gms | German Medical Science

Purpose: Invasive aspergillosis (IA) remains a major complication in patients with haematological malignancies (HM) and after allogeneic haematopoietic cell transplantation (HCT). In these patients, IA is the most common cause of mortality due to infection. Early detection of Aspergillus infections would have the potential to facilitate a more effective management of invasive disease. The optimal blood fraction for detecting Aspergillus DNA has still to be determined and different fractions will provide different DNA sources. Free circulating fungal DNA (DNAemia) is likely to be available in plasma and serum, but the differences in obtaining these cell free fractions (presence of blood clot) may alter the availability of free DNA.

Therefore we evaluated the sensitivity of different blood fractions.

Methods: The centres in Cardiff and Wuerzburg collected consecutive samples from HM patients undergoing chemotherapy or HCT. Blood samples were collected twice weekly from all study patients classified according to the European Organization for Research and Treatment of Cancer consensus criteria. In detecting Aspergillus DNA in plasma (centre 1) or serum (centre 2) methods conforming with European Aspergillus PCR Initiative recommendations were used. Galactomannan and PCR assays were performed prospectively.

Results: In total 64 patients were included over a one-year study period. There were 12 probable and 6 possible IA cases as well as 46 unclassified patients who served as controls.

Overall, 696 blood samples were collected. The PCR positivity for centre 1 testing 388 plasma samples was 10.8% and for centre 2 testing 308 serum samples was 3.3%. Centre 1 detected 9 of 10 probable cases (90%), one possible IA case (100%) and 13 of 27 unclassified patients. Ten of the 13 unclassified patients showed only one single positive PCR result. If a threshold of two PCR positive results was used 8 of 10 probable cases (80%) remained positive and specificity increased to 89%. Centre 2 detected 2 of 2 probable cases (100%), 1 of 5 possible cases (20%) and 2 of 19 unclassified patients.

PCR positivity rate in centre 1 was 15.3%, 15.5% and 7.8% and for centre 2 10%, 4.4% and 1.3% for probable, possible and unclassified patients, respectively. Negative Predictive Values (NPV) were high for centre 1 and 2 (0.93 and 1.00; possible IA cases were excluded for the analysis).

Conclusion: PCR positivity rates appear greater in plasma compared to serum. However, both blood fractions, plasma and serum, allowed the detection of probable IA cases with high sensitivity. Further research testing the concomitant serum samples at centre 1 and plasma samples at centre 2 is being performed to permit direct performance comparison.