"Dr Alexander, we have an order for a STAT West Nile test." "You have what?" I replied, not wanting to hear the answer. After all, it was a hot July Friday, about 1:00 PM, and I had just returned to the Pathology department from giving a noon conference to residents about laboratory testing for neurological conditions. I was planning on taking the afternoon off as part of a long weekend with the family. Our departure would have to wait. "It's a STAT West Nile PCR and antibody test," the medical technologist said.
"All right," I said. "Give me the patient's and physician's names."
The resolution of this case is interesting, and I will provide the specifics later in this article. The important question, whether the order that came down for both polymerase chain reaction (PCR) and antibody testing was appropriate, is the basis for this article. Molecular and serological assays are currently available for many infectious organisms, particularly viruses, but availability is not always a good reason for ordering a test. For the purpose of this article, I will define a molecular test as a procedure used to identify a nucleic acid, either DNA or RNA, and a serological procedure as an assay which can detect a protein, be it an antibody or antigen. When should one use molecular testing? When should one use serological testing? Is there ever a situation when both are recommended? This article will explore those issues for 3 infectious diseases (IDs)-human immunodeficiency virus (HIV) infection, Lyme disease, and West Nile infection.
GENERAL PRINCIPLES OF SEROLOGICAL TESTING
Serological (antibody) testing for ID diagnosis has been available since the early and middle part of the past century. From syphilis to Legionella pneumophila to HIV, antibody testing provided the primary laboratory diagnostic method to determine exposure to organisms that are difficult to culture. Because both specificity and sensitivity increased with better reagents and methods, antibody testing became the method of choice for both diagnosis and prognosis of many diseases. Unfortunately, serological testing is based upon the ability of an individual to mount an effective humoral immune response to a pathogen, and this response takes time to develop, usually weeks to months. It is also difficult to use serological methods to differentiate between a current infection and a past exposure. Immunoglobulin (Ig) M testing is often used to document a current infection because IgM is the earliest antibody response to infection and then converts to an IgG response. The IgM-specific assays are not available for most bacterial diseases; however, even in situations where IgM testing is available, such as for Mycoplasma pneumoniae or Toxoplasma gondii, interpretation of results is complicated by the fact that specific IgM may be present long after an acute infection.1,2 The criterion standard for serological identification of an infection remains the testing of acute- and convalescent-phase specimens and demonstration of a 4-fold increase in titer; however, such paired specimens are extremely rare in the real world of clinical laboratory testing.
The predominant means of antibody testing is the enzyme-linked immunosorbent assay (ELISA) test. For detection of antibody, microbial antigen is bound to plastic wells of a plate; upon addition of patient serum, specific antibody will be "immunosorbed" to the plate. The presence of the human antibody is detected by the binding of an enzyme-conjugated antihuman antibody to the well and conversion of its substrate to color. The use of a standard curve allows this technique to be quantitative. The Western blot technique, often used to confirm results of an ELISA reaction, analyzes the specificity of the patient's antibody. Microbial proteins, such as from HIV, are separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and then transferred to a sheet of special filter paper. After incubation with patient's serum and washing, an enzyme-conjugated antibody is added, followed by its substrate. Positive reactions are indicated by colored or luminescent protein bands corresponding to the molecular weights of the microbial antigens.
GENERAL PRINCIPLES OF MOLECULAR TESTING
Molecular techniques detect and identify the infecting organism by the presence of its DNA or RNA. Detection of microbial DNA, such as from a virus or bacteria, can be accomplished by the PCR. The use of specific DNA sequences from unique portions of the microbial genome as primers allows specific amplification of the microbial DNA, followed by detection on an agarose gel. Quantitation of the DNA sequence can be obtained by PCR methods or the branched chain DNA assay. The branched chain DNA assay is a signal amplification method, whereas PCR is a target amplification method. In branched chain DNA, nucleic acid sequences homologous to the microbial DNA or RNA are affixed to wells of a plate. Binding of the pathogen's nucleic acid from the patient's sample to the affixed DNA oligomers is detected by the addition of a specific probe coupled with an artificially branched chain of DNA, having many identical sticky ends. An enzyme-labeled DNA probe which is complementary to the sticky ends is added, followed by the substrate of the enzyme to provide a signal that can be quantitated in a manner similar to an ELISA test. As a result, each piece of microbial DNA will be identified by many sticky ends and an amplification of the ELISA-like signal.
