ABSTRACT
When used with BAL samples, they often reveal distinctive fluorescent clusters of trophozoites and cysts enmeshed in exudate that is also labeled. Although IF A provides high sensitivity and specificity, this diagnostic tool may not be available due to drawbacks of reagent cost, fluorescence microscope requirement and the time, skill and equipment needed for processing the specimens. PeR-based nucleic acid detection offers exquisite sensitivity and specificity, but suffers from lack of FDA-approved kits as well as the general problems of PeR-based diagnoses: costs of special reagents, special equipment, technical training, inclusion of extensive positive and negative controls and need for rigid procedures to prevent false positives due to contamination. Furthermore, the extreme sensitivity of peR presents a problem in that colonization with small numbers of P. jiroveci can be detected even if they are not causing disease. A well-designed and fully developed quantitative-peR protocol with a clear "initial target copy number" threshold could resolve many of these issues and encouraging clinical studies have been done; however, even if fully developed, general disadvantages of peR diagnosis will remain. Reported success rates for the various diagnostic methods vary from study to study, partially due to the experience and expertise of physicians and laboratories, but trends are clear. Open lung or transbronchial biopsy specimens stained for cysts provide diagnostic sensitivity and specificity of 98+ and 100%, respectively, but both procedures are invasive and present risks to the patient. Sensitivities and specificities for BAL fluid samples stained for cysts are reported as high as 9S and 100%, respectively; IFA can improve sensitivity to 98% while retaining selectivity. Reports of efficacy using induced sputum samples have a much broader spread with sensitivity ranging from 30 to 90%; specificity remains high.
