The use of the serum galactomannan assay in high -RISK hematology patients on primary prophylaxis with posaconazole : towards an efficient diagnostic strategy

Ref ID: 19568

Author:

I Sánchez-Ortega1*, I Cuesta2, B Patiño1, M Arnan1, A Fernández de Sevilla1, C Gudiol1, J Ayats3,
M Cuenca-Estrella2, RF Duarte1

Author address:

1Hematology, Institut Catalí  d’Oncologia (ICO) – Hospital Duran i Reynals, Barcelona, Spain
2Microbiology, Centro Nacional de Microbiologí­a – Instituto de Salud Carlos III, Majadahonda – Madrid,
Spain
3Microbiology, Hospital Universitario de Bellv

Full conference title:

6th Advances Against Aspergillosis 2014

Abstract:

Purpose and Methods:
We have recently reported that primary antifungal prophylaxis with posaconazole is effective to
prevent invasive mould infections in high-risk hematology patients, with only 5 cases of probable/
proven breakthrough invasive aspergillosis (IA; 1.9%) in a series of 262 consecutive high-risk
episodes (161 AML/MDS, 79 allogeneic transplantation, 22 graft-versus-host disease) from
121 patients in our center between June 2007 and June 2011 [Sí¡nchez-Ortega; P950 EBMT 2013].
The management algorithm for this series included a comprehensive diagnostic approach for
symptomatic patients as well as preemptive serum galactomannan tests (GM; Platelia, Bio-Rad)
performed twice weekly from the start of the risk episodes. Based on the sensitivity (70%) and
specificity (90%) of serum GM tests in hematology patients (Pfeiffer, CID 2006) and a 1.9% reallife
incidence of IA, the negative predictive value of the test remains very high (>99%) but the
expected positive predictive value would drop sharply to 12%, potentially limiting the diagnostic
performance of the assay when used as a screening technique in this setting. Here, we present an
analysis of our current biweekly serum GM test screening strategy in high-risk hematology patients
on posaconazole primary prophylaxis and describe possible improvements of its use in response to
the low pre-test incidence of IA.
Results:
A total of 2972 GM tests were performed in 262 high-risk episodes (median 11 per episode, 3-30).
In the vast majority of episodes (188, 71.7%), all GM tests were negative. In all 5 episodes of IA,
positive GM tests in the serum (Optical Index [OI] >0.7 x1 or OI >0.5 x 2 consecutive samples)
and in the BAL (OI >1 x1) contributed to the diagnosis (Table). However, in 30 additional episodes
(11.4%) all positive GM tests were false positive results (FP, 1-6 per episode; OI: 0.59-2.85) as
defined as those positive tests performed on patients who remained on posaconazole prophylaxis
without receiving other antifungal therapy, survived the risk episode and never developed other
criteria of IA. In 87% of these episodes (n=26) all FP GM tests were performed as a screening test in
asymptomatic cases. In only 4 episodes (13%) such FP tests were performed as part of the diagnostic
workup in persistently febrile symptomatic patients (p<0.001) (Table). Conclusion: Our experience in high-risk hematology patients receiving effective antifungal prophylaxis with posaconazole suggest that the serum GM assay provides no clinical or diagnostic benefit when performed as a surveillance test in asymptomatic patients. In this screening setting tests were always negative or FP. However, serum GM tests remain very useful as part of the management strategy for symptomatic patients, both for their contribution to the diagnosis of the small percentage of breakthrough IA, as well as for its high negative predictive value. Overall, these data suggest that the diagnostic accuracy of GM tests can be improved when used in high-risk hematology patients on posaconazole primary prophylaxis who yield a low pre-test incidence of IA.

Abstract Number: 94

Conference Year: 2014

Link to conference website: http://www.AAA2014.org

New link: NULL

Conference abstracts, posters & presentations

Showing 10 posts of 17325 posts found.

  • Title

    Author

    Year

    Number

    Poster