The RNA, as for an RNA virus such as HIV, can be detected by reverse-transcriptase PCR (RT-PCR) and quantitated by real-time PCR (qPCR) or branched chain DNA assay. For RT-PCR, RNA from the sample is converted to DNA using a retrovirus reverse-transcriptase polymerase and the DNA is amplified by PCR. For qPCR, the rate of DNA amplification is monitored as an indication of the initial concentration of sample DNA. In qPCR, the DNA concentration is indicated by the release of a fluorescent molecule during the amplification reaction (TaqMan; Roche Diagnostics, Basel, Switzerland) or by the quantitative binding of a fluorescent molecule to the DNA.
Molecular tests provide several advantages over serological tests. They detect the microbe instead of a serological history of the infection and are usually able to detect the presence of a microbe much earlier in the course of an infection. The empirical sensitivity of most molecular methods, especially nucleic acid amplification assays, also tends to decrease the time from infection to a positive assay. This increased sensitivity is not without cost, however, as false-positive amplification assays may occur.3 In some diseases, it is also difficult to obtain an appropriate specimen for molecular studies. If a pathogen is not routinely found in a blood or urine specimen, molecular methods may not be the most appropriate assay. Finally, because most infections are completely cleared by an individual's innate and antigen-specific immune response, molecular testing is not a useful procedure to document past exposure.
With the above as an introduction, the basis for using the different molecular and serological applications for 3 specific clinical pathogens, HIV, Lyme disease, and West Nile virus (WNV), will be reviewed. Having discussed the logic for choosing different tests for these 3 pathogens, we will conclude with the resolution of the incident presented in the opening paragraph.
Human Immunodeficiency Virus
The HIV diagnosis uses both molecular and serological testing. The principal assay used for HIV diagnosis is the HIV antibody test; an ELISA is followed by a Western blot confirmation of positive specimens. However, the antibody test may take up to 4 weeks to become positive after an exposure or infection.4 Specific IgM anti-HIV antibody testing has not been shown to be useful in diagnosing early infection. In some clinical situations, however, it is useful to have the diagnosis of HIV infection before the antibody test becomes positive. These situations include blood or organ donor screening and neonatal testing of babies born to HIV-positive mothers. Testing the neonate for antibody will yield results that parallel the mother's antibody status. The HIV ELISA assays are designed to detect either IgG antibodies or an IgG/IgM combination. Thus, infants born to an HIV-positive mother will always test positive in an HIV antibody test immediately after birth. Detection of the HIV p24 antigen has been used for neonatal diagnosis; however, molecular methods have yielded greater sensitivity. The 2 molecular methods that can be used are the HIV qualitative DNA PCR or the quantitative HIV viral load procedure. The DNA PCR detects HIV that has integrated into the host genome. The viral load assay measures the amount of free virus present in the plasma. The viral load testing can be performed using either the ultrasensitive RT-PCR assay or the branched chain DNA assay, either of which has the sensitivity to detect low levels of HIV that may be present in the neonate.
