An EORTC multicentre prospective survey of invasive aspergillosis in haematological patients: diagnosis and therapeutic outcome. EORTC Invasive Fungal Infections Cooperative Group

Abstract

Objectives: The EORTC Invasive Fungal Infections Cooperative Group (IFICG) conducted a prospective survey by questionnaire of all cases of invasive aspergillosis (IA) in cancer patients to ascertain current diagnostic and therapeutic approaches.

Methods: All members of the IFICG were asked prospectively to complete a detailed questionnaire for each IA case identified in their institution over a 12-month period.

Results: One hundred and thirty questionnaires were returned. All cases were independently evaluated (DWD & JC) and 123 were eligible. Cases came from 20 hospitals in eight countries and the number of cases per institution varied from 1-21. Acute myeloid leukaemia (AML) (60, 49%), acute lymphoblastic leukaemia (ALL) (21, 17%) and lymphoma (11, 9%) were the most frequent underlying diseases, and 16 (12%) patients had received an allogeneic bone marrow transplant. Pulmonary involvement was present in 87%, infection of sinuses/nose in 16% and brain in 8%. The chest radiograph was initially normal in 9% of those with primary pulmonary disease. The diagnosis was confirmed in 50%, probable in 31% and possible in 19%. The evidence for IA was on the basis of clinical and radiological features alone in 28%, with culture or histology in another 31% and 9%, respectively, and with both culture and histology in 29%. In three (2%) patients with diagnosis was based on culture or histology alone. Treatment was given to 120 patients (98%)-amphotericin B 75%, lipid-associated amphotericin B 36%, itraconazole 40%, flucytosine 12%, growth factors 33%, lobectomy 5%. At 3 months after diagnosis or first suspicion of IA, 44 (36%) patients were alive and 79 (64%) dead. Outcome was best in those with AML (30% death and 46% with a complete antifungal response or cure). Growth factors (mostly granulocyte colony stimulating factor) appeared not to influence outcome (P = 0.99).

Conclusion: IA remains a considerable diagnostic and therapeutic challenge. No single diagnostic procedure was universally successful and a multifaceted approach including surgery is necessary. There was no discernable difference in outcome between initial therapy with amphotericin B, itraconazole or lipid-associated amphotericin B, although numbers are limited and the study was retrospective.