UBC Faculty Research and Publications

Novel polymorphism of interleukin-18 associated with greater inflammation after cardiac surgery Shaw, David M; Sutherland, Ainsley M; Russell, James A; Lichtenstein, Samuel V; Walley, Keith R Jan 29, 2009

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-13054_2008_Article_7131.pdf [ 360.09kB ]
JSON: 52383-1.0223430.json
JSON-LD: 52383-1.0223430-ld.json
RDF/XML (Pretty): 52383-1.0223430-rdf.xml
RDF/JSON: 52383-1.0223430-rdf.json
Turtle: 52383-1.0223430-turtle.txt
N-Triples: 52383-1.0223430-rdf-ntriples.txt
Original Record: 52383-1.0223430-source.json
Full Text

Full Text

Available online http://ccforum.com/content/13/1/R9Open AccessVol 13 No 1ResearchNovel polymorphism of interleukin-18 associated with greater inflammation after cardiac surgeryDavid M Shaw, Ainsley M Sutherland, James A Russell, Samuel V Lichtenstein and Keith R WalleyCritical Care Research Laboratories, Heart + Lung Institute, University of British Columbia, 1081 Burrard Street, Vancouver, BC, V6Z 1Y6, CanadaCorresponding author: Keith R Walley, kwalley@mrl.ubc.caReceived: 3 Sep 2008 Revisions requested: 28 Oct 2008 Revisions received: 1 Dec 2008 Accepted: 29 Jan 2009 Published: 29 Jan 2009Critical Care 2009, 13:R9 (doi:10.1186/cc7698)This article is online at: http://ccforum.com/content/13/1/R9© 2009 Shaw et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.AbstractIntroduction Interleukin (IL)-18 is a key modulator of thecytokine response that leads to organ dysfunction andprolonged intensive care unit (ICU) stay after cardiopulmonarybypass surgery. We hypothesised that variation in the pro-inflammatory gene IL-18 is associated with adverse clinicaloutcome because of a more intense inflammatory response.Methods Haplotypes of the IL-18 gene were inferred fromgenotypes of 23 Coriell Registry subjects. Four haplotype tagsingle nucleotide polymorphisms (-607 C/A, -137 G/C, 8148C/T and 9545 T/G) identified four major haplotype clades.These polymorphisms were genotyped in 658 Caucasianpatients undergoing cardiopulmonary bypass surgery. Clinicalphenotypes were collected by retrospective chart review.Results Patients homozygous for the T allele of the 9545 T/Gpolymorphism had an increased occurrence of prolonged ICUstay (6.8% for TT genotype versus 2.7% for GG or GTgenotype; p = 0.015). Patients homozygous for the T allele alsohad increased occurrence of low systemic vascular resistanceindex (62%) compared with the GG and GT genotypes (53%;p = 0.045). Patients homozygous for the T allele had increasedserum IL-18 concentrations 24 hours post-surgery (p = 0.018),increased pro-inflammatory tumour necrosis factor alphaconcentrations (p = 0.014) and decreased anti-inflammatoryserum IL-10 concentrations (p = 0.018) 24 hours post-surgery.Conclusions The TT genotype of the IL-18 9545 T/Gpolymorphism is associated with an increased occurrence ofprolonged ICU stay post-surgery and greater post-surgicalinflammation. These results may be explained by greater serumIL-18, leading to greater pro-versus anti-inflammatory cytokineexpression.IntroductionThe balance of pro-inflammatory (e.g. tumour necrosis factoralpha (TNF-α)) and anti-inflammatory (e.g. interleukin (IL)-10)cytokine gene expression is highly correlated with organ dys-function and adverse outcome after cardiopulmonary bypass(CPB) surgery [1]. IL-18 is a key cytokine regulator of this bal-ance that, among other clinical and cytokine measures, standsout as predictive of organ dysfunction and adverse outcomesafter CPB [2]. IL-18 acts with IL-12 in a synergistic fashion tostimulate the release of interferon-gamma (IFN-γ) from lym-phocytes [3]. High serum levels of IL-18 are associated withincreased production of the pro-inflammatory cytokine TNF-α[4] and decreased production of the anti-inflammatorycytokine IL-10 [5]. Serum levels of IL-18 increase in responseto CPB surgery [2] and in other inflammatory conditions suchas sepsis [6] and type 1 diabetes [7].Inflammatory gene polymorphisms have been linked to theintensity of the post-operative inflammatory