03

do

Experimental Hematology 34 (2006) 1435–1441

Mannan-binding lectin pathway deficiencies and invasivefungal infections following allogeneic stem cell transplantation

Miquel Granella, Alvaro Urbano-Ispizuaa, Belen Suarezb,Montserrat Roviraa, Francesc Fernandez-Avilesa, Carmen Martıneza, Mar Ortegac,

Carla Uriburua, Anna Gayaa, Josep Ma Ronceroa, Alfons Navarrod, Enric Carrerasa,Josep Mensac, Jordi Vivesb, Ciril Rozmana, Emili Montserrata, and Francisco Lozanob

aDepartments of Hematology; bImmunology; cInfectious Diseases, Hospital Clınic of Barcelona, Barcelona, Spain;dInstitut d’Investigacions Biomediques August Pi i Sunyer (IDIBAPS), Faculty of Medicine, University of Barcelona, Barcelona, Spain

(Received 13 February 2006; revised 24 April 2006; accepted 7 June 2006)

Objective. The Mannan-binding lectin (MBL) pathway involves recognition of fungal surfacesby MBL and cleavage of C2 and C4 by MBL-associated serine protease (namely, MASP-2).Recent data show that MBL pathway deficiency might result not only from polymorphismsof the MBL2 gene but also of MASP2. The aim of the study was to assess whether polymor-phisms of these genes are associated with invasive fungal infections (IFIs) following allogeneicstem cell transplantation (allo-SCT).

Methods. The promoter and the exon 1 of MBL2 and the exon 3 of MASP2 were sequenced in106 donor-recipient pairs from HLA-identical sibling allo-SCTs performed in a singleinstitution.

Results. Ten percent of the donors and 11% of the recipients carried the MBL-low (O/O,LXA/O) genotypes; 7% of the donors and 3% of the recipients were heterozygous for theMASP2 Asp105Gly variant. Factors associated with a higher probability of IFIs were donor’sMBL-low genotype (38% vs 12%, p [ 0.01), recipient’s MASP2 variant (67% vs 14%, p [0.01), and acute graft-versus-host disease (GVHD) grades II to IV (27% vs 11%, p [0.04); in the multivariate analysis MBL-low genotype (relative risk [RR] 7.3, p [ 0.003),MASP2 variant (RR 6.4, p [ 0.002), and acute GVHD II to IV (RR 3.8, p [ 0.02) retainedindependent prognostic value.

Conclusion. These results show for the first time that polymorphisms responsible for not onlyMBL but also MASP-2 deficiency are independent predictive factors for IFI after allo-SCT. � 2006 International Society for Experimental Hematology. Published by Elsevier Inc.

Invasive fungal infection (IFI) is one of the most life-threat-ening complications after allogeneic stem cell transplanta-tion (allo-SCT) [1–3]. Length of neutropenia, presence ofgraft-versus-host disease (GVHD), and steroid and/or otherimmunosuppressant treatments have been identified as im-portant risk factors for developing IFI following allo-SCT[4,5]. The importance of individual non-HLA encodedgenetic variability in the development of infections recentlyhas been recognized [6]. Thus, polymorphisms of severalgenes, namely myeloperoxidase (MPO) and mannan-bind-

Offprint requests to: Alvaro Urbano-Ispizua, M.D., Ph.D., Department

of Hematology, Hospital Clınic, Villarroel 170, 08036 Barcelona, Spain;

E-mail: aurbano@clinic.ub.es

01-472X/06 $–see front matter. Copyright � 2006 International Society fo

i: 10.1016/j.exphem.2006.06.005

ing lectin (MBL) have been associated with a higher inci-dence of infections [7,8]. These genetic polymorphismscould also influence the susceptibility to IFI afterallo-SCT, but this possibility has not been fully addressed.

MBL, a plasma protein, belongs to the collectin familyand is characterized by an N-terminal collagen-like stalkdomain and a C-terminal C-type carbohydrate recognitiondomain [9]. The MBL subunits associate to form multi-meric complexes capable of recognizing carbohydrate pat-terns (mannose, fucose, glucose, and N-acetylglucosamine)displayed at high density in many microorganisms (virus,fungi, bacteria, parasites) but not in mammalian cells[10]. Several fungi species (including Aspergillus fumiga-tus, Candida albicans, and Criptococcus neoformans) are

r Experimental Hematology. Published by Elsevier Inc.

