!DOCTYPE HTML>

CHAPTER 3

PRIMER ON DRUG INTERFERENCES WITH TEST RESULTS

MARY LEE

Objectives

After completing this chapter, the reader should be able to

• Distinguish between in vivo and in vitro drug interferences with laboratory tests
  
• Identify suspected drug–laboratory test interference in a logical, systematic manner given a drug and a laboratory test
  
• Devise a stepwise process to confirm that a drug is causing a clinically significant drug–laboratory test interference
  
• Analyze differences among tertiary, secondary, and primary literature resources about drug–laboratory test interferences
  
• Apply a systematic method to search and identify medical literature relevant to a suspected drug–laboratory test interference situation

Through a variety of mechanisms, drugs can interfere with laboratory test results. If the clinician who has ordered the laboratory test is not aware that the drug has altered the results of the test, inappropriate management of the patient may follow including unnecessary hospitalization, extra office visits, or additional laboratory or clinical testing—all of which may increase the cost of healthcare. This chapter addresses this situation and provides resources that can be used by health professionals to better interpret laboratory tests when a drug is suspected to cause an interference with test results.

IN VIVO AND IN VITRO DRUG INTERFERENCES WITH LABORATORY TESTS¹

When a drug interferes with a laboratory test result, it alters the lab value. Mechanisms for drug interference of clinical laboratory tests can be classified as either in vivo or in vitro. In vivo drug interferences can also be called biological and can be subclassified as pharmacological or toxicological. In vivo interferences account for most effects of drugs on laboratory tests.² In contrast, the term in vitro interference is used synonymously with analytical or methodological.

In Vivo Interference

An in vivo interference is an actual change in the analyte concentration or activity prior to specimen collection and analysis. The assay measurement is true and accurate and reflects a change in the measured substance that has occurred in the patient. Therefore, an in vivo interference will always change a laboratory test result, independent of the assay methodology. A drug can produce an in vivo interference in several ways. By a direct extension of its pharmacological effects, a drug can produce changes in some lab test results. For example, thiazide and loop diuretics will commonly cause increased renal elimination of potassium. Therefore, decreased serum potassium levels can occur in treated patients. In these patients, hypokalemia is true and accurate. Similarly, increased blood urea nitrogen (BUN) levels can occur as a result of excessive fluid loss during treatment with thiazide and loop diuretics.

Other drugs produce changes in lab test results by producing in vivo toxicological effects. As the drug damages a particular organ system, abnormal laboratory tests may be one of the first signs of the problem. For example, as isoniazid and rifampin produce hepatotoxicity, elevated hepatic transaminases will herald the onset of liver inflammation. Similarly, as a prolonged course of high-dose aminoglycoside antibiotic causes acute renal failure, serum creatinine and serum trough aminoglycoside levels will increase steadily. In the face of cyclophosphamide-induced bone marrow suppression, neutropenia will become evident 10–14 days after the dose has been administered.

In Vitro Interference

Drugs in a patient’s body fluid or tissue can directly interfere with a clinical laboratory test during the in vitro analytical process. This type of drug–laboratory test interaction is highly dependent on the laboratory test methodology, as the reaction may occur with one specific assay method but not another. For example, serum digoxin levels are commonly determined using a radioimmunoassay, a fluorescent polarization immunoassay, or a TDx assay. However, these assays are based on the three-dimensional structure of the digoxin molecule, and many other drugs with a similar chemical structure to digoxin (e.g., spironolactone, estrogen replacement products, cortisol, or digoxin-like substances) can cross-react with the assay.³ A falsely increased or decreased serum digoxin level can result.^3,4 To determine the true serum digoxin level in this situation, another assay technique (e.g., high pressure liquid chromatography) may be used. A similar problem occurs with fosphenytoin, which cross-reacts with phenytoin when measured with immunoassay methods.⁵ In addition, substances that are prepackaged in or added to the in vitro system before or after sample collection can cause laboratory test interference in vitro. As an example, test tubes sometimes contain lithium heparin or sodium fluoride. Heparin can interfere with aminoglycoside assays, and fluoride can cause false increases in BUN when measured by the Ekatchem assay.

