b i l i r u b i n and thyroxine. Other drugs such as coumarin may also be displaced with a sudden change i n prothrombin time. Drug-drug interactions may also cause enzyme induction. Ethchlorvynol, glutethimide and haloperidol are known to i n h i b i t the effectiveness of warfarin therapy by this mechanism (7). These drug-drug interactions may result i n a reduced prothrombin time. Enzyme i n h i b i t i o n i s another result of drug-drug interaction and the impairment of the hepatic glucuronyl transferase system by phenothiazine derivatives i s a good example. Whenever several drugs are administered concurrently to a patient, their combined effect on laboratory results should be c a r e f u l l y considered. Effectiveness of Computerized Drug Interference F i l e The computerized drug interference f i l e has been used i n two modes. In one, possible interactions are l i s t e d i n response to direct queries (the r o l e for which i t was o r i g i n a l l y designed). In the other, the f i l e has been used to provide automatic i n t e r ­ pretation of effects of drugs on t e s t s . In this mode, the f i l e i s used to provide a report a l e r t i n g a physician to a possible interaction between a drug and a test (Figure 1). Without a computerized hospital information system, which could be accessed to provide information about a patient's c l i n i c a l state and non therapeutic-drug procedures, the drug f i l e cannot be completely effective i n providing explanations for changes i n laboratory data. I t s r o l e must be that of a warning system to highlight possible causes of changes i n test values which can be accepted or rejected by the patient's physician on the basis of his c l i n i c a l judgement. Our experience i n a University teaching h o s p i t a l , i n which over 10,000 patient days were monitoried, indicated that the drug f i l e provided the correct explanation for changes i n laboratory data i n approximately 21 percent of a l l the occasions i n which a report was produced (21). A report, similar i n content and format to an entry i n the master f i l e (20), was generated when an abnormal test result was produced and one of the administered drugs had been described as causing this effect. The y i e l d of correct explanations would undoubtedly be increased by elimination of effects that were u n l i k e l y to arise i n hospital practice, e.g. effects due to overdoses or toxic effects of drugs. To make i t appropriate for on-line use i n any particular hospital the f i l e should be modified to eliminate these effects. The only method­ ological interferences that should be included are those involving the procedures i n use i n the hospital laboratory. The f i l e i s

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

Drug Interference in Laboratory Testing

281

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

10. FORMAN AND YOUNG

Figure 1. Report alerting physician to drug interference in laboratory testing

In Clinical Chemistry; Forman, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1976.

Downloaded by EAST CAROLINA UNIV on September 16, 2015 | http://pubs.acs.org Publication Date: June 1, 1976 | doi: 10.1021/bk-1976-0036.ch010

282

CLINICAL CHEMISTRY

being modified for use i n the automatic warning role i n a com­ munity hospital to determine i f i t can provide a useful service for physicians with a different orientation from those i n a university hospital. In the o f f - l i n e mode of use of the drug f i l e the data-base, stored i n a time-shared computer, i s interrogated through a remote device such as a cathode ray terminal. The questions that may be asked include a l l the affects of a drug, a l l the drugs that affect a test, or more s p e c i f i c a l l y does a particular drug affect a particular test procedure by a particular mechanism. In each of these situations i t i s possible to l i s t the appro­ priate l i t e r a t u r e reference. The search procedure has been designed i n an interactive mode so that even an individual who has never used i t may obtain an explanation for a particular problem, but the experienced searcher can by-pass the longer interactive search routine. This mode of operation i s i n use only at the National Institutes of Health and a simpler procedure i s used i n Evanston Hospital. The most important use of a computer i n setting up the CLAUDE (computer l i s t i n g of abnormal and usual drug effects) concept may be that of a text editor and sorter. Scanning a printout i s much more e f f i c i e n t than searching the f i l e i n the computer except for any information that i s not yet available on a printout. I t i s our intention to reproduce printouts of the f i l e i n a journal with a wide c i r c u l a t i o n to the appropriate audience when a s i g n i f i c a n t amount of new information has been added to the master-file. The second edition has recently been published (22). I t includes some information on the p r o b a b i l i ­ t i e s of certain effects occurring as w e l l as an indication of the concentration of drug at which the effects occur, at least for the methodological effects. The f i l e i s used routinely i n the laboratory at the National Institutes of Health i n an attempt to explain abnormal test results. The resident physicians a f f i l i a t e d with the C l i n i c a l Chemistry Service discuss the results with the patient-care physicians and determine i f the results were due to the patient's c l i n i c a l state or to a drug effect. This close monitoring of test results has led to recognition of deficiencies i n what i s believed are specific enzymatic procedures for the measurement of glucose and u r i c acid. Likewise, the guaiac procedure for occult blood i n feces was found to y i e l d false negative results under certain circumstances. This has prompted the development of a more s p e c i f i c procedure (Jaffe et a l . unpublished). The f i l e may be queried i n response to questions from c l i n i c a l staff at the National Institutes of Health and physi