1436 M. Granell et al./ Experimental Hematology 34 (2006) 1435–1441

among the microorganisms that show high to intermediatebinding to MBL [11,12]. Upon binding, MBL activatesthe complement system, leading to opsonization or directkilling of microorganisms. Complement activation viaMBL is achieved through association with two MBL-asso-ciated serine proteases (MASP-1/3 and MASP-2/MAPp19),of which MASP-2 appears to be the most important [9]. In-terestingly, MASP-2 also associates with L- and H-ficolin,likely playing a key role in the lectin pathway of comple-ment activation [13].

MBL deficiency is one of the most common human im-munodeficiencies. It was first reported as an opsonophago-cytosis defect in which patients failed to opsonize baker’syeast (Saccharomyces cerevisiae). Since then, MBL defi-ciency has been associated with increased susceptibilityto many infectious diseases, particularly when adaptive im-munity is compromised (e.g., in early childhood [14,15] orfollowing chemotherapy [16–18]). Low serum levels ofMBL are found in association with single nucleotide poly-morphisms (SNPs) in the promoter and the structuralgene-coding region (exon 1) of the MBL2 gene (MBL1 isa pseudogene). Three SNPs at codons 52, 54, and 57 en-code for variant alleles D, B, and C, respectively, whichdramatically reduce the functional MBL levels by impair-ing the assembly of MBL monomers into functional oligo-mers [19]. These variants are collectively named O, and Aindicates the wild-type variant. Three additional SNPs atpositions �551 (H/L), �221 (X/L), and þ4 (P/Q) in the50-flanking region of the MBL2 gene also influence the se-rum MBL levels in individuals with the wild-type codingregion [20]. Of these, two promoter haplotypes, HY, whichis associated with high levels of MBL, and LX, which is as-sociated with low MBL levels, appear to be the most impor-tant. Through the combination of structural gene andpromoter polymorphisms, MBL concentration can varyconsiderably (a thousand-fold) in apparently healthy indi-viduals [21]. The correlation between genotype and pheno-type is, however, relatively strong and so MBL-sufficientlevels are associated with YA/YA, YA/LXA, LXA/LXA,and YA/O genotypes, and MBL-low levels are associatedwith O/O and LXA/O genotypes [19–21].

An inherited deficiency of MASP-2 has been recentlydescribed in a patient with severe autoimmune and infec-tious disease [22]. This deficiency is due to a mutation inexon 3 of MASP2, resulting in an amino acid substitutionat position 105 in the mature protein (Asp105Gly). The mu-tation is present in 5.5% of healthy Danish individuals andis associated with low serum levels of MASP-2 [23].Accordingly, reduced MBL pathway function has beenreported in individuals heterozygous for this specificMASP2 gene mutation [24].

No single study has analyzed the relationship betweenMBL2 and MASP2 gene polymorphisms and IFIs in immu-nocompromised patients. With this background, we investi-gated the association of donor and recipient variant alleles

of MBL2 and MASP2 with the incidence of IFI in a group of106 allogeneic HLA-identical sibling transplants performedin a single institution.

Material and methods

Patients and donorsThe basis of this study was 106 donor-patient sibling pairs under-going conventional myeloablative allo-SCTs at the Hospital Clinicof Barcelona from 1995 to 2002. Donor and patient characteristicsare given in Table 1. All donor-patient pairs were of Caucasian or-igin. All patients received standard supportive care, including iso-lation in high-efficiency particulate-air-filtered rooms andadministration of prophylactic antimicrobials; these includedoral ciprofloxacin (500 mg/12 h) until neutrophil count recovery,oral fluconazole (50 mg/d) until 2 months after transplantation,and acyclovir (250 mg/m2/12 h intravenously or 800 mg/12 horal) if the patient had a herpes simplex virus positive serology,until 1 month after transplantation. Primary prophylaxis with itra-conazole (200 mg/12 h) was given to patients with a second caseof acute GVHD requiring steroids. Secondary prophylaxis withamphotericin B (0.3 mg/kg/48 h) was given to patients who hadan IFI previous to transplantation. The Ethics Committee of thehospital approved the study. All patients and donors consentedto obtain peripheral blood samples.