Alternatively, a drug may cause discoloration of the body fluid specimen, which may interfere with colorimetric, photometric, or fluorometric laboratory-based assay methods. For example, phenazopyridine causes an orange–red discoloration of urine that may be mistaken for blood. Nitrofurantoin may cause a brown discoloration of the urine that may cause alarm for the patient. These types of drug interference with lab testing can be detected visually and appropriate attribution of the abnormality should be made by knowledgeable clinicians.

Other common mechanisms by which drugs cause in vitro interferences with laboratory tests include the following:

1. A drug reacts with reagent to form a chromophore (e.g., cefoxitin or cephalothin) with the Jaffe-based creatinine assay.
   
2. A drug reacts with immunoassay’s antibody that is intended to be specific for the analyte. For example, caffeine cross-reacts in the theophylline assay; digitoxin, digoxin metabolites, antigen-binding fragments derived from antidigoxin antibodies (used for treating digoxin intoxication), spironolactone, and canrenone (the major metabolite of spironolactone) cross-react with the digoxin immunoassays.³
   
3. A drug alters the specimen pH (usually urine) so that reagent reactions are inhibited or enhanced. For example, acetazolamide produces an alkaline urinary pH that causes false-positive proteinuria with reagent dip strips.
   
4. A drug has chemical properties similar to the analyte. For example, patients who receive radiographic contrast media, which contain iodine, may exhibit altered laboratory values for protein-bound iodine.
   
5. A drug chelates with an enzyme activator or reagent used in the in vitro laboratory analysis.
   
6. A drug absorbs at the same wavelength as the analyte. For example, methotrexate interferes with analytic methods using an absorbance range of 340–410 nm.

In addition to the parent drug, other drug-related components may cause significant interferences with laboratory tests. Metabolites can cross-react with the parent drug in an assay, such as in the case with cyclosporine. Its metabolites cross-react with the parent drug in high-pressure liquid chromatography assays and can produce a falsely high measurement of the concentration of cyclosporine.⁶ Inactive ingredients of some drug products may influence assay results. Inactive ingredients in dosage formulations include excipients such as lactose or starch, preservatives, colorants, or flavoring agents. Although most manufacturers do report the inactive ingredients in their products, little systematic research has been performed to assess the impact of these substances on laboratory tests. Compounding these factors, many laboratory test interferences are concentration-related, and many drug metabolites and their usual plasma concentrations have yet to be identified. Therefore, systematic study of all of these potential causes of interactions is difficult to conduct and is not available in many cases.⁷

Simultaneous In Vitro and In Vivo Effects

Some drugs can affect an analyte both in vivo and in vitro. In these rare situations, interpretation is extremely difficult because the degree of impact in each environment cannot be determined easily. The prototypic example involves reaction of aminoglycosides with penicillins, which leads to a loss of antibacterial activity in vivo and can decrease measured aminoglycoside concentrations and antibacterial activity in vitro. Although this inactivation mechanism is unclear, it seems to involve the formation of an adduct between the aminoglycoside and beta-lactam ring of the penicillins.

