The utility of real-time polymerase chain reaction in detecting Coccidioides immitis among clinical specimens in the Central California San Joaquin Valley.

Coccidioidomycosis, the fungal infection caused by dimorphic Coccidioides species, is typically diagnosed by histopathologic identification of spherules, by culture, or by serology. These tests are reliable but time-intensive, delaying diagnosis and treatment. Rapid real-time polymerase chain reaction (RT-PCR) can be performed and was validated to identify Coccidioides immitis using an in-house developed assay for the Becton Dickinson molecular instrument (BD MAXTM). These studies were performed using patient samples that had been shown to be positive on previously set up fungal cultures. To evaluate this new RT-PCR test in the clinical setting, we conducted a retrospective chart review of patients (N = 1160) who underwent Coccidioides PCR (Cocci PCR) on clinical samples between March 1, 2014, and Dec 31, 2016. We abstracted clinical, microbiologic, serologic, radiographic, treatment, and follow-up data. Specimens of cerebrospinal fluid (CSF), bronchioalveolar lavage fluid (BAL), lung tissue biopsy (LTB), sputum, and pleural fluid were evaluated to determine sensitivity and specificity. Of the 113 specimens that tested positive for Cocci PCR, all had clinical disease defined by traditional clinical criteria, yielding 100% specificity. Overall sensitivity was 74% versus 46% for fungal culture and was available in 4 hours rather than 1-2 weeks. Sensitivities varied by source material and clinical setting. CSF had a sensitivity of 59%, BAL for acute pneumonia 91%, sputum for acute pneumonia 94%, pleural fluid 86%, but LTB for lung nodules only 44%. Overall positive predictive value (PPV) was 100%, while negative predictive value (NPV) was 96%, but again this varied by specimen and clinical setting. Our experience with clinical testing of >1160 specimens over 2-3 years shows we can utilize this technology to improve our ability to diagnose disease but that the sensitivity varies by specimen source and clinical setting.