Table 1. Patient characteristics

Patient characteristics n (%)

Sex

Male 54 (50.9)

Female 52 (49.1)

Mean age (range) 39 (20–59)

Diagnosis

AML 35 (33.0)

NHL 17 (16.0)

CML 16 (15.1)

ALL 13 (12.3)

MM 8 (7.5)

MDS 8 (7.5)

CLL 6 (5.6)

ALBiph 2 (1.9)

Histiocytose X 1 (0.9)

Phase of the disease

Early 50 (47.2)

Advanced 56 (52.8)

Conditioning regimen

CyTBI 90 (84.9)

BuCy 9 (8.5)

Others 7 (6.6)

GVHD prophylaxis

CsA and T-cell depletion (CD34þ selection) 68 (64.1)

CsA and MTX 29 (27.4)

Others 9 (8.5)

AML, acute myeloblastic leukemia; NHL, non-Hodgkin’s lymphoma;

CML, chronic myeloid leukemia; ALL, acute lymphoblastic leukemia;

MM, multiple myeloma; MDS, myelodisplastic syndrome; CLL, chronic

lymphocytic leukemia; ALBiph, acute biphenotypic leukemia; HL; Cy,

cyclophosphamide; TBI, total body irradiation; Bu, busulphan; GVHD,

graft-versus-host disease; CsA, cyclosporine A; MTX, methotrexate.

1437M. Granell et al. / Experimental Hematology 34 (2006) 1435–1441

MBL2 and MASP2 genotypingBlood samples were obtained from 102 recipients, 106 donors,and 104 healthy volunteer blood donors from the geographicalarea of Barcelona. Genomic DNA was extracted from ethylenedi-amine-tetraacetic acid (EDTA)-treated whole blood samples byusing the QIAamp DNA blood mini kit following the manufac-turer’s instructions (QIAGEN GmbH, Hilden, Germany) andthen stored at �20�C until used. EDTA-treated plasma sampleswere obtained in some cases and kept frozen at �20�C until used.

Six SNPs in the MBL2 gene (�550 G/C,�221 C/G, 4 C/T, codon52 CGT/TGT, codon 54 GGC/GAC, and codon 57 GGA/GAA)and the Asp105Gly SNP in the MASP2 gene were analyzed by asequencing-based typing method [25]. Briefly, a 969-bp fragmentencompassing from the promoter to the end of exon 1 of MBL2was obtained by polymerase chain reaction (PCR) amplification us-ing the sense 50-GGGGAATTCCTGCCAGAAAGT-30 and anti-sense 50-CATATCCCCAGGCAGTTTCCTC-30 primers and theExpand 20kbPLUS PCR System (Roche Diagnostics GmbH, Man-nheim, Germany). Similarly, a 354-bp fragment from the exon 3 ofMASP2 was PCR-amplified by using the sense 50-GCGAGTAC-GACTTCGTCAAGG-30 and antisense 50-CTCGGCTGCATA-GAAGGCCTC-30 oligonucleotides and the Expand High FidelityPCR System (Roche Diagnostics GmbH, Mannheim, Germany).The cycling conditions were 94�C for 8 minutes; 35 cycles of94�C for 45 seconds, 58�C for 30 seconds, and 72�C for 90 seconds;and finally 72�C for 10 minutes. Five microliters of the resultingPCR reactions were treated with ExoSAP-IT (USB Corporation,Cleveland, OH, USA) and then subjected to direct sequencingwith the BigDye Terminator v1.1 Cycle Sequencing Kit (AppliedBiosystems, Warrington, UK) following the manufacturer’s instruc-tions with the sense and antisense gene-specific primers mentionedabove.

DefinitionsIFIs were classified according to the recommendations fromAscioglu et al. [26]: possible IFIs were cases of prolonged feverwithout microbiologic documentation not responding to broad-spectrum antibiotics and with radiologic images suspicious of fun-gal infection, occurring after neutrophil recovery; probable IFIswere cases of fever not responding to broad-spectrum antibioticswith isolation of filamentous mould on sputum or with two con-secutive galactomannan antigen positive on blood; proven IFIswere cases of fever with histologic documentation of filamentousfungi or isolation of yeast in blood. Individuals carrying O/O orLXA/O genotypes were considered as ‘‘MBL-low producers,’’and the remaining MBL genotypes were considered as ‘‘MBL-sufficient producers.’’

Statistical analysisCumulative incidence of IFI was calculated using relapse or deathwithout IFI as competing risks and statistically compared byGray’s method [27]. All reported p values are two-sided, and a sig-nificance level of a 5 0.05 was used. All prognostic variables inthe univariate analysis with a p value # 0.05 (Gray’s method)were included for the multivariate analysis to eliminate the redun-dancy among highly correlated characteristics, each of which maybe individually significant. This was performed using a multivari-ate regression model that takes into account competing risks [28](cmprsk package of the R Foundation for Statistical Computingat www.R-project.org). The proportional hazard assumption of

the model was checked separately for each covariate beforeperforming the regression analysis (SPSS version 12.0 statisticalsoftware).