The HIV causes a chronic disease, and monitoring the rise or fall of antibody titers does not provide an indication as to the disease state of the patient. The HIV infection is staged using a combination of clinical parameters and the absolute CD4 T-cell count. Treatment guidelines show the CD4 T-cell count and the HIV viral load (amount of virus present per milliliter of plasma).5 Most laboratories performing CD4 counts use a flow cytometric technique for the assays. Although most laboratories use a "single-platform" flow cytometry technique for determining absolute CD4 counts, some laboratories use what is known as a dual-platform technique. The dual-platform method combines results from both a flow cytometer and a hematology analyzer to determine the absolute CD4 count. However, values are not identical between the 2 methods, and serial monitoring is best interpreted when the values are consistently obtained from a single laboratory or, at least, from laboratories using identical methods. The HIV viral load is determined by 2 major methods, RT-PCR and branched chain DNA analysis. The values obtained by these 2 methods also do not agree. Interestingly, even values obtained using different RT-PCR assays may not agree, as seen in Tables 1 and 2. The Bayer branched chain DNA technique (Bayer Diagnostics, Lowerkusen, Germany) uses a single assay to cover a linear range between 75 and 500,000 HIV-1 copies/mL. The standard Roche RT-PCR assay (Roche Diagnostics, Basel, Switzerland) is linear from 400 to 750,000 copies/mL, whereas the Roche Ultrasensitive PCR assay is calibrated to report out values from 50 to 100,000 copies/mL. Values obtained by College of American Pathologists (CAP)-accredited laboratories' testing of recent CAP survey specimens are shown in Table 1. Similar values obtained from Centers for Disease Control and Prevention (CDC) specimens are shown in Table 2.
Serology can be an aid in the diagnosis of Lyme disease; however, the antibody response may not occur until weeks after an infection, if at all.6 Specific IgM Lyme disease antibodies have limited use in diagnosing an early infection. Most Lyme antibody screening assays are IgM/IgG combination ELISAs or rapid tests. These assays detect both IgG and IgM anti-Borrelia burgdorferi antibodies without differentiating between the 2 isotypes. The CDC recommends that reactive Lyme antibody ELISA tests be confirmed with a Western blot.7 Specific IgM and IgG Lyme Western blots are available, and interpretative guidelines are different for each blot. The IgM blot must demonstrate antibody recognition of 2 specific B. burgdorferi protein bands to be considered positive. The IgG blot must identify 5 protein bands to be considered positive.8 The antigenic cross reactivity among the various spirochete species is the primary reason for the multiple-band requirements. Molecular techniques, particularly PCR, have limited use in the diagnosis of Lyme disease. The principal problem is determining the type of specimen to be analyzed. There are conflicting reports as to whether an infected individual would have spirochetemia at a time when symptoms would be present, leading to diagnosis. Thus, blood is a poor specimen for Lyme PCR diagnosis.
Laboratory monitoring of Lyme disease is an inexact science. Antibody levels do not decline with treatment, and the level of antibody does not correlate with disease activity. If B. burgdorferi is detected by a molecular assay, then that procedure may be repeated after the treatment to determine if the organism is still present in the particular tissue or fluid in which the organism was originally detected. However, obtaining a follow-up biopsy or a cerebrospinal fluid (CSF) specimen is not routinely performed.
West Nile Virus
The WNV infection became a clinical concern in the western hemisphere at the turn of the century. Diagnosis initially was performed at autopsy using molecular techniques. The PCR was the initial diagnostic regimen performed until antibody assays became available. The antibody assays provide a less expensive and highly sensitive alternative to PCR for diagnosis. The WNV-specific IgM in the spinal fluid of infected patients has been reported to be the most sensitive indicator of West Nile encephalitis.9 False-positive West Nile antibody results have been reported; however, the currently used third generation WNV ELISA assay has a false-positive rate of less than 1%.10 The use of a control antigen to identify serological cross reactivity has the potential to eliminate the false-positive assays, thus providing a specificity of 100% in 1 study.10
No serological or molecular assay is currently available for monitoring the disease activity of West Nile-infected patients. However, assaying for WNV IgG avidity can, at least, provide a relative time indicator of infection. The IgG avidity is a measurement of the strength of antibody-antigen binding. Avidity increases during an immune response. Thus, measuring IgG avidity can provide a relative indicator of time of infection.11 Increasing avidity levels between acute- and convalescent-phase specimens point to a recent infection. High avidity levels detected in a single specimen point toward a past exposure or infection, whereas lower avidity levels point to a more recent infection. Avidity may be determined by adding a dissociation reagent, such as urea, to an immunofluorescent or ELISA assay. The dissociation reagent will break weak antigen-antibody bonds. An avidity index, related to the amount of reactivity lost with the addition of the dissociation reagent, is calculated and used in the interpretation of the assay.
