Methods: Samples had been prospectively banked in our BAL repository. Forty-two samples, 14 from pt with proven/probable IPA by EORTC/MSG criteria and 28 from control pt without IPA, were tested with AspID and AspLFD. For AspID, DNA extraction and qRT-PCR were performed per manufacturer instructions. For AspLFD, 100μl of sample was applied to the device. AspID and AspLFD results were each read by 3 different blinded observers. Only pt with a valid result for both tests were included in the analysis. Sensitivity, specificity, and accuracy of AspID alone and in combination with AspLFD were calculated.
Results: Of the 42 samples, 22 were excluded because the AspID internal extraction control showed the assay to be invalid and 1 sample was excluded because the AspLFD internal control line was not visible. Thus, 19 pt were analyzed, 8 with IPA and 11 without IPA. Among 8 IPA cases, 7 were positive by AspID and 1 was negative; 2 tested positive by AspLFD and 6 were negative. Of the 11 control pt without IPA, 4 were positive by AspID and 7 were negative; all 11 were negative by AspLFD. AspID sensitivity was significantly higher than that of AspLFD (87.5% vs. 25%, p=.0001), but specificity of AspLFD was superior to that of AspID (100% vs. 64%, p=.049). Accuracy was 74% for AspID and 68% for AspLFD. When deciding whether doing both tests was beneficial for diagnosis, union analysis showed the sensitivity to be 87.5% and the specificity to be 64%. Accuracy was not improved and remained at 74%.
Conclusion: AspID had higher sensitivity than AspLFD and AspLFD had higher specificity than AspID. Using both tests in combination did not improve the ability to diagnose IPA in pt with classic risk factors.
K. A. Linder,
S. Zhou, None
J. A. Diaz, None
C. A. Kauffman, None
M. H. Miceli, None