Results

MBL2 and MASP2 allele frequenciesIndividuals were genotyped for the presence of SNPsreported in the promoter and exon 1 of the MBL2 gene;of them, 102 were recipients undergoing allo-SCT, 106were their HLA-identical sibling donors, and 104 corre-sponded to healthy blood donors obtained through theblood bank of our institution. No significant differenceswere found in the prevalence of polymorphisms in the threegroups. Allele frequencies of MBL2 gene in these groupsare described in Table 2. Eleven (11%) recipients, 11(10%) donors, and 16 (15%) healthy volunteers had MBL-low genotypes (Table 2).

Ninety-five recipients, 103 donors, and 104 healthyblood donor volunteers were also genotyped for SNPs inthe exon 3 of MASP2. Exon 3 Asp105Gly polymorphismwas detected in 3 (3%) recipients, 7 (7%) donors, and 3(3%) healthy blood donor volunteers (Table 2). No signifi-cant differences in the prevalence of polymorphisms in thethree groups were seen.

Clinical characteristics and infectionsAfter a median follow-up of 24 months (0.6–77 months),overall survival in the overall group was 52% (95% confi-dence interval [CI]: 39.0–65.3), relapse rate was 30%(95% CI: 46.6–15.3), and transplant related mortality24% (95% CI: 15.7–33.3). Twenty-two patients (21%)

Table 2. MBL2 and MASP2 frequency

Patients,

n (%)

Donors, n

(%)

Healthy volunteers,

n (%)

MBL2 haplotype

HYPA 68 (33.3) 63 (29.7) 52 (25.0)

LYQA 49 (24.0) 58 (27.3) 50 (24.0)

LXPA 38 (18.6) 37 (17.4) 37 (17.8)

LYPB 26 (12.7) 28 (13.2) 35 (16.8)

LYPA 10 (4.9) 12 (5.7) 17 (8.2)

HYPD 11 (5.4) 10 (4.7) 13 (6.2)

LYQC 2 (1.0) 4 (1.9) 4 (1.9)

MBL2 genotypes

Sufficient

A/A 63 (61.8) 64 (60.4) 58 (55.8)

A/O 23 (22.5) 27 (25.4) 27 (25.9)

LXA/LXA 5 (4.9) 4 (3.8) 3 (2.9)

Low

O/O 5 (4.9) 4 (3.8) 8 (7.7)

LXA/O 6 (5.9) 7 (6.6) 8 (7.7)

MASP2 Asp105Gly

Wild-type 92 (96.8) 96 (93.2) 101 (97.1)

Heterozygous 3 (3.2) 7 (6.8) 3 (2.9)

1438 M. Granell et al./ Experimental Hematology 34 (2006) 1435–1441

developed acute graft-versus-host disease (aGVHD) gradesI to -IV. Eighty-six (81%) patients survived more than 100days and were valuable for chronic graft-versus-host dis-ease (cGVHD); of these, 29 developed cGVHD (33%), 14limited (16%) and 15 extensive (17%).

Fifteen (14%) patients experienced a total of 16 cases ofIFI after the transplant. Of these, 8 (50%) were possible, 7(44%) were probable, and 1 (6%) was proven. The meantime of presentation of this complication was 94 days(range 8–300) after the transplant. Of the seven probablecases, in three cases Aspergillus was detected in sputum,and in four cases galactomanann antigen was detected inblood with increasing levels in two consecutive measures.One patient was diagnosed of a proven IFI by isolation ofCandida tropicalis in blood cultures and a clinical pictureof chronic systemic candidiasis. In this patient, galacto-mannan antigen was repeatedly detected in blood. IFI wasconsidered the cause of death in five patients.

Fifty (47%) recipients experienced a total of 73 cases ofbacteriemia; mean time of appearance of the first case wasday 44 after the transplant (range, 2–300 days). Fifteen(14%) recipients experienced a total of 20 cases of bacterialpneumonia; mean time was day 317 after transplant (range,1–1100 days). Forty-one (39%) recipients experienced a to-tal of 54 cases of cytomegalovirus (CMV) antigenemia;mean time of appearance of CMV reactivation was 61day after the transplant (range, 30–270 days). No associa-tion was found between the rate of bacterial and viral infec-tions with neither MBL nor MASP2 genotypes (data notshown).