response and toclinical outcomes after CPB surgery [8]. Therefore, we postu-lated that IL-18 gene single nucleotide polymorphisms (SNPs)may be important in recovery from CPB surgery. Two polymor-phic loci within the IL-18 gene have been investigated forassociations between genotype and serum concentrations ofIL-18 (intermediate phenotype) as well as associationsbetween genotype and clinical outcome (phenotype). The -BMI: body mass index; bp: base pair; CI: cardiac index; CPB: cardiopulmonary bypass; CVP: central venous pressure; ELISA: enzyme-linked immu-Page 1 of 8(page number not for citation purposes)nosorbent assay; htSNP: haplotype tag single nucleotide polymorphism; ICU: intensive care unit; IL: interleukin; INF-γ: interferon-gamma; MAP: mean arterial pressure; PCR: polymerase chain reaction; SE: standard error; SIRS: Systemic Inflammatory Response Syndrome; SNP: single nucleotide polymorphism; SVRI: Systemic Vascular Resistance Index; TNF-α: Tumour necrosis factor-alphaCritical Care    Vol 13 No 1    Shaw et al.607 C/A and -137 G/C SNPs were initially discovered to influ-ence promoter activity of the IL-18 gene [9]. The -137 C allelehas been associated with adverse outcomes and higher serumIL-18 cytokine levels, and the -607 A allele has been associ-ated with improved outcomes and lower serum IL-18 levels intype 1 diabetes [3].Our study design was based on the above observation thatincreased serum IL-18 is associated with adverse outcomesafter CPB, including cardiovascular dysfunction [2], associ-ated with increased serum TNF-α and decreased serum IL-10[10]. Therefore, we tested the hypothesis that IL-18 gene pol-ymorphisms are associated with adverse outcome, as meas-ured by prolonged stays in the intensive care unit (ICU)(primary clinical outcome) and cardiovascular dysfunction(secondary clinical outcome). We then confirmed that the pol-ymorphism is associated with increased serum concentrationsof IL-18 and altered serum concentrations of TNF-α and IL-10post-CPB, thus providing a possible explanation for our clini-cal results.Materials and methodsPatient cohortAll patients admitted to the cardiac surgery ICU of St Paul'sHospital in Vancouver having elective CPB surgery betweenFebruary 2001 and December 2003 were eligible for entryinto this study. For inclusion, 890 were screened, of which815 had CPB pump-driven circulation of blood during first-time elective coronary artery bypass surgery. Of these, werestricted our analyses to the 658 patients who were Cauca-sian in order to decrease the potentially confounding influenceof population admixture secondary to ethnic diversity on asso-ciations between genotype and phenotype. All 658 patientswere successfully genotyped for four polymorphisms in the IL-18 gene and made up our final cohort for analysis.Our Research Ethics Board approved analysis of fully anony-mous data and, for DNA samples, collection of blood that wasbeing discarded as part of routine clinical care in a consecu-tive cohort of all patients meeting inclusion criteria and admit-ted to our cardiac surgical ICU. Using this fully anonymousstudy design the Research Ethics Board of Providence HealthCare and the University of British Columbia approved thisstudy and waived informed consent.Primary clinical phenotypeThis cohort was confined to patients undergoing first-timeelective coronary artery bypass surgery, so a mortality rateendpoint would be ineffective unless sample sizes were verylarge. Many authors, therefore, use prolonged ICU stay as ameasure of adverse outcome [11,12]. Most recently, Nakasujiand colleagues found that prolonged ICU stay of more thanthree days was a sensitive and specific measure of adverseoutcome that reflected measures of cardiovascular and pulmo-the Parsonnet score with a maximum predictive efficiency wasa score of 10. Patients having a Parsonnet score of 0 to 9 hada mean ICU stay of about 1.5 days, while those having a scoreof 10 or more had a mean ICU stay of about three days [12].Therefore, we used the proportion of patient having a post-operative ICU stay of greater