In vitro, carbenicillin is the most potent inactivator of aminoglycosides when compared to the other penicillins; tobramycin is the most susceptible when compared to the other aminoglycosides; and amikacin and netilmicin are degraded the least. Refrigeration does not significantly slow the reaction but centrifugation with freezing does. Unreasonably low aminoglycoside concentrations in patients receiving concomitant high-dose penicillins should alert the practitioner to this in vitro interaction.^8-10

In vivo inactivation of aminoglycosides by a parenteral penicillin is a problem primarily in patients with renal failure. With normal renal function, these antibiotics are excreted at a rate faster than their interaction rate. In one study, the in vivo inactivation rate constant of gentamicin in patients with renal failure receiving carbenicillin was 0.025 hour^–1. The inactivation rate constant was defined as the difference between the elimination rate constant for gentamicin alone and that for its combination with carbenicillin. Thus, in renal failure patients, carbenicillin apparently eliminates gentamicin at the same rate as do the kidneys. The change corresponded to a shortening of the gentamicin half-life from 61.6–19.6 hours after carbenicillin was added. Fortunately, carbenicillin is no longer commercially available by the parenteral route. However, this drug–laboratory test interaction occurs with other penicillins.

The effect of penicillins on aminoglycosides can be minimized several ways⁷:

1. To eliminate in vitro problems, the sample can be centrifuged and frozen soon after it is drawn.
   
2. Antibiotics with less interaction potential (e.g., amikacin and piperacillin, or amikacin and a cephalosporin) can be used.
   
3. Administration times can be separated or blood can be drawn just before the next scheduled penicillin dose to decrease the penicillin concentration in the specimen tube, the concentration-exposure time (in vivo), and, thus, the degradation time.

This last maneuver would be less applicable in patients with renal failure, particularly if the penicillin dose was not adjusted. However, decreasing the daily penicillin dose according to renal function and target serum concentration would make testing more accurate even for these patients.

Simultaneous in vitro and in vivo effects also occur when drugs produce lipemia, hemolysis, or hyperbilirubinemia.⁷ Drugs that increase cholesterol blood levels (e.g., protease inhibitors, estrogens, corticosteroids) do so by a variety of different mechanisms. The increase in lipids in the blood specimen produce a cloudy appearance, which interferes with light transmission, an essential component of nephelometric and turbidometric analytic procedures.¹¹ Similarly, drugs that produce hemolytic anemia cause in vivo toxic effects on red blood cells. As hemoglobin is released from lysed red blood cells into the circulation, in vitro interference with analytic methods results from a variety of mechanisms: (1) hemoglobin interferes with assay reactions directly; (2) cytoplasmic constituents of lysed red blood cells enter the bloodstream and produce elevations of plasma potassium and magnesium; and (3) hemoglobin interferes with optical absorbance measures on spectrophotometric assays.^12,13

Some immunoassays use mouse monoclonal antibody as the reagent. Human antimouse antibody can interfere with the mouse monoclonal antibody reagent leading to incorrect assay results for creatine kinase and thyroid function.¹³ Human antimouse antibody can be produced by patients who are regularly exposed to mice as pets, by people whose occupation is working with mice in a pet store, or by humans who are exposed to foods or other items contaminated by mice.⁷ To avoid this lab test interference, another assay method should be chosen, or alternatively, employ an immunoassay that does not rely on mouse monoclonal antibody.¹³

IDENTIFYING DRUG INTERFERENCES

Incidence of Drug Interferences

The true incidence of drug interferences with laboratory tests is unknown. This is because many situations probably go undetected. However, as the number of laboratory tests and drugs on the U.S. commercial market increase, it is likely that the number of cases of in vivo interferences will also increase.

As a reflection of this, consider the number of drug–laboratory test interferences reported by D. S. Young, author of one of the classic literature references on this topic. In the first edition of Effects of Drugs on Clinical Laboratory Tests, published in the journal Clinical Chemistry in 1972, 9000 such interactions were included.¹⁴ In the second edition of the same publication, which was published in 1975, 16,000 such interactions were reported.¹⁵ In 2007, this resource, which had been converted to an online searchable database, included over 135,000 interactions.

As for in vitro interferences, the number of drug–laboratory test interferences may be moderated over time because of newer, more specific laboratory test methodologies that minimize cross-reactions with drug metabolites or drug effects on reagents or laboratory reactions.^12,16 In addition, manufacturers of commonly used laboratory equipment systematically study the effects of drugs on assay methods.¹⁷ Therefore, this information is often available to clinicians who confront problematic laboratory test results in patients. This increased awareness reduces the number of patients who are believed to have experienced newly reported drug–laboratory test interferences.