Risk factors of IFIIn the univariate analysis, three factors were associated withIFI: donor MBL-low genotype, recipient MASP2 As-p105Gly variant, and aGVHD II to IV (Table 3). Of the11 recipients whose donors had an MBL-low genotype, 4(36%) experienced an IFI, while in the 95 patients whosedonors had an MBL-sufficient genotype, 11 (12%) experi-enced IFI (cumulative incidence 38% vs 12%, p 5 0.018;Fig. 1A). Of note, of the seven patients whose donors hadthe MBL-low LXA/O genotype, four developed IFI, andin the remaining 99 patients, 11 developed this complica-tion (57% vs 11%, p ! 0.0004). When we restricted theanalysis to the MBL-sufficient group, those recipients

Table 3. Risk factors of IFI (univariate)

Risk factor

Cumulative

incidence (%) p

Donor MBL-low genotype 38 vs 12 0.018

Recipient MASP2 Asp105Gly variant 67 vs 15 0.010

aGVHD II–IV 27 vs 11 0.048

Unmanipulated grafts 22 vs 10 0.079

Extensive cGVHD 27 vs 12 0.14

Advanced phase of disease 18 vs 11 0.29

whose donors carried genotypes resulting in MBL-high se-rum levels (A/A) had an incidence of IFI of 11%, whereasthe group of recipients whose donors carried genotypes re-sulting in MBL-intermediate levels (A/O, LXA/LXA) hadan incidence of IFI of 20% (log-rank, p 5 0.06; Fig. 1B).

Of the three recipients with the Asp105Gly variant ofMASP2, two (67%) experienced an IFI, whereas of theremaining 95 recipients only 13 (13%) experienced this

Figure 1. (A): Cumulative incidence of IFI according to donors MBLgroup. MBL-low indicates LXA/O and O/O genotypes; MBL-sufficient

indicates A/A, A/O, and LXA/LXA genotypes. (B): Cumulative incidence

of IFI according to donors MBL group. MBL-low indicates donor LXA/O

and O/O genotypes; MBL-intermediate indicates donor MBL A/O and

LXA/LXA genotypes; MBL-high indicates donor A/A genotype.

1439M. Granell et al. / Experimental Hematology 34 (2006) 1435–1441

complication (cumulative incidence 67% vs 15%, p 5 0.01;Fig. 2). Donor MBL haplotypes of the three patients withthe Asp105Gly variant of MASP2 were LYPA/LYPA,LYPA/HYPA, and LYQA/LYQA, corresponding to anMBL-sufficient (A/A) genotype.

The following characteristics were included in the mul-tivariate analysis for risk factors of IFI: donor MBL-low ge-notype, aGVHD grades II to IV, and MASP2 Asp105Glyvariant. Using this approach, donor MBL-low genotype (rel-ative risk [RR] 7.3, 95% CI 1.9–27.3, p 5 0.003), MASP2Asp105Gly variant (RR 6.4, 95% CI 2.0–20.6, p 5 0.002),and aGVHD II to IV (RR 3.8, 95% CI 1.2–12.1, p 5 0.02)remained significant (Table 4).

DiscussionAllo-SCT is increasingly used worldwide for treatinghematologic and nonhematologic diseases [29]. Unfortu-nately, allo-SCT has a procedure-related mortality rangingbetween 18 and 50%, the principal cause of death being in-fections. IFI is a frequent and often mortal infectious com-plication after allo-SCT. Results of this report show, for thefirst time, that donor and/or recipient genetic variants ofMBL2 and MASP2 are independent risk factors for develop-ing IFI after allo-SCT. In vitro studies have demonstratedthat MBL is capable of binding and favors opsonizationof Aspergillus fumigatus, Candida albicans, and Criptococ-cus neoformans, the main microorganisms involved in IFI[10,12]. Moreover, clinical studies have shown an associa-tion of MBL gene polymorphisms with chronic necrotizing

Figure 2. Cumulative incidence of IFI according to recipient MASP2 ge-

notype. MASP2 Asp105Gly indicates recipient MASP2 Asp105Gly variant.

pulmonary aspergillosis [30] and with recurrent vulvovagi-nal candidiasis [31].

In this series of allo-SCTs from HLA-identical siblingdonors, the incidence of IFI was of 16% (95% CI: 8.35–23.6), very similar to that observed in other reports ofallo-SCT [1]. IFIs were classified according to standard rec-ommendations [26]. To avoid the inclusion of misdiagnosesof IFI, only cases with prolonged fever without microbio-logic documentation, not responding to broad-spectrum an-tibiotics, with compatible radiology, and occurring afterneutrophil recovery were considered as ‘‘possible IFI.’’ Be-sides, an independent expert in infectious diseases reviewedthese cases before results of MBL2 and MASP2 genotypingwere available.

