Significance of Isolated Positive IgM Serologic Results by Enzyme Immunoassay for Coccidioidomycosis
- First Online:
- Cite this article as:
- Blair, J.E. & Currier, J.T. Mycopathologia (2008) 166: 77. doi:10.1007/s11046-008-9129-9
- 109 Downloads
Serologic testing is important for diagnosis of coccidioidomycosis. Many methods are available for diagnostic testing. Enzyme immunoassay (EIA) can be performed quickly and locally but has the potential for false-positive results in patients manifesting a positive EIA for immunoglobulin M (IgM) antibodies and a negative EIA for immunoglobulin G (IgG). We retrospectively reviewed the charts of 405 patients with coccidioidal serologic testing performed between 1999 and 2003. Of 706 EIAs, 37 (5%) produced test results for 28 patients that showed isolated IgM positivity. Among these 28 patients, 24 (86%) had positive serologic findings by other methods (complement fixation or immunodiffusion or both), and 7 (25%) had positive microbiologic or histopathologic findings. All 4 (14%) patients without other positive serologic results had diagnostic tests with positive microbiologic or histopathologic results. No false-positive IgM assays were observed. We conclude that the false-positive rate of the EIA IgM is low, and that an isolated positive EIA IgM should prompt further follow-up and diagnostic testing.
KeywordsCoccidioides Immunoenzyme techniques False-positive reactions Predictive value of tests
Arizona Department of Health Services
Human immunodeficiency virus
Coccidioidomycosis is an endemic fungal infection found primarily in the southwestern US, northern Mexico, and limited areas of Central and South America and is acquired from inhalation of airborne arthrospores of Coccidioides sp [1, 2]. Approximately 60% of infections are asymptomatic, and most of the remaining 40% result in a respiratory illness that ranges in severity from mild to life threatening . Depending on the population studied, extrapulmonary infections are identified in less than 5% of symptomatic patients .
Serologic studies have become the mainstay in diagnosing and managing patients with coccidioidomycosis. Although serologic tests for coccidioidomycosis are described as the most reliable among those for mycoses , their sensitivity is variable, depending on the type of serologic test and its timing with respect to the onset of symptoms .
Within the endemic area, the enzyme immunoassay (EIA) is useful as a sensitive, qualitative test with a rapid turnaround time, because it can be performed by local laboratories, whereas complement fixation (CF) and immunodiffusion (ID) tests require that the serum specimen be sent to a reference laboratory. However, questions have arisen about the routine use of the EIA serologic test. Studies comparing the EIA immunoglobulin M (IgM) to standard methods have not only shown that it demonstrates superior sensitivity in very early infections [5, 6] but have also raised concern that the EIA IgM has low specificity and thus requires confirmation with another serologic test . The reputation of the EIA IgM as a low-specificity assay makes it difficult to accurately interpret an EIA that is both IgG negative and IgM positive.
We reviewed our experience with the diagnosis of coccidioidomycosis using the EIA method. Then we correlated the EIA results that were both IgM positive and IgG negative with the strength of the coccidioidal diagnosis. This paper presents the results of this investigation and calls into question the advisability of disregarding as false-positive any EIA findings that are IgM positive and IgG negative within an endemic area.
A retrospective chart review was conducted for all patients with any positive coccidioidal serologic result obtained between January 1, 1999, and December 31, 2003, and reported to the Arizona Department of Health Services (ADHS). Patients whose positive serologies were limited to an isolated EIA IgM positivity or an indeterminate result were included in this review. Only patients with an onset of symptoms after January 1, 1998, and with adequate documentation of clinical illness were included. Patients were identified by a manual search of the reports made to the ADHS. To ascertain that no cases with a positive serology were missed, we also used a computerized search of the ICD-9 (International Classification of Diseases, Ninth Revision)  codes for coccidioidomycosis. This study was approved by the Mayo Foundation Institutional Review Board.
Abstracted information included age, sex, race, comorbid conditions, dates, and results of serologic tests for coccidioidomycosis, microbiologic tests, radiographs, and histopathologic tests. The clinical symptoms, location, treatment, and follow-up of the coccidioidal illness were also abstracted.
A patient was defined as having confirmed coccidioidomycosis if Coccidioides sp were isolated from any culture or if spherules of Coccidioides were identified histologically.
Highly Probable Coccidioidomycosis
A patient was defined as having highly probable coccidioidomycosis if the illness included a compatible clinical syndrome (including but not limited to fever, cough, headache, rash, malaise, or myalgia) AND positive serologic results AND a compatible radiographic image. A diagnosis of highly probable coccidioidomycosis required serologic results other than an isolated positive EIA IgM.
