Reply to “Low-Positive Histoplasma Antigen Results in the MVista Assay Should Not Be Assumed To Be False Positive”

We read with interest the letter to the editor from Hage and Wheat regarding our recent publication entitled “Clinical significance of low-positive Histoplasma urine antigen results” and consider that a few points require clarification (1, 2).

Hage and Wheat comment that 12 of 57 patients had a MiraVista (MVista) Histoplasma capsulatum urine antigen (UAg) result below the limit of quantification (BLQ), which was considered falsely positive in our publication. Our study, however, focused only on the clinical utility of “positive, BLQ” MVista results from patients without a history of histoplasmosis, and the medical charts of only this subgroup of individuals (n = 25) were reviewed. As noted in our report, of the 57 patients who met the inclusion criteria (i.e., a “positive, BLQ” UAg result), 32 were excluded from analysis because they had a previously positive MVista Histoplasma UAg result and the most recent testing was assumed to have been ordered for monitoring purposes. The medical records of these 32 patients were not reviewed, and therefore, these patients should not be included in the above calculation. This does, however, present an interesting question: what is the clinical significance of persistent, low-level Histoplasma antigenuria in patients who have completed an appropriate course of antifungal therapy who are asymptomatic? The most recent Infectious Diseases Society of America guidelines note that continued low-level antigenuria may not be an adequate reason to prolong treatment in the absence of ongoing clinical disease (3). Additional studies to define the clinical significance of such results are needed.

Hage and Wheat state that 10 of the 12 patients classified as falsely positive in our review were misclassified. We respectfully disagree for the following reasons. First, 5 of the 10 subjects in question were ultimately diagnosed with another endemic mycosis (blastomycosis [n = 3] or coccidioidomycosis [n = 2]) with no evidence of histoplasmosis. We considered these cases falsely positive since H. capsulatum was neither recovered in culture nor identified by nucleic acid amplification testing (NAAT), serology, or histopathology. MVista offers individual UAg assays for each of these three fungal agents, which, albeit indirectly, implies a certain level of analytical specificity. Yet, UAg cross-reactivity between the Histoplasma and Blastomyces assays is acknowledged (www.miravistalabs.com, accessed on 17 September 2014). Our own observations show that 29% (44/150) of the patients tested for Histoplasma UAg between 19 August 2013 and 10 February 2014 also had a Blastomyces UAg test ordered and that the two tests showed 97.7% qualitative agreement (43/44; 6 Histoplasma UAg positive, of which 5 were also Blastomyces UAg positive) (unpublished data).

Second, Hage and Wheat note that the two patients with sarcoidosis in our study were seropositive for antibodies to Histoplasma and that death has occurred in this patient group because of undiagnosed histoplasmosis. While we do not disagree, neither of the two patients in our review received antifungal therapy and neither developed histoplasmosis in the year following testing. Further, the complement fixation (CF) titers of both patients were low (1:16) and did not change over time. Additionally, such low titers are of limited utility in differentiating acute infection from previous exposure (4). These data, along with the absence of other diagnostic findings positive for Histoplasma, were interpreted by both the managing health care providers and our independent review as not associated with an active H. capsulatum infection.

Third, Hage and Wheat indicate that subacute pulmonary histoplasmosis (SPH) cannot be ruled out in the last three patients, as approximately 30% of individuals with SPH have a positive MVista Histoplasma UAg result and subsequently recover without treatment (5). However, the authors do not provide a range or average quantitative (ng/ml) antigen value for this subgroup of patients. Additionally, Hage and Wheat indicate that the diagnosis of SPH is ultimately dependent on the detection of antibodies to H. capsulatum, as 95% of SPH patients are positive by CF and/or immunodiffusion testing (5). These three patients were all seronegative for antibodies to H. capsulatum, supporting the absence of SPH (only positive laboratory findings were included in Table 1 of reference 2).

Ultimately, as indicated in our original publication and acknowledged by Hage and Wheat, we recommend that clinicians use caution when interpreting “positive, BLQ” Histoplasma UAg results of patients not previously diagnosed with histoplasmosis. Such results should be correlated with other laboratory findings, including serology, culture, repeat UAg testing, NAAT, and histopathology, as available.