Presented in part as a poster at the annual meeting of the Coccidioidomycosis Study Group, Tucson, Arizona, March 2012.
The Utility of Diagnostic Testing for Active Coccidioidomycosis in Solid Organ Transplant Recipients
Version of Record online: 7 FEB 2013
© Copyright 2013 The American Society of Transplantation and the American Society of Transplant Surgeons
American Journal of Transplantation
Volume 13, Issue 4, pages 1034–1039, April 2013
How to Cite
Mendoza, N. and Blair, J. E. (2013), The Utility of Diagnostic Testing for Active Coccidioidomycosis in Solid Organ Transplant Recipients. American Journal of Transplantation, 13: 1034–1039. doi: 10.1111/ajt.12144
- Issue online: 30 MAR 2013
- Version of Record online: 7 FEB 2013
- Manuscript Revised: 14 DEC 2012
- Manuscript Accepted: 14 DEC 2012
- Manuscript Received: 26 JUL 2012
- organ transplantation;
- polymerase chain reaction;
- sensitivity and specificity;
- serologic tests
- Top of page
- Materials and Methods
Solid organ transplant recipients who acquire coccidioidomycosis have high rates of disseminated infection and mortality, and diagnosis of infection in these immunosuppressed patients is challenging because of suboptimal sensitivity of diagnostic tests. To characterize the utility of diagnostic tests for coccidioidomycosis in this population, we conducted a retrospective chart review of all solid organ transplant recipients with newly acquired coccidioidomycosis who were seen at our institution from 1999 to 2011. We identified 27 solid organ transplant recipients with newly acquired, active coccidioidomycosis. The positivity of any single serologic test ranged from 21% (5/24; immunoglobulin M by immunodiffusion) to 56% (14/25; immunoglobulin G by enzyme immunoassay), compared with 77% (20/26) seropositivity for a battery of serologic tests (enzyme immunoassay, immunodiffusion and complement fixation). Serology performed approximately 1 month later increased positive test findings to 92%. Culture of respiratory or tissue specimens yielded Coccidioides sp in 54% (14/26) of the cultures submitted, and 10/16 (63%) of patients tested. Chest-computed tomography was abnormal in 86% (19/22). Multiple test modalities may be needed to diagnose coccidioidomycosis in solid organ transplant recipients, and repeat studies over time may increase sensitivity of the diagnostic assays.
European Organization for Research and Treatment of Cancer
Mycoses Study Group
polymerase chain reaction
- Top of page
- Materials and Methods
Coccidioidomycosis is an endemic fungal infection of the desert southwestern United States, and its incidence in Arizona has markedly increased over the past 2 decades [1-3]. Solid organ transplant recipients are exquisitely susceptible to severe coccidioidal infection because of the iatrogenic suppression of the cell-mediated immunity required to control this infection. Indeed, coccidioidal infection in transplant patients is frequently disseminated, with high mortality rates [4, 5]. Recently, the Transplant-Associated Infection Surveillance Network reinforced the notion that endemic fungi, including Coccidioides, are important pathogens causing invasive fungal infections in organ transplant recipients .
Tests currently used to diagnose coccidioidomycosis include serology, culture, cytology, histology, molecular analysis with tools such as polymerase chain reaction (PCR), and detection of antigenuria . Small case series [8-11] of transplant recipients with coccidioidomycosis have indicated that such patients often display lower seroreactivity compared with that of immunocompetent patients ; nonetheless, little is known about the sensitivity of these and other diagnostic tests in patients who have undergone solid organ transplantation. Therefore, we conducted a retrospective review to better characterize the utility of various diagnostic tests in the evaluation of newly acquired, active coccidioidomycosis in solid organ transplant recipients.
Materials and Methods
- Top of page
- Materials and Methods
To identify patients who had undergone solid organ transplantation and who had active coccidioidomycosis, we initially constructed a list of all patients from our institution whose coccidioidomycosis was reported to the Arizona Department of Health Services between the dates of January 1, 1999, and August 31, 2011. We then cross-referenced that list with all recipients of a transplanted kidney, liver, pancreas or heart, and included in the study those patients whose posttransplant coccidioidal symptoms began within 2 months of the diagnostic evaluation. Patients were excluded who had evidence of coccidioidomycosis before transplantation or coccidioidal infection for longer than 2 months. Clinical records were subsequently examined, and the results of all coccidioidal diagnostic tests were tabulated, including: cultures (blood, respiratory secretions, cerebrospinal fluid [CSF], urine, other); histopathology and cytology (tissue or fluids); serology results (enzyme immunoassay [EIA], immunodiffusion [ID] and complement fixation [CF]); PCR; and radiology. Other abstracted information included patient demographics, features of the coccidioidal illness and transplant characteristics.