A patient was defined as having probable coccidioidomycosis if there was a strong, positive serologic finding (positive by at least 2 methods [an isolated EIA IgM positive result did not count as 1 positive method], or if the CF alone was ≥1:8), AND if the patient had EITHER but NOT BOTH the radiographic abnormalities and symptoms that are needed to achieve a diagnosis of “highly probable coccidioidomycosis.”
Neither Rejected nor Confirmed Coccidioidomycosis
The patient met minimal criteria of a positive test for report to ADHS, but did not meet criteria for a diagnosis of confirmed, highly probable, or probable coccidioidomycosis and had no alternate diagnosis.
When histopathologic or serologic results identified the presence of coccidioidomycosis without any identifiable symptomatic illness, patients were diagnosed with asymptomatic coccidioidal infection. Examples of these situations included lung nodule resections for suspected cancer or serologic screening for transplantation evaluation.
Our local laboratory performed the EIA to detect IgM and IgG antibodies using an assay kit from Meridian Bioscience, Inc (Cincinnati, Ohio). Positive, indeterminate, and negative results were defined in accordance with the manufacturer’s recommendations (positive = absorbance value ≥0.200; indeterminate = absorbance value ≥0.150 but ≤0.199; negative = absorbance value <0.150).
CF and ID studies were performed at the Mayo Clinic Infectious Diseases Serology Laboratory, Rochester, Minnesota. The Laboratory Branch CF test of the Centers for Disease Control and Prevention, as described previously [4, 9], was used to detect IgG antibodies. D. Pappagianis of the University of California at Davis provided the antigen for the CF test. Between January 1, 1999, and May 31, 2002, the ID test was performed using an assay kit from Gibson Laboratories, Inc (Lexington, Kentucky) that used antigen F to detect IgG antibodies. Beginning June 1, 2002, the ID test was performed using a Meridian Biosciences kit to detect both TP (IgM, early) antibodies and F (IgG, late) antibodies. Infrequently, serum samples were sent to the external laboratory of D. Pappagianis for confirmation of test results.
A total of 405 patients (357 symptomatic patients, 48 asymptomatic) with any positive coccidioidal serologic findings and with adequate chart information were identified and included for study. In all, 706 EIA, 797 CF, and 797 ID tests were performed among these 405 patients.
Among the 28 patients with 37 EIA tests showing a positive IgM and a negative IgG, a repeat EIA performed on serum drawn on a subsequent date resulted in 5 seroconversions, such that the repeat serologic test resulted in identifying a positive IgG. Further review of the charts of these 28 patients revealed confirmed coccidioidomycosis in 7 patients (25%), highly probable coccidioidomycosis in 18 patients (64%), and probable infection in 3 patients (11%). None of the 28 patients with EIA negative IgG and positive IgM had a rejected diagnosis. A total of 24 of the 28 patients had 1 or more concurrent or subsequent positive serologies, either by subsequent EIA seroconversion (5 of 28; 18%) or by concurrent positive CF (16 of 28; 57%) or ID (8 of 28; 29%) results, which left 4 (14%) patients without positive serologic findings on the initial or subsequent assays. A review of these 4 patients indicates that each had a diagnostic culture or biopsy confirming the diagnosis of coccidioidomycosis. None of the 28 patients had asymptomatic coccidioidomycosis.
The demographics of the 28 patients with EIA IgM positive and IgG negative findings were compared with those of the 377 patients with other serologic patterns. No differences were found in the proportions of patients in each group with respect to age, race, tobacco use, comorbid conditions (e.g., diabetes mellitus; lung, rheumatologic, or cardiovascular diseases; cancer; organ transplantation; human immunodeficiency virus [HIV] infection; or immunosuppression).
Acute coccidioidomycosis is most often characterized as a self-limited respiratory infection manifested by the nonspecific symptoms of fever, cough, headache, rash, myalgia, and fatigue . In self-limited illness, diagnostic studies are limited, because blood cultures are rarely positive and dry cough prevents many patients from being able to produce a specimen of sputum for culture or cytology. Most patients do not manifest lesions that require a diagnostic biopsy. More recently, the diagnostic utility of urinary antigens has been tested, but they are not adequately sensitive as a screening test . Therefore, serologic tests are still the mainstay for diagnosis of this infection.
Serologic tests have been performed reliably for decades and have been well described . The EIA method has been in use as an assay for coccidioidomycosis for more than a decade, and it has the advantage of ease of use, rapid results, and ability to be performed in a local laboratory. EIA testing is seemingly the most sensitive of all the assays for identification of early coccidioidal disease [6, 11]; in one study, patients with coccidioidal infection identified only by the EIA test had fewer or milder clinical abnormalities than those detected by standard testing .
