European Journal of Haematology 85 (58–64)
ORIGINAL ARTICLE
Sepsis in acute myeloid leukaemia patients receiving high-dose chemotherapy: No impact of chitotriosidase and mannose-binding lectin polymorphisms Anja Klostergaard1, Rudi Steffensen2, Jens K. Møller3, Niels Peterslund1, Caroline Juhl-Christensen1, Ingolf Mølle1 1
Department of Haematology, Aarhus University Hospital, Aarhus C, Denmark; 2Department of Clinical Immunology, Aalborg University Hospital, Aalborg, Denmark; 3Department of Clinical Microbiology, Aarhus University Hospital, Aarhus C, Denmark
Abstract Infections after chemotherapy often cause significant morbidity in patients with acute myeloid leukaemia (AML). Chitotriosidase (CHIT) and mannose-binding lectin (MBL) are part of the innate immune system. Polymorphism in the CHIT-coding gene (CHIT1) may be associated with Gram-negative sepsis in children with AML, and polymorphism in the MBL-coding gene (MBL2) seems to modify the risk of infections in several patient groups. The purpose of this study was to investigate the possible associations between polymorphisms in CHIT1, MBL2 and sepsis in adult patients treated with high-dose chemotherapy for AML. We included 190 patients treated with 526 cycles of chemotherapy. The follow-up period was 6 months from the diagnosis of AML. Prophylactic antibiotics were not used. We identified 604 febrile episodes with 246 episodes of sepsis. Thirty-two patients (17%) either died from infection or infection was a major concomitant factor for death. No significant associations between CHIT1 polymorphism and sepsis (P = 0.85) or death caused by sepsis (P = 0.14) were found. Furthermore, no significant associations between MBL2 polymorphism and sepsis (P = 0.76) or death caused by sepsis (P = 0.24) were observed. The severe and long-lasting neutropenia and mucositis after chemotherapy may explain why the MBL system does not protect against sepsis in patients with AML. Replacement therapy with recombinant MBL is not likely to decrease the risk of sepsis in patients with AML. Key words acute leukaemia; acute myeloid leukaemia; chitotriosidase; CHIT-coding gene; infection; innate immune system; mannose-binding lectin; MBL-coding gene; polymorphism Correspondence Anja Klostergaard, Department of Haematology, Aarhus University Hospital, Tage-Hansensgade 2, 8000 Aarhus C, Denmark. Tel: 0045 89497551; Fax: 0045 89497599; e-mail:
[email protected] Accepted for publication 8 March 2010.
Introduction
Infections are a major cause of mortality and morbidity in patients with acute myeloid leukaemia (AML) receiving high-dose chemotherapy. Fever is present in 85% of patients during the neutropenia caused by chemotherapy, and microbiologically documented infections are seen in 45% of the patients (1). Several studies demonstrated a lower risk of microbiologically documented infections, when prophylactic antibiotics were used (1–3). One meta-analyse showed reduced mortality, when fluoroquinolones were used (3), but in most studies no survival benefit was seen (1, 2). In the future the possibility of
58
doi:10.1111/j.1600-0609.2010.01443.x
risk stratifying patients may provide basis for riskadapted prophylaxis with antibiotics or possibly immune modulation. In this aspect the impact of polymorphisms of the innate immune system has been investigated in a number of studies. Accordingly, polymorphisms in chitotriosidase gene (CHIT1) and mannose-binding lectin gene (MBL2) may modify the risk of infections in various cancer patients treated with chemotherapy (4–7). Variant CHIT1 seems to be associated with increased risk of Gram-negative bacterial infections in children receiving chemotherapy for AML (6). In addition, variant CHIT1 has been associated with increased
ª 2010 John Wiley & Sons A/S
Klostergaard et al.
susceptibility to filarial infection (8). In contrast, no association between chronic disseminated candidiasis and variant CHIT1 was seen in forty patients with acute leukaemia (9). Chitotriosidase (CHIT) is a glycosylhydrolase synthesised by activated macrophages. It binds to and degrades the glycopolymer chitine. Chitine is found on the surface of fungi, parasites and possible other micro-organisms. CHIT1 is located at chromosome 1 (10). Variant CHIT1 (WM ⁄ MM) is caused by a 24-base-pair duplication. This results in abnormally spliced mRNA, leading to inactivated and low levels of CHIT compared with the wildtype (WW) (6, 11). Variant MBL2 seems to increase the risk of infections in patients with multiple myeloma after autologous stem cell transplantation (4), in children treated with chemotherapy (5) and in patients admitted to intensive care units (12, 13). MBL status pretransplantation also seems to influence the risk of infection after myeloablative allogeneic hemopoietic stem cell transplantation, whereas the importance of the donor MBL status differs in different studies (14–17). Furthermore, deficiency of MBL may contribute to severe infections in patients treated with chemotherapy for other haematological diseases (7). In patients with AML, in contrast, a previous smaller study showed no association between the serum level of mannose-binding lectin (MBL) and the risk of infections (18). MBL is a liver-derivated lectin, which is able to bind to carbohydrate structures of micro-organisms, thereby opsonizing phagocytosis and activating the complement system (19, 20). MBL2 is located at chromosome 10. Inherited MBL deficiency is caused by one of three point mutations located at codons 52, 54 and 57 in exon 1 of MBL2. These point mutations are denoted D, B and C, or just O, when considered only as the opposite to the wild-type variant A. MBL serum concentrations vary widely and, besides from the mutations in exon 1, also depend on three single nucleotide polymorphisms (SNPs) in the promoter region of MBL2. At position –221, especially the presence of the promoter allele X, instead of the Y allele, significantly down-regulates the gene expression (21, 22). Because of linkage disequilibrium, only seven common haplotypes are formed by the three promoter region SNPs and the four allelic markers in exon 1 (23). The diplotypes based on these haplotypes are classified according to their known association with a high (YA ⁄ YA, YA ⁄ XA), intermediate (XA ⁄ XA, YA ⁄ YO) or low (XA ⁄ YO, YO ⁄ YO) MBL concentration (24). The purpose of this study was to investigate, whether polymorphisms in CHIT1 and MBL2 modify the risk of sepsis, the type of sepsis and death caused by sepsis in adult patients treated with high-dose chemotherapy for AML.