Recent conclusive evidence on the important role playedby MBL in the defense against infections comes from MBL-A and MBL-C double knockout mice. These MBL-null miceare highly susceptible to intravenous inoculation of Staph-ylococcus aureus compared with wild-type mice [32]. Inter-estingly, intraperitoneum inoculation of S. aureus resultedin enhanced infectious complications only when MBL-null mice were rendered neutropenic [33]. Information onsusceptibility to fungal infections in this mouse model is,however, still due. In humans, the potential association ofMBL2 gene polymorphisms responsible for low MBL se-rum concentrations with infectious complications followingstem cell transplantation [7,8,33] is currently a matter ofdebate. Mullighan et al. [8] was the first to find an associ-ation of donor MBL genotype with the incidence of majorinfections posttransplant. However, most of the infectionsanalyzed were bacterial and viral, and no specific associa-tion with fungal infections was reported, likely due to thesmall number of patients experiencing IFI for meaningfulstatistical analysis. Rocha et al. [7] could not find an asso-ciation of the coding (exon 1) mutations in MBL2 gene withthe incidence of the first episode of bacterial, viral, andfungal infection. Unfortunately, these authors did not ana-lyze polymorphisms in the promoter region of the MBL2gene, thus missing the contribution of MBL-low LXA/Oindividuals.

The MASP2 Asp105Gly variant showed a low allele fre-quency in our population study compared with Danes. Nev-ertheless, we found an association of patients carrying thisvariant with IFI in both the univariate (Fig. 2) and multivar-iate analysis (Table 4). Of note, the effect of MASP2 varianton the incidence of IFI was independent of the effect of

Table 4. Risk factors of IFI (multivariate)

Risk factor RR 95% CI p

Donor MBL-low genotype 7.3 1.9–27.3 0.003

Recipient MASP2 Asp105Gly variant 6.4 2.0–20.6 0.002

aGVHD II–IV 3.8 1.2–12.1 0.02

RR, relative risk; CI, confidence interval.

1440 M. Granell et al./ Experimental Hematology 34 (2006) 1435–1441

MBL low-genotype. Although this result deserves furthervalidation due to the low number of recipients with theMASP2 variant in this series, it is in agreement with the ma-jor role played by MASP2 in the MBL pathway function[24,34]. It also represents the first study showing the clini-cal relevance of MASP2 polymorphisms resulting in partialMASP-2 deficiency.

The results of our study show that independent geneticfactors from both donors and recipients play an importantrole in the outcome of HLA-matched sibling bone marrowtransplantation. The results of our study show that recipientMASP2-low and donor MBL-low genotypes influence theincidence of major infections after allo-SCT. The later arein agreement with the results of Mullighan et al. [8] and im-ply that donor cells engrafted after allo-SCT could playa role in the synthesis of MBL. MBL and its associated ser-ine proteases are acute phase proteins mainly produced inthe liver, but substantial extrahepatic transcription hasbeen recently reported in mouse and humans [35,36]. Infact, MBL2 transcripts are detected in human cord bloodmononuclear cells and differentiated cultured monocytesbut not in lymphocytes [36]. Kilpatrick et al. [37] reportedtwo patients with low MBL serum levels transplanted fromdonors with normal MBL serum levels. Although completedonor chimerism was obtained in both cases, serum MBLlevels remained low after the transplant in patients. Theyconcluded that MBL sufficiency of the donor does not cor-rect a recipient’s MBL insufficiency. Our findings and thoseof Mullighan et al. [8], showing that MBL-low donor geno-type is a risk factor for IFI and major bacterial infections,respectively, would argue against Kilpatrick’s conclusion.Recent studies indicate that regulation of MBL2 andMASP2 expression is more complex than previously antic-ipated and that local extrahepatic expression of these twogenes may be relevant in local immune defense, particu-larly in restricted areas of the gastrointestinal and reproduc-tive tracts [36]. Interestingly, C. albicans, which is acommon cause of morbidity and mortality in immunocom-promised individuals, is part of the normal commensal florain the human gastrointestinal and female lower genital tracts.