The discrepancies between the results of the EIA and those of more traditional assays seem to be most marked in the detection of IgM . Previous studies have noted high sensitivity of the assay , but the specificity of the EIA IgM, when compared with that of other serologic methods, ranges from 75% to 96% [7, 12]. The EIA IgM has also shown reactivity in persons with confirmed blastomycosis and other noncoccidioidal disease . With such a range of specificity, experts in the field have cautioned that the EIA IgM test produces substantial numbers of false-positive reactions and therefore requires confirmation (e.g., by ID) .
In this study, we attempted to correlate the serologic result with the strength of the coccidioidal diagnosis. There is no dispute that an illness accompanied by a positive culture or histopathology result is coccidioidal. Likewise, in the vast majority of patients with compatible clinical illness, radiographic findings, and positive serology, the diagnosis is firm, although not to the same level of certainty. We termed these latter illnesses “highly probable coccidioidomycosis” but not “confirmed,” because of the theoretical possibility that the current illness is due to another cause and the serology is a carryover from a separate recent infection with Coccidioides sp. The strength of the coccidioidal diagnosis is much more difficult when fewer of the criteria (clinical, serologic, or radiographic) are present. The criteria for “probable coccidioidomycosis” were created to include persons with some positive serology but without either clinical or radiographic findings. The problems we faced in classifying this group relate to the fact that a patient could be in any phase of resolved coccidioidomycosis (with waning but not resolved serology) but may actually have an illness due to another cause. This group was created to show that coccidioidomycosis is still a likely diagnosis, but not with the same degree of certainty.
Our findings indicate that coccidioidal illness accompanied by an EIA positive IgM and an EIA negative IgG was observed in 37 (5%) of 706 EIA assays performed between 1999 and 2003. However, when we further evaluated the 28 patients represented by those 37 studies, 7 (25%) had confirmed infection by microbiologic or histopathologic results, and 24 (86%) had positive serologic results by other methods (subsequent EIA, CF, or ID). All 4 (14%) of the 28 patients without other positive serologic findings went on to have a diagnostic test that confirmed the presence of infection with Coccidioides sp, either by histologic results or by growth of the organism in the microbiology laboratory. Therefore, in this study, there were no false-positive EIA IgM results. In addition, there were no identified clinical or demographic characteristics of patients that increased the risk of isolated EIA IgM positivity.
A clinical study by Crum et al.  found that 48 patients (representing 18% of all positive EIA IgM assays) with positive EIA IgM had false-positive results, defined by the fact that these patients did not develop a clinical illness. Whether the patients had other serologic, microbiologic, histologic, or radiographic studies was not reported, nor was the frequency of follow-up evaluations reported. In that study, the false-positive rate was found to be higher among asymptomatic persons undergoing surveillance, African Americans, and persons with immunocompromised conditions such as HIV infection . These results stand in contrast to our findings of no false-positives, even among 57 asymptomatic patients. It is difficult to compare the other risks identified in the study by Crum et al. , because our study population did not contain a large percentage of African Americans or HIV-infected persons.
Despite this difference in findings, it is still the impression of the senior investigator of this report (J.E.B.) and her clinical colleagues that in the years since the conclusion of this study in 2003, a patient may infrequently manifest a mild illness similar to coccidioidomycosis that is associated with isolated EIA IgM positivity but not confirmed by subsequent serologic, microbiologic, or histologic studies. This situation may reflect a low rate of false-positive tests, but an alternate explanation is that very mild coccidioidal illnesses are not always accompanied by a positive serologic test. Although a different antibody was measured, this concept was demonstrated more than 50 years ago by Charles Smith , who found only 7% seropositivity in a sample of patients with asymptomatic skin test conversion but no clinical manifestations of infection.
Parenthetically, it should be stated that while this study focused on EIA specificity, it may appear that the 199 of 706 negative EIA IgM and IgG results are falsely negative. In this study, we reviewed 706 EIA tests from 405 patients; clearly, many patients had serial EIA serologic tests performed. In a previous examination of the serologic tests from this same cohort of patients , we found that the sensitivity of the EIA test was highest (up to 90% sensitivity) in the first 4 months after symptom onset and that it dropped off markedly thereafter. Since we evaluated all serologic results regardless of time from symptom onset in the present study, no statements can be made regarding the sensitivity of the EIA test.
In summary, numerous assays are available for the serologic diagnosis of coccidioidomycosis. Because none appear to have perfect sensitivity and specificity, the use of a combination of studies may be useful in the diagnostic process. In patients with suspected coccidioidal illness, an isolated positive EIA IgM result should trigger careful observation of the patient, with consideration of repeat serologic tests or use of another serologic method or of adjunctive diagnostic testing (culture or biopsy) as warranted by the clinical illness.