Active coccidioidal illness was defined according to the European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) Consensus Group . Briefly, proven coccidioidomycosis was defined as a case with positive culture or histopathology, or positive serology in the cerebrospinal fluid. Probable coccidioidomycosis required a consistent clinical picture and mycological evidence from urine, blood, or cerebrospinal fluid. Mycological evidence included a positive serology, PCR or antigen assay .
We used the definitions of radiographic abnormalities described previously . Chest radiograph or computed tomography (CT) scans were considered positive if they revealed signs of acute disease (segmental or lobar consolidations, multifocal nodular or patchy opacities, peribronchial thickening, hilar and/or mediastinal adenopathy, or pleural effusion) or of disseminated disease (miliary or reticulonodular pattern or pericardial effusion).
For all patients, serology was performed in a similar manner, and the specific tests used at our institution have been described previously . EIA to detect immunoglobulin M (IgM) and immunoglobulin G (IgG) was performed using a commercial kit (Meridian Bioscience Inc. Cincinnati, OH, USA), with manual pipetting and an automated plate washer. Indeterminate EIA or ID results were counted as negative. The CF and ID tests were performed at the Mayo Medical Laboratories in Rochester, Minnesota, using antigen from Dr. Pappagianis at the Coccidioidomycosis Serology Laboratory at the University of California at Davis. The Laboratory Branch methodology of the Centers for Disease Control and Prevention was used to detect IgG CF antibodies. Between January 1, 1999, and May 30, 2002, the ID test was performed using a kit from Gibson Laboratories Inc. (Lexington, Kentucky); from June 1, 2002 onward, the ID test was performed using a kit from Meridian Bioscience Inc. (Cincinnati, OH, USA).
The PCR tests were performed using a laboratory developed test, which has been described previously . Total nucleic acid was extracted using the MagNA Pure Compact automated extractor (Roche Applied Science, Roche Diagnostics Corp., Indianapolis, Indiana) with a 0.2 mL sample and a 0.1 mL elution. Amplification was performed with the LightCycler 2.0 real-time PCR platform (Roche Applied Science) using 5 microliter of sample extract in a 20 microliter total reaction.
Eosinophilia was defined as >0.4×109 cells/L, which is the upper limit of normal for our laboratory at Mayo Clinic, Scottsdale, AZ, USA.
The results of all coccidioidal diagnostic tests were tabulated using Epi Info 7 (Centers for Disease Control and Prevention, Atlanta, Georgia) and were analyzed using Microsoft Excel (Microsoft Corp, Redmond, Washington). Simple descriptive statistical comparisons were calculated. The Fisher exact test was used to compare proportions, and a P value ≤.05 was considered significant.
This study was approved by the Mayo Clinic Institutional Review Board.
- Top of page
- Materials and Methods
From January 1999 through August 2011, a total of 2246 recipients of a transplanted organ (liver, kidney, pancreas, heart, or a combination) had medical care provided at our institution. Of these, 239 had 1 or more positive coccidioidal diagnostic tests. Of the 239, most (n = 212) had evidence of coccidioidomycosis prior to transplantation, or coccidioidal infection for longer than 2 months; the rest (n = 27) had newly acquired, symptomatic, active coccidioidomycosis. The focus of the current study was the latter group of 27, whose baseline patient characteristics are presented in Table 1. No patients with heart or pancreas transplants met our case definition. All patients were on immunosuppressive medication.
|Age, mean (range), y||55 (36–74)|
|Male sex||19 (70)|
|Deceased donor||17 (63)|
|Active coccidioidomycosis definition|
|Immunosuppression at diagnosisd|
|Mycophenolate preparations||19 (70)|
|Prednisone >10 mg/day||2 (7)|
|Prednisone ≤10 mg/day||6 (22)|
A summary of the results of the serologic tests (EIA, ID and CF) is presented in Table 2. These serologic assays were typically performed on the date the patients presented medically with their fungal illness, and they were repeated approximately 1 month later. The positive rates of most of the serologic tests improved with repeat testing. Of the serologic assays, the EIA IgG was the most sensitive (overall, 56% [14/25] of the patients tested positive initially, and 64% [16/25] tested positive upon repeat testing 1 month later), and the two assays for IgM (EIA and ID) were the least sensitive (32% [8/25] and 29% [7/24] positive on the 1-month repeat test, respectively) in these patients. After two tests, most patients (92% [24/26]) who underwent serologic testing in our study had at least one positive serologic result. However, two patients with negative serologic findings on the first test did not have a second set of serologic tests; thus, we do not know if they later underwent seroconversion.