In conclusion, we show that MBL pathway dysfunctiondue to genetically determined deficiencies of MBL andMASP-2 influence the outcome of allo-SCT by increasingthe susceptibility to IFIs. Recipient and donor MBL-low ge-notypes have been previously associated with major bacte-rial infections after autologous [35] and allo-SCT [8], butthis is the first study showing its association to IFI. Thelater also applies to MASP-2 deficiency, although in thiscase represents the first study showing the clinical relevanceof this condition. Thus, it should be emphasized thatMASP-2 should not be neglected in clinical studies of theMBL pathway. Finally, the findings here reported haveclinical implications in at least two areas: selection of thedonor and selection of patients for primary antifungalprophylaxis.

AcknowledgmentsThis work was supported by RTICCC03/10 from Instituto deSalud Carlos III, by 2003-020510 from Fundacio La Marato deTV3, FIS2006 and 02/350, and BM05-219-0 from Fundacio LaCaixa. M.G. is supported by a grant from the Spanish Society ofHaematology (AEHH). We are deeply indebted to Dr. MyriamLabopin for revising the statistical analysis of this article.

References1. Martino R, Subira M. Invasive fungal infections in haematology: new

trends. Ann Hematol. 2002;1:233–243.

2. Herbrecht R, Denning DW, Patterson TF, et al. Voriconazole versus

Amphotericin B for primary therapy of invasive aspergillosis.

N Engl J Med. 2002;347:408–415.

3. Cornely OA, Ullmann AJ, Karthaus M. Evidence-based assessment of

primary antifungal prophylaxis in patients with hematological malig-

nancies. Blood. 2003;101:3365–3372.

4. Wald A, Leisenring W, van Burik JA, Bowden RA. Epidemiology of

Aspergillus infection in a large cohort of patients undergoing bone

marrow transplantation. J Infect Dis. 1997;175:1459–1466.

5. Baddley JW, Stroud TP, Salzman D, Pappas PG. Invasive mold infec-

tions in allogeneic bone marrow transplant recipients. Clin Infect Dis.

2001;32:1319–1324.

6. Dickinson AM, Middleton PG. Beyond the HLA typing age: genetic

polymorphisms predicting transplant outcome. Blood Rev. 2005;19:

333–340.

7. Rocha V, Franco RF, Porcher R, et al. Host defense and inflammatory

gene polymorphisms are associated with outcomes after HLA-identi-

cal sibling bone marrow transplantation. Blood. 2002;100:3908–3918.

8. Mullighan CG, Heatley S, Doherty K, et al. Mannose-binding lectin

gene polymorphisms are associated with major infection following

allogeneic hemopoetic stem cell transplantation. Blood. 2002;99:

3524–3529.

9. Guardia A, Lozano F. Mannose-binding lectin deficiencies in infec-

tions and inflammatory disorders. Reviews in Medical Microbiology.

2003;14:41–52.

10. Jack DL, Klein NJ, Turner MW. Mannose-binding lectin targeting the

microbial world for complement attack and opsonophagocytosis.

Immunol Rev. 2001;180:86–99.

11. Cross CE, Bancroft JG. Ingestion of acapsular Cryptococcus neofor-

mans occurs via mannose and beta-glucan receptors, resulting in cyto-

kine production and increased phagocytosis of the encapsulated form.

Infect Immun. 1995;63:2604–2611.

12. Neth O, Jack DL, Dodds AW, Holzel H, Klein NJ, Turner MW. Man-

nose-binding lectin binds to a range of clinically relevant microorgan-

isms and promotes complement deposition. Infect Immun. 2000;68:

688–693.

13. Holmskov U, Thiel S, Jensenius JC. Collections and ficolins: humoral

lectins of the innate immune defense. Annu Rev Immunol. 2003;21:

547–578.

14. Summerfield JA, Sumiya M, Levin M, Turner MW. Association of

mutations in mannose binding protein gene with childhood infection

in consecutive hospital series. Br Med J. 1997;314:1229–1232.

15. Koch A, Melbye M, Sorensen P, et al. Acute respiratory tract infec-

tions and mannose-binding lectin insufficiency during early childhood.

J Am Med Assoc. 2001;285:1316–1321.

16. Peterslund NA, Koch C, Jensenius J, Thiel S. Associations between

deficiency of mannose-binding lectin and severe infections after che-

motherapy. Lancet. 2001;358:637–638.

17. Neth O, Hann I, Turner MW, Klein NJ. Deficiency of mannose-bind-

ing lectin and burden of infection in children with malignancy: a pro-

spective study. Lancet. 2001;358:614–618.