Suspecting a Drug Interference

A clinician should suspect a drug–laboratory test interference when an inconsistency appears among related test results, or between test results and the clinical picture. Specifically, clinicians should become suspicious whenever

1. Test results do not correlate with the patient’s signs, symptoms, or medical history.
   
2. Results of different tests—assessing the same organ anatomy or organ function or the drug’s pharmacologic effects—conflict with each other.
   
3. Results from a series of the same test vary greatly over a short period of time.
   
4. Serial test results are inconsistent.

No Correlation with Patient’s Signs, Symptoms, or Medical History

As emphasized elsewhere in this book, when an isolated test result does not correlate with signs, symptoms, or medical history of the patient, the signs and symptoms should be considered more strongly than the test result. This rule is particularly true when the test result is used to confirm suspicions raised by the signs and symptoms in the first place or when the test results are being used as surrogate markers or indirect indicators of underlying pathology.

For example, serum creatinine is used in various formulae to approximate the glomerular filtration rate, which is used to assess the kidney’s ability to make urine. However, actual urine output and measurement of urinary creatinine excretion is a more accurate method of assessing overall renal function. If a patient’s serum creatinine has increased from a baseline of 1–5 mg/dL over a 3-day period, but the patient has had no change in urine output, urinary creatinine excretion, or serum electrolyte levels, then the serum creatinine level may be elevated because of a drug interference with the laboratory test. Similarly, if a patient has a total serum bilirubin of 6 mg/dL, but the patient is not jaundiced or does not have scleral icterus, then a drug interference with the laboratory test should be considered.

Conflicting Test Results

Occasionally, pharmacological or toxicological effects of a drug produce conflicting results of two tests that assess the same organ function. For example, a presurgical test screen shows a serum creatinine of 4.2 mg/dL in an otherwise healthy 20-year old patient with a BUN of 8 mg/dL. Usually, if a patient had true renal impairment, BUN and serum creatinine would likely be elevated in tandem. Thus, in this patient, a drug interference with the laboratory test is suspected. Further investigation revealed that the patient received cefoxitin shortly before blood was drawn for the lab test. Cefoxitin can falsely elevate serum creatinine concentrations. Thus, the elevated serum creatinine is likely due to drug interference with the laboratory test and not to renal failure. To confirm that this is the case, cefoxitin should be discontinued and the serum creatinine repeated after that. If due to the drug, the elevated serum creatinine should return to the normal range.¹⁸

Varying Serial Test Results Over a Short Time Period

Typically, the results of a specific laboratory test should follow a trend in a patient. However, in the absence of a new onset of medical illness or worsening of existing disease, a sudden change in the laboratory test result trend should cause examination of a possible drug interference with a laboratory test. For example, prostate specific antigen (PSA) is a tumor marker for prostate cancer. It is produced by glandular epithelial cells of the prostate. Whereas the normal serum level is less than 4 ng/mL in a patient without prostate cancer, the level is typically elevated in patients with prostate cancer. However, it is not specific for prostate cancer. Elevated PSA serum levels are also observed in patients with benign prostatic hyperplasia, prostatitis, or following instrumentation of the prostate. A 70-year-old male patient with stage T₃ (locally invasive) prostate cancer has a PSA of 20 ng/mL and has decided to undergo no treatment. Four serial PSA tests over the course of 1 year and done at 3-month intervals show no change. Despite the absence of any changes on pelvic computerized axial tomography, bone scan, or chest x-ray, his PSA is 10 ng/mL at his most recent office visit. After a careful interview of the patient, the urologist discovers that the patient has been treated for alopecia for the past 6 months with Propecia^® (finasteride). The patient received the prescription from another physician. It is likely that his use of Propecia^® caused the decrease in PSA.¹⁹

