A Comprehensive Diagnostic Approach Improves the Diagnostic Accuracy of Invasive Fungal Disease (IFD) in Adult Haemato-Oncology Patients Undergoing HSCT or High Dose Chemotherapy- Results of the King's Prospective Aspergillosis Study (NCT0081608

M. Mansour Ceesay, MRCP, FRCPath, Lisa Berry*, Sujal R Desai, MBBS, MD, FRCR*, Joanne Cleverely, FRCR*, Christopher Kibbler, FRCPath*, Jim Wade, FRCPath*, Melvyn Smith, PhD*, Ghulam J. Mufti, MD8 and Antonio Pagliuca, MBBS, FRCP, FRCPath, MA

Author address: 

Haematological Medicine, Kings College Hospital, London, England, Haematological Medicine, Kings College Hospital, London, United Kingdom, Radiology, Kings College Hospital, London, United Kingdom, Radiology, Royal Free Hospital, London, Uni


Background: Invasive fungal disease (IFD) is a difficult diagnosis in patients undergoing haematopoietic stem cell transplantation (HSCT) or high dose chemotherapy. For clinical trials the revised EORTC/MSG criteria is a useful diagnostic tool but there are few data on its usefulness in clinical practice. The aim of this study was to evaluate the ’real world’ incidence and outcome of IFD in patients undergoing HSCT or high dose chemotherapy with expected neutropenia of 10 days or more using all approved EORTC/MSG diagnostic tools. Methods: Patients were recruited prospectively between December 2008 and May 2010 and followed for at least 4 months after chemotherapy/HSCT or until death. During admission twice weekly galactomannan (GM) was performed and neutropenic sepsis unresponsive to antimicrobials triggered diagnostic work-up with continuous volume acquisition computed tomography (CT) and tissue diagnosis was obtained where possible. β -D-glucan (BDG) was performed on all positive GM samples (n=139), patients with relevant CT abnormalities but negative GM (n=41), and 79 control samples. The CT scans were reviewed independently by two chest radiologists and all cases were independently verified before assigning EORTC status. All patients had antifungal prophylaxis during the period of neutropenia or continuing immunosuppression (fluconazole for low risk patients, itraconazole suspension or posaconazole for high risk patients, and voriconazole for secondary prophylaxis). Results: Two hundred and three patients were recruited [123 male, 80 female; median age 54y (range 1973); median follow-up 194 days (range 12647)]. The underlying diagnoses were: acute leukaemia 62 (myeloid 55, lymphoid 7), chronic leukaemia 4 (myeloid 3, lymphoid 1), MDS/MPD 33, aplastic anaemia 19, lymphoma 37 (non-Hodgkin's 29, Hodgkin's 8), multiple myeloma 45, and others 3. Main treatments received were: HSCT 165 (81%) [allogeneic 94, cord 5, autologous 66], chemotherapy 28 (14%), and immunosuppressive therapy (IST) 8 (4%). The total number of treatments received during study period was 263 (allografts 106, chemotherapy 77, autografts 67, IST 13). The patients were heavily pre-treated; 85% had at least one prior therapy with a median of 5 cycles of therapy/patient (range 117). There were 44 (21%) cases of IFD in 40 patients: proven 14 (Aspergillus fumigatus 2, Fusarium spp 1, mould IFD 5, non-candida albican spp 5, Pneumocystis jirovecii 1), probable 30. Invasive pulmonary aspergillosis was the commonest presentation (82%). The treatment-specific incidences were 17%, 12%, 7%, and 4% for allogeneic HSCT, chemotherapy, immunosuppressive therapy, and autologous HSCT respectively. Patients with AML/MDS had the highest risk (27 cases) while multiple myeloma represented the lowest risk (1 case). Median time to diagnosis of IFD was 42 days (range 3 to 524) from index chemotherapy or HSCT. Using galactomannan (GM) or β -D-glucan (BDG) alone (plus host and CT scan evidence) the apparent incidence of IFD would be 13% and 17% respectively, 8% and 4% lower than the true incidence. The sensitivity, specificity, positive predictive value, and negative predictive value was 63%, 61%, 34%, and 84% for GM and 89%, 21%, 44%, and 75% for BDG respectively. The overall and IFD-attributable mortality rates were 32% (64/203) and 6% (13/203) respectively and this was significantly higher among IFD cases (55% mortality, 33% attributable mortality, P=0.001). The most common causes of death were relapse/progression of underlying disease (38, 59%) and IFD (13, 20%). Conclusion: Our findings demonstrate that a multi-diagnostic approach is necessary in order to improve diagnostic accuracy of IFD. Relying on GM or BDG alone underestimates the true incidence of IFD. Our data suggests that the ’real world’ incidence of IFD is higher than usually reported especially in tertiary referral transplant units. This data has important implications for clinical practice as well as clinical trials where improved diagnostic tools would improve recruitment. Disclosures: Ceesay: Pfizer, Gilead: Research Funding. Berry: Pfizer, Gilead: Research Funding. Kibbler: Gilead: Consultancy, Research Funding. Wade: Pfizer, Gilead: Research Funding. Smith: Pfizer, Gilead: Research Funding.

abstract No: 


Full conference title: 

53rd American Society of Haematology
    • ASH 53rd (2011)