|First test||Second test|
|Type of test||No. of patients positive/no. tested (%)||No. of patients retested||Mean time from the first test, days||Cumulative no. of patients positive/no. tested (%)|
|EIA IgM (all patients)||7/25 (28)||10||31||8/25 (32)|
|Proven infection||3/11 (27)||2||29||4/11 (36)|
|Probable infection||4/14 (29)||8||32||4/14 (29)|
|EIA IgG (all patients)||14/25 (56)||10||31||16/25 (64)|
|Proven infection||5/11 (45)||2||29||5/11 (45)|
|Probable infection||9/14 (64)||8||32||11/14 (79)|
|ID IgM (all patients)||5/24 (21)||13||24||7/24 (29)|
|Proven infection||2/10 (20)||5||12||3/10 (30)|
|Probable infection||3/14 (21)||8||32||4/14 (29)|
|ID IgG (all patients)||9/24 (38)||13||24||9/24 (38)|
|Proven infection||3/10 (30)||5||12||3/10 (30)|
|Probable infection||6/14 (43)||8||32||6/14 (43)|
|CF ≥1:2 (all patients)b||7/25 (28)||14||24||9/25 (36)|
|Proven infection||1/10 (10)||5||12||1/10 (10)|
|Probable infection||6/15 (40)||9||30||8/15 (53)|
|Any positive serology (all patients)c||20/26 (77)||15||NA||24/26 (92)|
|Proven infection||7/11 (64)||5||NA||9/11 (82)|
|Probable infection||13/15 (87)||10||NA||15/15 (100)|
The quantitative titer of the CF assay was ≥1:2 in 7 of 25 patients (28%) on the first test and ≥1:2 in 9 of 25 patients (36%) on the second test. The median (range) of the initial CF titer was 1:2 (range, 1:2–1:64); for the follow-up test performed 1 month later, the median result was 1:16 (range, 1:2–1:16). One patient had anticomplementary serum both initially and at follow-up.
Specimens were collected for culture from 16 patients, as summarized in Table 3. Many patients had at least two specimens collected for culture. Twenty-six respiratory, pleural, or tissue cultures were performed in 16 patients, and 14 of these tests were positive. Ten of the 16 patients (63%) had at least one positive culture. Cultures were negative for all urine (11 patients [14 tests]), blood (10 patients [15 tests]), catheter tip (4 patients [4 tests]) and CSF (9 patients [9 tests]) samples.
|Type of text||No. of cultures positive/no. tested (%)|
|Respiratory specimenb||9/17 (53)|
|Tissue biopsy or swabc||4/7 (57)|
|Pleural fluid||1/2 (50)|
|Any positive cultured||14/26 (54)|
|Tissue pathologyf||6/8 (75)|
|Rapid PCRg||3/5 (60)|
PCR testing was performed on multiple specimens from four patients, including bronchoalveolar lavage (2/3 positive), induced sputum (1/1 positive) and cerebrospinal fluid (0/1 positive). One patient had a positive bronchoalveolar lavage PCR and a negative cerebrospinal fluid PCR.
Nine patients had CSF testing performed for the purpose of excluding coccidioidal meningitis. One of these nine had confirmed meningitis, demonstrating both CSF pleocytosis (84 nucleated cells/microliter) and elevated protein (61 milligrams/deciliter). The CSF serologic tests were negative by ID and CF, and positive by EIA for both IgM and IgG.
Abnormalities were initially identified by chest radiograph in 16 of 25 patients (64%). Specific radiographic findings included consolidation alone (n = 11); pleural effusion alone (n = 1); consolidation with pleural effusion (n = 3); and multifocal nodules (n = 1). The chest radiograph was repeated in 14 patients an average of 7 days after the first test, and the performance of a second radiograph increased the number of patients with identified chest radiographic abnormalities to 21 of 25 (84%). On repeat tests, the abnormalities included consolidation (n = 8, of which 3 were previously negative, 1 was previously consolidation and effusion, and 4 were unchanged from first test); pleural effusion (n = 1, which was previously negative); and pleural effusion and consolidation (n = 1, which was previously pleural effusion alone).