Remember the STAT West Nile PCR and antibody test order that I received? My first thought was this-why did the physician want both molecular and serological assays? As previously noted, the most sensitive indicator of West Nile encephalitis is specific IgM testing of the CSF. I called the physician in question, who was an endocrinologist, to discuss the patient's condition and to inform her of the use of both the West Nile PCR and IgM test. Her patient was a young lady with type 1 diabetes who had presented unresponsive. The lady had been camping with her family the previous week but not in an area where WNV had been reported to be present. The physician stated that she needed to be able to confirm that this was West Nile, so she could inform the family before the patient died. Our institution does not perform in-house West Nile testing; hence, I told her that there was no way to obtain a "STAT" result. Clinically, a STAT test was not indicated because the result would not affect immediate treatment. I told her that I could have an antibody report from the reference laboratory late on the following Monday. An ID physician, who was consulted in the meantime, did not believe that the patient had been exposed to WNV. Nonetheless, a CSF specimen was obtained and referred for West Nile antibody testing. The antibody test was negative, supporting the ID physician's opinion. However, upon the insistence of the endocrinologist, a subsequent WNV PCR test was performed on the CSF and yielded a negative result. The patient went on to make a full recovery. As a follow-up, the physician ordered WNV serology on serum obtained approximately a month after the initial presentation. Interestingly, the serum IgM anti-WNV ELISA assay showed a low-positive result; the result of the serum IgG anti-WNV test was negative. However, further testing showed that the patient's IgM reactivity was caused by a cross-reacting antibody and not because of WNV-specific IgM.
Was the physician correct in ordering this entire battery of WNV assays? As previously mentioned, the most sensitive indicator of active WNV infection is IgM antibody in the CSF. The PCR may be used to confirm a positive IgM assay but does not offer increased sensitivity beyond the ELISA test. Thus, there was no basis for ordering the PCR assay. The follow-up serum assay could be useful because most individuals infected with WNV will become serum antibody positive, and obtaining serum is less invasive than obtaining CSF.
So, now you can ask and answer the question: serology or molecular diagnosis?
1. Nir-Paz R, Michael-Gayego A, Ron M, et al. Evaluation of eight commercial tests for Mycoplasma pneumoniae
antibodies in the absence of acute infection. Clin Microbiol Infect
2. Gorgievski-Hrisoho M, Germann D, Matter L. Diagnostic implications of kinetics of immunoglobulin M and A antibody responses to Toxoplasma gondii
. J Clin Microbiol
3. Rich JD, Merriman NA, Mylonakis E, et al. Misdiagnosis of HIV infection by HIV-1 plasma viral load testing: a case series. Ann Intern Med
4. Weber B, Fall EH, Berger A, et al. Reduction of diagnostic window by new fourth-generation human immunodeficiency virus screening assays. J Clin Microbiol
5. Hammer SM, Saag MS, Schechter M, et al. Treatment for adult HIV infection: 2006 recommendations of the international AIDS Society-USA panel. JAMA
6. Steere AC. Lyme disease. N Engl J Med
7. CDC. Recommendations for test performance and interpretation from the Second National Conference on Serologic Diagnosis of Lyme Diseases. MMWR Morb Mortal Wkly Rep
8. Engstrom SM, Shoop E, Johnson RC. Immunoblot interpretation criteria for serodiagnosis of early Lyme disease. J Clin Microbiol
. 1995;33: 419-422.
9. Prince HE, Lape-Nixon M, Moore RJ, et al. Utility of the Focus Technologies West Nile virus immunoglobulin M capture enzyme-linked immunosorbent assay for testing cerebrospinal fluid. J Clin Microbiol
10. Hogrefe WR, Moore R, Lape-Nixon M, et al. Performance of immunoglobulin G (IgG) and IgM enzyme-linked immunosorbent assays using a West Nile virus recombinant antigen (preM/E) for detection of West Nile virus-and other flavivirus-specific antibodies. J Clin Microbiol
11. Fox JL, Hazell SL, Tobler LH, et al. Immunoglobulin G avidity in differentiation between early and late antibody responses to West Nile virus. Clin Vaccine Immunol