ª 2010 John Wiley & Sons A/S
Polymorphisms and sepsis in AML patients
We only found a few studies concerning CHIT1 and the risk of infections in patients with AML (6, 9), and therefore aimed at confirming previous findings (6) in a larger group of adult patients. Most importantly, however, we intended to further define the role of the MBL system in patients with AML. Despite previous findings (18), the emerging possibility of substitution therapy with recombinant MBL (25), in our opinion, justified a confirmatory study in these patients, in whom targeted prophylaxis against infections could be of great importance. Methods
Patients were identified at a single centre between 1 ⁄ 11993 and 1 ⁄ 9-2004 using the Immune-Haematological Laboratory Database at Aarhus University Hospital. Inclusion criteria were age over fifteen and treatment with high-dose chemotherapy. The follow-up period was 6 month after diagnosis of AML. Patients receiving transplantation, changing to palliative treatment or transferring to another hospital were censored. Genomic DNA was extracted from diagnostic bone marrow aspirates. Mononuclear cells were isolated by Lymphoprep TM density centrifugation (Axis-Shield PoC AS, Oslo, Norway), cryo-preserved in 10% dimethylsulfoxide and stored until use at )180C. DNA was purified from approximately one million cells using the MagNa-Pure LC Robot (Roche Diagnostics, Basel, CH, Switzerland) or the TRIZOL reagent (Invitrogene Corporation, Carlsbad, CA, USA), according to the respective manufacture¢s protocol. The CHIT1 genotypes were identified using real-time polymerase chain reactions (rtPCR) as described by Boot¢s (10). MBL2 genotyping was performed using rt-PCR with TaqMan SNP Genotyping Assays (Applied Biosystems, Foster City, CA, USA). MBL2 –D (codon 52, rs 5030737) and MBL2 – C (codon 57, rs1800451) were genotyped using predesigned ⁄ validated TaqMan Genotyping assay id C___2336610_10 and C___2336608_20, respectively (Applied Biosystems, Foster City, CA, USA). Previously described assays were used for genotyping MBL2 –B (codon 54, rs1800450), MBL2 –X ⁄ Y (-221, rs7096206), MBL2-H ⁄ L (-550, rs11003125) and MBL2-P ⁄ Q (+4, rs12780112) (26, 27). For all TaqMan assays, DNA amplification was carried out in 5-lL volume polymerase chain reaction (PCR) containing 20 ng DNA, 0.9 lm primers and 0.2 lm probes (final concentrations) amplified in 384-well plates. Reactions were performed with the following protocol on GeneAmp PCR 9700 or a 7900 HT Sequence Detection System: 95C for 10 min, followed by 40 cycles of 95C for 15 s and 60C for 1 min. To determine genotypes, end-point fluorescence was read on the 7900 HT Sequence Detection
59
Polymorphisms and sepsis in AML patients
Klostergaard et al.
System using sds version 2.3 software (Applied Biosystems, Foster City, CA, USA). Infections were identified using clinical records, blood culture results compiled from the Laboratory Information System MADS at the Department of Microbiology, and biochemical data from the Department of Biochemistry. Infections were classified as ‘febrile episodes’ according to the general practice in the department, when the temperature exceeded 38.5C without other obvious reason for fever, e.g. cytarabine infusion or reaction to blood transfusion. First-line antibiotic treatment was an aminoglycoside in combination with a broad-spectrum beta-lactam. Prophylactic antibiotics after chemotherapy were not used. Diagnostic criterion for ‘sepsis’ was a febrile episode and a significant pathogen grown from a blood culture. Single positive blood cultures with coagulase-negative staphylococci, coryneforms, bacillus species and propionibacteria were regarded as contaminated. Death of patients having a positive blood culture within the last 14 d, persistently elevated acute phase reactant or fever, and no other obvious cause of death was classified as ‘death caused by sepsis’. Death of patients having clinical or autopsy signs of infection, no positive blood culture during the last 14 d, persistently elevated acute phase reactant or fever and no other obvious cause of death was classified as ‘death probably because of infection’. We analysed the median age, number of cycles of chemotherapy and neutrophil count with Kruskal–Wallis¢ and Mann–Whitney’s tests. The gender and reason for exclusion (death ⁄ censored ⁄ followed 6 month) were compared by chi-square test. We used Fisher’s exact test to compare the outcome (sepsis, type of sepsis and death caused by sepsis) in the different groups. The proportion of patients without a septic episode was analysed using Statistical Package for the Social Sciences (SPSS) computer program. The number of septic episodes and the number of patients with ⁄ without sepsis were tested by chi-square test. A P-value of
