LETTER TO THE EDITOR
Limited utility of serum galactomannan assay after auto-SCT
Invasive Aspergillosis (IA) is a dreaded infectious compli-
cation among immunocompromised patients, including
those receiving treatment for hematologic malignancies.
Preemptive diagnosis of IA, using serum markers alone or
in combination with computed tomography (CT), has
received significant attention over the last decade but
In 2003, an ELISA to detect serum galactomannan (GM)
was approved by the US Food and Drug Administration.4
In a meta-analysis of 27 studies using the GM assay for
surveillance, the sensitivity and specificity were 71 and
80%, respectively, for cases of definite IA. However, there
was significant heterogeneity among studies. Also, most
studies of serum GM have included patients with acute
leukemia or allogeneic hematopoietic SCT recipients. In
these populations, the risk for IA varies between 5 and
15%.5–9In contrast, IA among autoSCT recipients is
significantly less common, ranging between 0 and 8%.5,7–9
This lower risk raises the question whether the GM assay
has utility among auto-SCT recipients.
Memorial Sloan-Kettering Cancer Center (MSKCC) is a
tertiary care cancer center in New York City. Medical and
laboratory records were reviewed for all patients who
underwent auto-SCT at MSKCC between September 2003
and June 2007. The study was approved by the MSKCC
Institutional Review Board. NIAID/MSG criteria10were
used to categorize definite, probable or possible IA. Mold-
active prophylaxis was not routinely given to patients
undergoing auto-SCT, but was administered to a limited
number of patients at the treating physician’s discretion.
Galactomannan assays were performed only at the
treating physician’s discretion, generally in response to
persistent fever and neutropenia or the development of
pulmonary lesions identified by CT. A positive serum GM
was defined as X0.500. The Student’s t-test was used to
compare continuous variables. STATA software (version 7)
was used to calculate the area under the receiver operating
characteristic (ROC) curve.
During the study period, 583 patients received 698 auto-
SCTs (Table 1). Of the 583 patients, 43 (7.4%) underwent
91 GM assays (Table 1). Ten (23.3%) of the 43 patients
received peri-transplant prophylaxis with a mold-active
antifungal, 29 (67.4%) received fluconazole and 4 (9.3%)
received no antifungal prophylaxis. Twenty-three (53.5%)
patients received mold-active antifungal therapy within 72h
before their first serum GM.
The patients who underwent GM testing tended to be
younger and more frequently male than the remaining
population (Table 1). The spectrum of malignancies also
differed between the groups, with a greater proportion of
patients who underwent GM testing having Hodgkin’s
lymphoma (32.6 vs 12.6%) (Table 1).
Only 2 (2.2%) of 91 GM assays were positive, both from
the same patient who received a second auto-SCT for
multiple myeloma. He also grew Aspergillus fumigatus from
sputum specimens. He died despite a 36-day course of
combination therapy with voriconazole and micafungin.
An autopsy confirmed the presence of pulmonary mold
infection. No other patients had positive GM assays. None
of the other 582 patients who underwent auto-SCT had a
positive culture for mold by day þ100 after transplant.
Beside the one patient with definite IA, none had
probable IA, 14 had possible IA based on CT of the chest,
and 28 were found to have no evidence of IA. Assuming
that all patients with definite, probable or possible IA had
IA infection, the sensitivity of the GM assay was 6.7%. The
mean (0.117 vs 0.106; P¼0.37) and range of GM values for
patients with possible IA mirrored the range among
patients with no evidence of IA (Figure 1). ROC analysis
illustrated a minimal improvement of the GM assay over
chance, with an area under the ROC curve¼0.5857.
Differences in treatment appeared to depend on clinical
factors rather than GM assay values. The 14 patients who
were found to have possible or definite IA received
prolonged courses (mean, 52 days; median, 25 days) of
mold-active therapy. In contrast, patients who underwent
GM testing but had no evidence of IA received shorter
courses (mean, 17 days; median, 7 days).
Five additional patients were believed to have fungal
infection based on ICD-9 coding, but they did not undergo
GM testing. All had possible IA, and two were diagnosed
with possible IA before auto-SCT and continued on
voriconazole through the transplant period.
The serum GM assay has been used both as a preemptive
screen, typically on a weekly or twice-weekly basis, and as a
diagnostic test in the setting of clinical signs or symptoms,
in patients at high-risk for IA. The risk of IA among
patients receiving auto-SCT is significantly lower than that
among high-risk groups such as allo-SCT recipients. In our
series, only 19 (3.3%) of 583 auto-SCT recipients developed
possible, probable or definite IA, and only one of these had
a positive culture for Aspergillus, a positive GM assay or
consistent pathologic findings, raising the likelihood that
the true frequency of IA was much lower. It is noted that
none of the patients with possible IA had GM values 40.2
and that the spectrum of values in these patients did not
differ from those with no evidence of IA (Figure 1). Thus,
the lower risk of IA among auto-SCT recipients makes the
cost and effort of a preemptive approach difficult to justify.