Abnormalities were identified by chest CT in 19 of 22 patients (86%). The identified abnormalities included consolidation alone (n = 4); multifocal nodules alone (n = 7); consolidation and multifocal nodules (n = 4); adenopathy and multifocal nodules (n = 1); adenopathy and consolidation (n = 1); pleural effusion and multifocal nodules (n = 1); and lung consolidation with multifocal nodules in liver (n = 1). A repeat CT scan was performed in three patients, an average of 47 days after the first test, and although an evolution of abnormalities was observed in patients who previously had abnormalities, the repeat scan did not identify abnormalities in patients with previously normal chest CT scans.
While not a fungus-specific test, an absolute eosinophil count of >0.4 × 109 cells/L was identified in 5 of 26 patients (19%). A second test in nine patients increased the number of patients with eosinophilia to 6 (23%).
- Top of page
- Materials and Methods
Coccidioidomycosis is a fungal infection acquired by inhaling airborne arthroconidia of the soil-dwelling species Coccidioides, which are endemic to the southwestern United States . Immunocompetent persons with coccidioidomycosis may experience either no symptoms or any of a wide range of symptoms that vary in severity from minimal to fulminant and life-threatening. Transplant recipients who acquire this infection have high rates of disseminated infection and mortality (33%  and 28% , respectively, as reported in recent series from our institution). Thus, while the diagnosis of symptomatic, newly acquired coccidioidomycosis was identified in as few as 27 of 2246 transplant recipients (1.2%) in the current study, prompt diagnosis and treatment of this infection are essential. To our knowledge, the current study is the largest thus far to examine the performance of diagnostic tests for newly acquired coccidioidomycosis among recipients of solid organ transplants, and the current cohort includes previously reported patients [5, 8, 15, 17].
The diagnosis of coccidioidomycosis requires a high index of suspicion, because symptoms and initial tests at presentation can be nonspecific and negative test results can be misleading. Positive cultures and histopathologic findings from infected specimens provide definitive results, since there is no colonized state. Individuals infected with Coccidioides often have a nonproductive cough, which makes it difficult to obtain a diagnostic specimen from patients whose clinical condition does not justify an invasive procedure. Among larger cohorts with proven or probable coccidioidomycosis, culture has been found to be 40% to 100% sensitive [8, 18]; similarly, a small study of renal transplant recipients with proven or probable primary coccidioidomycosis demonstrated positive cultures in roughly half the patients .
Serologic testing has long been used for the diagnosis of coccidioidomycosis [19, 20]. Several different methodologies are available, each with unique strengths and pitfalls. For example, the EIA appears to be more sensitive in early coccidioidal infection, but false-positives may render the IgM problematic. The CF test is positive later in infection, but it is quantitative; CF titers decrease with convalescence, making it useful as a serial test. ID testing is characterized as very specific, and it is useful for confirming the results of other serologic assays . The sensitivity of serologic testing may improve with repeat testing . When positive, serologic test results can be helpful, but a negative test result does not exclude the diagnosis, even in immunocompetent hosts.
Prior studies of transplant recipients with coccidioidomycosis have indicated that serologic responses may be absent or low [9, 10, 12], possibly because they are hindered by immunosuppressive antirejection medications . More recent assessments of this issue demonstrate that immunosuppressed patients may have positive serologic tests, but typically at lower rates compared with immunocompetent patients . Prior studies indicate that 1 or more serologic tests were positive in 5 of 7 (71%)  and in 12 of 12 (100%)  solid organ transplant recipients. Any single serologic test appears to have a suboptimal rate of positivity; however, the sensitivity of the serologic inquiry may be improved with a combination of serologic tests, and by repeating these tests over time. These prior studies were hindered by the lack of a gold standard benchmark (i.e. positive culture or histology). The present study demonstrated that among transplant recipients with proven coccidioidomycosis, any single serology by any methodology was positive in 10–45%, but repeating the serology and using multiple methods improved the positive rate to 82%.