Changing Serial Test Results Which Are Inconsistent with Expected Results

Leuprolide, a luteinizing hormone-releasing hormone (LHRH) agonist, is useful in the management of prostate cancer, which is an androgen dependent tumor. Persistent use of leuprolide causes down-regulation of pituitary LHRH receptors, decreased secretion of luteinizing hormone, and decreased production of testicular androgens. A patient with prostate cancer, treated with leuprolide, should experience a sustained reduction in serum testosterone levels from normal (300–1200 ng/dL) to castration levels (less than 50 ng/dL) after 2–3 weeks. The serum testosterone level should remain below 50 ng/dL as long as the patient continues treatment with leuprolide, making sure that he makes visits to the clinic for repeated doses on schedule. However, one of the adverse effects of leuprolide is decreased libido and erectile dysfunction, which is a direct extension of the drug’s testosterone-lowering effect. Such a patient may seek medical treatment of sexual dysfunction, and he may be inappropriately prescribed depot testosterone injections. Thus, in this case, depot testosterone injections will cause a change in serum testosterone levels in the wrong direction. If serum testosterone levels increase, this should be a signal that the patient has serial test results, which are inconsistent with expected results of leuprolide, and an investigation should be done as to the cause.²⁰

MANAGING DRUG INTERFERENCES

When a drug is suspected to interfere with a laboratory test, the clinician should collect appropriate evidence to confirm the interaction. Important information includes

1. Establishing a temporal relationship between the change in the laboratory test and drug use and ensuring that the change in the laboratory test occurred after the drug was started or after the drug dose was changed
   
2. Ruling out other drugs as causes of the laboratory test change
   
3. Ruling out concurrent diseases as causes of the laboratory test change
   
4. If possible, discontinuing the causative agent and repeating the test to see if dechallenge results in correction of the abnormal laboratory test
   
5. Choosing another laboratory test that will provide assessment of the same organ’s function but is unlikely to be affected by the drug (the clinician can conduct the lab test, compare the results against the original lab test result, and check for dissimilarity or similarity of results)³
   
6. Finding evidence in the medical literature that documents the suspected drug–laboratory test interference
   
7. Contacting the head of diagnostic labs who maintains or has access to computerized lists of drugs that interfere with laboratory tests (the person would also provide assistance in interpreting aberrant laboratory test results)¹³

For any particular patient case, it is often not possible to obtain information on all seven of the above items. However, the first four items are crucial in any suspected drug–laboratory test interference.

LITERATURE RESOURCES

A systematic search of the medical literature is essential for providing the appropriate background information to address steps 2–6. This search will ensure that a complete and comprehensive review—necessary in making an accurate diagnosis—has been done. When searching the literature, it is recommended to use the method originally described by Watanabe et al. and, subsequently, modified by C. F. Kirkwood.^21,22 Using this technique, the clinician would search tertiary, secondary, and then primary literature.

Tertiary literature includes reference texts and monograph databases, which provide appropriate foundational content and background material essential for understanding basic concepts and historical data relevant to the topic. Secondary literature, functioning as a gateway to primary literature, includes indexing and abstracting services (i.e., PubMed and International Pharmaceutical Abstracts). Primary literature includes case reports, experimental studies, and other nonreview types of articles in journals about the topic. These represent the most current literature on the topic. By systematically scanning the literature in this order, the clinician can be sure to have identified and analyzed all relevant literature, which is crucial in developing appropriate conclusions for these types of situations.