Instead, we use the GM assay solely in the setting of clinical
Bone Marrow Transplantation (2009) 44, 59–61
& 2009 Macmillan Publishers Limited All rights reserved 0268-3369/09 $32.00
suspicion for IA, either because of pulmonary symptoms,
findings on CT imaging or persistent fever.
Empiric mold-directed therapy was common in our
population and may have contributed to the low rate of
positive GM assays. Patients with possible IA received a
median of 25 days of anti-mold therapy, despite having one
or more negative GM assays. Thus, treating physicians
generally did not value the assay as a negative predictor
We used the GM assay as a diagnostic test for IA within
the first 100 days after auto-SCT. In our experience, this
approach was not helpful for confirming or excluding the
diagnosis of IA and did not appreciably influence manage-
ment decisions regardless of test results.
Conflict of interest
The authors do not have a commercial or other association
that might pose a conflict of interest (for example,
pharmaceutical stock ownership, consultancy, advisory
board membership, relevant patents or research funding).
Financial support was provided by the MSKCC Depart-
ment of Medicine.
A Jathavedam1, DC Dure ´1, Y Taur1and DM Weinstock1,2
1From the Infectious Disease Service, Department
of Medicine, Memorial Sloan-Kettering Cancer Center,
New York, NY, USA and
2Adult Allogeneic Bone Marrow Transplant Service,
Department of Medicine, Memorial Sloan-Kettering Cancer
Center, New York, NY, USA
1 Maertens J, Theunissen K, Verhoef G, Verschakelen J,
Lagrou K, Verbeken E et al. Galactomannan and computed
neutropenic patients at high risk for invasive fungal infection:
a prospective feasibility study. Clin Infect Dis 2005; 41:
2 Marino IR, Panarello G, Singh N. Efficacy of Aspergillus
galactomannan-directed preemptive therapy for the prevention
of invasive aspergillosis in organ transplant recipients. Transpl
Infect Dis 2002; 4: 226–227; author reply.
2007 and the subset of 43 patients who had serum galactomannan (GM) testing performed
Demographic factors for the 583 patients who underwent auto-SCT at Memorial Sloan Kettering between September 2003 and June
All patients (%)No GM assay sent (%) GM assay population (%)
Median age, years (range)
Age less than or equal to 18 years
Diffuse large B-cell lymphoma
Mantle cell lymphoma
Other lymphoid neoplasm
Germ cell tumor or neuroectodermal tumor
AML or myelodsyplasia
proven invasive Aspergillosis (IA) (top left), 14 patients with possible IA
(bottom left) and 28 patients with no evidence of IA (right).
A plot of serum galactomannan values for the one patient with
Letter to the Editor
Bone Marrow Transplantation
3 Maertens J, Deeren D, Dierickx D, Theunissen K. Preemptive
antifungal therapy: still a way to go. Curr Opin Infect Dis 2006;
4 Rapid test for Aspergillus infection. FDA Consum 2003; 37: 7.
5 Sinko J, Csomor J, Nikolova R, Lueff S, Krivan G, Remenyi P
et al. Invasive fungal disease in allogeneic hematopoietic stem
cell transplant recipients: an autopsy-driven survey. Transpl
Infect Dis 2008; 10: 106–109.
6 Steinbach WJ, Addison RM, McLaughlin L, Gerrald Q,
Martin PL, Driscoll T et al. Prospective Aspergillus galacto-
mannan antigen testing in pediatric hematopoietic stem cell
transplant recipients. Pediatr Infect Dis J 2007; 26: 558–564.
7 Rooney S, Sekiguchi J, Zhu C, Cheng HL, Manis J, Whitlow S
et al. Leaky Scid phenotype associated with defective V(D)J
coding end processing in Artemis-deficient mice. Mol Cell
2002; 10: 1379–1390.
8 Pagano L, Caira M, Nosari A, Van Lint MT, Candoni A,
OffidaniM et al. Fungal
hematopoietic stem cell transplants: results of the SEIFEM
Fungine Nelle Emopatie Maligne. Clin Infect Dis 2007; 45:
9 Post MJ, Lass-Floerl C, Gastl G, Nachbaur D. Invasive
fungal infections inallogeneic
cell transplant recipients: a single-center study of 166
10 Ascioglu S, Rex JH, de Pauw B, Bennett JE, Bille J, Crokaert
F et al. Defining opportunistic invasive fungal infections in
immunocompromised patients with cancer and hematopoietic
stem cell transplants: an international consensus. Clin Infect
Dis 2002; 34: 7–14.
infectionsin recipients of
Letter to the Editor
Bone Marrow Transplantation