Are Two Tests Better Than One for Detecting Invasive Aspergillosis?

Abstract & Commentary

Synopsis: The case for this PCR assay for aspergillus DNA remains unproven; and that proof can only be obtained by undertaking a formal prospective study in which samples for the detection of galactomannan and fungal DNA by PCR are taken at several predetermined intervals eg, 2-3 times weekly from all patients at risk.

Source: Challier S, et al. Development of a serum-based Taqman real-time PCR assay for diagnosis of invasive aspergillosis. J Clin Microbial. 2004;42:844-846.

Challier and colleagues had serum obtained from 7 proven cases of invasive aspergillosis (3 pulmonary, 1 trachea, and 3 disseminated), 19 probable (15 pulmonary, 4 cerebral and pulmonary) and 15 possible (10 pulmonary, 1 cerebral and pulmonary, and 4 without clinical or radiological signs). They compared a real-time PCR assay specific for Aspergillus fumigatus (Taqman) to the enzyme-linked immunosorbent assay for galactomannan (Platelia Aspergillus). Whilst the PCR technique appeared more specific than the galactomannan assay, the detection of invasive aspergillosis (IA) was only optimal by using the results of both tests (see Table below).

Comment by J. Peter Donnelly

The early diagnosis of invasive aspergillosis is one of the many unmet needs in hospitals and continues to present a challenge to the laboratory. Much is at stake since without such a reliable test clinicians will continue to rely on empirical antifungal therapy to deal with suspected fungal infection. This has become more attractive now that the choice extends beyond amphotericin B, its lipid formulations, and to include itraconazole and voriconazole and the echinocandin, caspofungin, and tempts some clinicians to opt for combinations of drugs just to be on the safe side even though there is no evidence to support this. Hence, I, for one, continue to scour the literature for reports of laboratory tests that look promising.

There are now several studies which purport to show the same results for PCR, namely that it provides a sensitive means for detecting aspergillosis. Like others before, Challier et al had to rely on serum that had been stored for other purposes. Hence, the number of samples varied from 2 to 19 with an average of 5.3. Only 66% of samples were taken at the onset of the disease. In fact, the samples had been obtained a maximum of 2 months before or after the onset. Consequently, galactomannan and fungal DNA were detected up to 3 weeks before the onset and up to a month afterwards. This makes the picture a lot less clear since the detection of Aspergillus assay, antigen, and probably its DNA decline during therapy. In fact, the tests were only done for 19 (46%) patients within a week of onset of suspected IA. Four proven IA and 6 probable IA were detected by both tests, 4 cases of probable IA and 5 cases of possible IA were detected only by galactomannan, and 1 case of probable IA was detected only by PCR. This makes the results look less appealing for PCR. So while Challier et al’s conclusions are correct that combining the 2 tests should improve diagnosis, their own results suggest another plausible conclusion, namely that PCR adds little to the diagnostic yield already attained by using galactomannan. However, it would be much fairer to conclude that the case for this PCR technique remains unproven and that proof can only be obtained by undertaking a formal prospective study in which samples for the detection of galactomannan and fungal DNA by PCR are taken at several predetermined intervals (eg, 2-3 times weekly from all patients at risk).

J. Peter Donnelly, PhD, Clinical Microbiologist University Hospital Nijmegen, The Netherlands Section Editor, Microbiology, is Associate Editor of Infectious Disease Alert.