1441M. Granell et al. / Experimental Hematology 34 (2006) 1435–1441

18. Eisen DP, Minchinton RM. Impact of mannose-binding lectin on sus-

ceptibility to infectious diseases. Clin Infect Dis. 2003;37:1496–1505.

19. Garred P, Larsen P, Madsen HO, Koch C. Mannose-binding lectin

deficiency-revisited. Mol Immunol. 2003;40:73–84.

20. Madsen HO, Garred P, Thiel S, et al. Interplay between promoter and

structural gene variants control basal serum levels of mannan-binding

protein. J Immunol. 1995;73:1–13.

21. Steffensen R, Thiel S, Varming K, Jersild C, Jensenius JC. Detection

of structural gene mutations and promoter polymorphisms in the man-

nam-binding lectin (MBL) gene by polymerase chain reaction with se-

quence-specific primers. J Immunol Methods. 2000;241:33–42.

22. Stengaard-Pedersen K, Thiel S, Gadjeva M, et al. Inherited deficiency

of mannan-binding lectin-associated serine protease 2. N Engl J Med.

2003;349:554–560.

23. Moller-Kristensen M, Jensenius JC, Jensen L, et al. Levels of mannan-

binding lectin-associated serine protease-2 in healthy individuals.

J Immunol Methods. 2003;282:159–167.

24. Carlsson M, Sjoholm AG, Eriksson L, et al. Deficiency of the mannan-

binding lectin pathway of complement and poor outcome in cystic

fibrosis: bacterial colonization may be decisive for a relationship.

Clin Exp Immunol. 2005;139:306–313.

25. Lozano F, Suarez B, Munoz A, et al. Novel MASP2 variants detected

among North African and Sub-Saharan individuals. Tissue Antigens.

2005;66:131–135.

26. Ascioglu S, Rex JH, de Pauw B, et al. Defining opportunistic invasive

fungal infections in inmunocompromised patients with cancer and

hematopoietic stem cell transplants: an international consensus. Clin

Infect Dis. 2002;34:7–14.

27. Gray RJ. A class of K-sample test for comparing the cumulative inci-

dence of a competing risk. Annals of Statistics. 1988;16:1141–1154.

28. Pepe MS, Mori M, Kaplan-Meyer. Marginal or conditional probability

curves in summarizing competing risk failure time data? Statistical

Medicine. 1993;12:187–200.

29. Gratwohl A, Baldomero H, Horisberger B, Schmid C, Passweg J,

Urbano-Ispizua A. Current trends in hematopoetic stem cell transplan-

tation in Europe. Blood. 2002;100:2374–2386.

30. Crosdale DJ, Poulton KV, Ollier WE, Thompson W, Denning DW.

Mannose-binding lectin gene polymorphisms as a susceptibility factor

for chronic necrotizing pulmonary aspergillosis. J Infect Dis. 2001;

184:653–656.

31. Babula O, Lazdane G, Kroica J, Ledger WJ, Witkin SS. Relation

between recurrent vulvovaginal candidiasis, vaginal concentrations

of mannose-binding lectin, and mannose-binding lectin gene polymor-

phism in Latvian women. Clin Infect Dis. 2003;37:733–737.

32. Shi L, Takahashi K, Dundee J, et al. Mannose-binding lectin-deficient

mice are susceptible to infection with Staphylococcus aureus. J Exp

Med. 2004;10:1379–1390.

33. Horiuchi T, Gondo H, Miyagawa H, et al. Association of MBL gene

polymorphisms with major bacterial infection in patients treated

with high-dose chemotherapy and autologous PBSCT. Genes Immun.

2005;6:162–166.

34. Sorensen R, Thiel S, Jensenius JC. Mannan-binding-lectin-associated

serin proteases, characteristics and disease associations. Springer

Semin Immunopathol. 2005;27:299–319.

35. Wagner S, Lynch NJ, Walter W, Schwaeble WJ, Loos M. Differential

expression of the murine mannose-binding lectins A and C in lym-

phoid and nonlymphoid organs and tissues. J Immunol. 2003;170:

1462–1465.

36. Seyfarth J, Garred P, Madsen HO. Extra-hepatic transcription of

the human mannose-binding lectin gene (mbl2) and the MBL-associ-

ated serine protease 1-3 genes. Mol Immunol; 2005; Epub ahead of

print.

37. Kilpatrick DC, Stewart K, Allan EK, McLintock LA, Holyoake TL,

Turner ML. Successful haemopoietic stem cell transplantation does

not correct mannan-binding lectin deficiency. Bone Marrow Trans-

plant. 2005;35:179–181.