Predicting azole treatment failure with PCR-based detection of A. fumigatus cyp51A mutations on bronchoalveolar lavage. A multi-centre validation study in 201 haematology patients with suspected invasive aspergillosis

Galai Chong, Martha van der Beek, Peter von dem Borne, Jerina Boelens, Eva Steel, Greetje Kampinga, Bart Spam, Katrien Lagrou, Johan Maertens, Gijs Dingemans, Giel Gaajetaan, Dennis van Tegelen, Jan Cornelissen, Alieke Vonk, Bart Rijnders

Abstract: 

Background: In patients with invasive aspergillosis (IA), azole resistance is associated with a high
mortality rate. However, when fungal cultures remain negative, azole resistance stay undetected. The
AsperGenius® multiplex real-time PCR assay identifies clinically relevant Aspergillus species and 4
resistance associated mutations (RAM: TR34/L98H/T289A/Y121F) in the cyp51A gene. The purpose
of this multicenter study was to evaluate the diagnostic performance of this PCR on bronchoalveolar
lavage (BAL) and to correlate the presence of RAMs with azole treatment failure and mortality.
Material/methods: BAL samples from hematology patients with suspected IA were used. A
galactomannan ≥1.0 or a positive culture was used as the gold standard to determine the best PCR
cycle treshold (Ct) cut-off value by ROC analysis. Using the RAM PCR results, the azole treatment
failure and 6-week mortality was compared in patients with and without RAMs.
Results: 201 patients contributed each 1 BAL sample, of which 88 were positive according to the gold
standard. The ROC analysis led to a diagnostic accuracy by AUC of 0.890 (CI 95% 0.84 – 0.94;
p<0.001). The closest to (0,1) analysis showed that the optimal Ct value cut-off was <38, while the
Youden index selected <36 as optimal. The Ct cut-off of <36 resulted in a sensitivity, specificity, PPV
and NPV of 70%, 96%, 93% and 81% while Ct of <38 gave values of 84%, 80%, 76% and 87%,
respectively. Because avoiding false negative results was deemed more important than avoiding false
positives, the Ct value of <38 was chosen as the optimal cut-off. Voriconazole treatment failure was
observed in 6/8 patients with a RAM compared to 14/50 patients without RAMs (P=0.02). Six-week
mortality was 30% higher in patients with RAMs (20% versus 50%, P=0.09), figure 1. The BAL culture
of 6 of the 8 patients with RAM was culture negative.
Conclusions: The AsperGenius® assay has a good diagnostic performance on BAL and
differentiated wildtype from A. fumigatus with RAMs, also in culture negative BALs. Most importantly,
detection of RAMs predicted azole treatment failure.

Tables: 

2016

abstract No: 

#1914

Full conference title: 

26th European Congress of Clinical Microbiology and Infectious Diseases
    • ECCMID 26th (2016)