Tertiary Literature

Tertiary literature, which contains useful information about drug–laboratory test interferences, includes the Physicians’ Desk Reference. Each complete package insert included in this book contains a precautions section that includes information on drug–laboratory test interferences. However, it is important to note that the Physicians’ Desk Reference does not include package inserts on all commercially available drugs, nor does it include complete package inserts for all of the products included in the text. Thus, additional resources will need to be checked. The drug monographs in the AHFS Drug Information, published by the American Society of Health-System Pharmacists, also include a section on lab test interferences. Although the information provided is brief, it can be used as an initial screen. Meyler’s Side Effects of Drugs, although well-referenced, includes extensive information on adverse effects of medications that are associated with in vivo interference with laboratory test results. Information on various drug categories is subdivided among a number of texts that focus on antimicrobial agents, anesthetics, antineoplastic agents, immunobiologics, endocrine and metabolic drugs, psychiatric drugs, and cardiovascular agents. To do a complete search through Meyler’s Side Effects of Drugs, the clinician may have to go through several texts. Also, unlike the other two texts, Meyler’s Side Effects of Drugs is usually only available in drug information centers or medical libraries because of its expense.

One of the most comprehensive compilations of drug–laboratory test interactions is Effects of Drugs on Clinical Laboratory Tests by D. S. Young. Although originally published for many years as a special issue of the journal, Clinical Chemistry, it is available as a textbook, which can be purchased from the American Association of Clinical Chemistry. The last and fifth edition of the textbook was published in 2000. The content is searchable by the name of the laboratory test; specific drug, herb, or disease name; preanalytic variable; type of body fluid specimen; and specific laboratory test abnormality. Search results include a description of the drug–laboratory test interaction and a reference citation. No detail is provided on the dosage of drug that produced the interference. Hence, a clinician will need to obtain the original references and evaluate the data independently as a separate step.

MINICASE 2

Ciprofloxacin-Induced Hypoprothrombinemia

WILLIAM R., A 65-YEAR-OLD MALE PATIENT, is started on ciprofloxacin 500 mg by mouth twice daily for chronic prostatitis due to E. coli. Antibiotic treatment will continue for 6 months. William R. has atrial fibrillation and is also taking digoxin 0.125 mg by mouth daily and warfarin 2.5 mg by mouth daily. William R. has been on warfarin for years and says that he is fully aware of all the DOs and DON’Ts of taking warfarin. His international normalized ratio (INR) regularly and consistently is 2.5, which is therapeutic. William R. has no history of liver disease and appears healthy and well-nourished. Prior to the start of the antibiotic, William R.’s serum sodium was 137 mEq/L, potassium 4.0 mEq/L, BUN 10 mg/dL, creatinine 1 mg/dL, and INR is 2.5. After 3 days of antibiotics, a repeat INR is 5, and William R. complains of slight gum bleeding when he brushes his teeth.

Question: What do you think is causing the laboratory abnormality? How should William R. be managed?

Discussion: Ciprofloxacin inhibits cytochrome 1A2, the principal hepatic enzyme that catabolizes warfarin, and decreases vitamin K–producing bacteria in the gastrointestinal tract. A search of the medical literature documents multiple cases of enhanced warfarin effect when ciprofloxacin is taken concurrently.²³ Although some other antibiotics in the same pharmacologic class of quinolones may have less inhibitory effect on this enzyme, the interaction may still occur.

In William R., the drug interaction occurred after ciprofloxacin was started. William R. is not taking any other medications that could cause the drug–laboratory test interaction and has no history of vitamin K deficiency or liver disease, which could be causing hypoprothrombinemia. To confirm that ciprofloxacin is causing the drug interaction, the physician could discontinue the drug and then see if William R.’s INR returns to the range of 2–3. However, since the ciprofloxacin–warfarin interaction is well-known, a better approach might be to continue ciprofloxacin, hold warfarin until the INR has decreased to 2.5, and then resume warfarin at a reduced daily dose.

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Like this:

Like Loading...

Related

Related posts:

1. Substance Abuse and Toxicological Tests
2. Lipid Disorders
3. Electrolytes, Other Minerals, and Trace Elements
4. Infectious Diseases

Stay updated, free articles. Join our Telegram channel

Join