Due to the lack of adequate sensitivity with serologic testing, some investigators have sought to identify, measure and correlate fungal antigen and nucleic acid assays as adjunct diagnostic tools for coccidioidomycosis. PCR was studied in 266 respiratory specimens (16 positive by culture) and found to have 100% sensitivity and 98.4% specificity compared with culture . A subsequent retrospective review of 158 PCR tests identified sensitivity of 75% in patients with confirmed or highly probable coccidioidomycosis (similar to that for fungal culture) ; PCR appeared to have similar sensitivity and specificity for either immunocompetent or immunosuppressed patients. Finally, PCR from bronchoalveolar lavage was described as the first test positive (enabling targeted treatment) in a liver transplant recipient with coccidioidomycosis whose serology remained negative . Since PCR was not available at our institution until recent years, only 4 of the 27 organ transplant recipients had a PCR test performed, and 3 of those 4 patients had a positive result. A positive PCR test was not the sole diagnostic criterion for the patients, and its utility is promising but bears further study.
The nonspecific finding of peripheral eosinophilia has been reported frequently in persons infected with coccidioidomycosis [23, 24], and it is our observation that, within the endemic area, eosinophilia in the presence of pneumonia often prompts a medical practitioner to consider a diagnosis of coccidioidomycosis. Coccidioidomycosis-associated eosinophilia may be extreme (more than 21% of the circulating white blood cells) . Eosinophilia is not diagnostic, but its presence in one-third to one-half  of transplant recipients with coccidioidomycosis may alert the clinician to the possibility of this fungal infection in the differential diagnosis.
Radiographic abnormalities are common in coccidioidal illness , and these are frequently mistaken in the general population to indicate community-acquired pneumonia [18, 26, 27]. In addition, radiographic manifestations of pulmonary coccidioidomycosis may resemble miliary infections (eg, tuberculosis) or malignancy . The most commonly identified abnormality using CT scanning is multiple bilateral nodules . A review  of 18 renal transplant patients identified multiple radiographic abnormalities, including nodular (n = 5), reticular (n = 5), and lobar (n = 3) infiltrates, as well as hilar adenopathy (n = 1) and cavities (n = 1); similar findings were identified in previous descriptions  and in the current study. We have found the chest CT to be an especially sensitive tool for diagnostic evaluation of solid organ transplant recipients with suspected coccidioidomycosis.
Assays to detect coccidioidal antigens have recently been developed, or are under current development. The identification of coccidioidal antigen from urine has been described  and is commercially available, with positive findings in 17 of 24 patients (71%) with coccidioidomycosis, including 19 who were immunosuppressed, 2 of whom were organ transplant recipients . Given the need for multiple diagnostic assays, the use of this urinary antigen test in solid organ transplant patients bears further investigation.
Because we used serology and culture results to identify patients for the study, then characterized the proportional positivity of each diagnostic test, we potentially introduced selection bias into the results. In addition, we did not look at a cohort without coccidioidal infection; therefore, the rates of positivity for any given diagnostic test are not measures of true sensitivity or specificity. Since most (although not all) of the patients with proven infection also had serology performed, we separately calculated the rates of seropositivity among subjects with proven infection and found that the rates of positive tests were not statistically different in the proven versus the probable groups (Table 2); the lone exception was the CF, which had a statistically lower rate of seropositivity (P = .04) in the group with proven infection.
Although this study is the largest study of diagnostic testing in transplant recipients with newly acquired coccidioidomycosis reported to date, its findings are limited by the small sample size and the retrospective design. Over the duration of the study, new diagnostic tests were introduced, which rendered the sample size for such testing even smaller. These patients were also under the care of a number of different physicians within our multispecialty group practice; therefore, not all patients had the same testing protocol. Because our study focused on a very specific population of patients, results may not be generalizable to the population at large.
In summary, a high index of suspicion is required for the diagnosis of coccidioidomycosis in the solid organ transplant recipient, because symptoms are nonspecific and serologic tests are not sensitive. Multiple test modalities may be needed, and repeat studies over time may increase the likelihood of assay positivity in the presence of infection. In the absence of an established alternate diagnosis, a negative test should not dissuade the clinician from continuing to search for coccidioidomycosis in the symptomatic transplant recipient with an epidemiologic history suggestive of a potential exposure to this organism.
- Top of page
- Materials and Methods
The authors thank Thomas Grys, PhD, for his review of the manuscript and for his helpful advice.
- Top of page
- Materials and Methods
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.
- Top of page
- Materials and Methods
- 7The diagnosis of coccidioidomycosis. F1000 Med Rep. 2010 Jan 18; 2..
- 13Revised definitions of invasive fungal disease from the European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) Consensus Group. Clin Infect Dis 2008; 46: 1813–1821., , , et al; European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group; National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) Consensus Group.
- 23Primary pulmonary coccidiodomycosis: Report of an epidemic of 75 cases. War Med 1943; 4: 299–317., .