than or equal to three days (72hours) as our primary outcome variable. In our institution, thedecision to discharge patients from the ICU after CPB surgeryis protocol driven and genotype was unknown to care provid-ers, making this measurement an unbiased outcome measure-ment for this study, particularly because genotype was fullyblinded and not measured until after complete clinical datahad been collected.Secondary clinical phenotypeKristof and Magder [13] identified low post-CPB systemic vas-cular resistance index (SVRI) as a clinical manifestation of sys-temic inflammation. This vasodilatory syndrome is associatedwith related parameters such as longer cross-clamp times andlower post-CPB platelet count [13]. We used Kristof andMagder's definition of two consecutive SVRI measurementsless than 1800 dyne.s/cm5/m2 (SVRI = ((MAP-CVP) × 80)/CI,where MAP is the mean arterial pressure, CVP is the centralvenous pressure and CI is the cardiac index) as a secondaryclinical phenotype.Intermediate phenotypeSerum concentrations of cytokines are useful intermediatephenotypes to test for biologically plausible explanations forgenotype – clinical phenotype associations. We measuredserum IL-18, TNF-α and IL-10 concentrations in an overlap-ping subset of patients (n = 44) within the current cohort forwhom serum was available from a related observational study.Inclusion and exclusion criteria for this cohort were identical.In our own preliminary time course experiments we found thatserum IL-18 concentrations peaked at 24 hours post-CPB.Therefore, we measured serum IL-18, TNF-α and IL-10 at thispost-operative time point. Serum IL-18 and TNF-α were meas-ured by ELISA (R&D Systems, Minneapolis, MN for IL-18; BDPharMingen, San Diego, CA for TNF-α). Serum IL-10 wasmeasured using the Luminex bead-based bioassay system(Luminex Corp, Austin, TX).Tag single nucleotide polymorphism selectionTo determine IL-18 gene haplotypes we used unphased gen-otype data from the University of Arizona's Innate ImmunityProgram in Genomics Application website. We used PHASEv 2.0 [14] to infer haplotype from unphased genotype data.The resulting haplotypes were clustered into four groups ofsimilar haplotypes (clades) using the molecular evolutionarygenetic analysis software package MEGA2 [15]. The programLDSelect [16] was used to select a set of maximally informa-tive haplotype tag (ht) SNPs with the restriction that the litera-ture SNPs -607 C/A and -137 G/C be included. Four htSNPsPage 2 of 8(page number not for citation purposes)nary organ failure [11]. Lawrence and colleagues found that were chosen to differentiate the four haplotype clades: -607Available online http://ccforum.com/content/13/1/R9C/A (rs1946518), -137 G/C (rs187238), 8148 C/T(rs360722) and 9545 T/G (rs5744249) (Figure 1).GenotypingDNA was extracted from peripheral blood samples using aQIAamp DNA Blood Maxi Kit (Qiagen Inc. Canada, Missis-sauga, ON, Canada). SNP genotypes were determined usingthe 5' nuclease, or Taqman PCR method (Applied Biosys-tems; Foster City, CA, USA) [17]. These htSNPs were thengenotyped in the 658 patients of our cohort undergoing CPB.StatisticsDifferences in continuous variables were assessed using Stu-dent's t-test for two groups or an analysis of variance for morethan two groups for normally distributed data, and a Mann-Whitney U Test or Kruskal-Wallis H Test for non-normally dis-tributed data. Fisher's exact test was used to test for signifi-cant differences in discrete variables. Data are reported asmean ± standard error (SE). A p ≤ 0.05 was taken to indicatea significant difference. Analysis was performed using SPSSv11.5 (SPSS, Chicago, IL, USA). Allele frequencies weretested for Hardy-Weinberg equilibrium using the test of GuoFigure 1Haplotypes of the IL-18 genetypes of the IL-18 gene. a) Haplotype diagram of the interleukin (IL)-18 gene. Haplotypes were inferred from unphased genotype data of 23 unrelated healthy Caucasians by the PHASE v2.0 program. Each column represents a polymorphic locus in the IL-18 gene, and is coded as either the common allele (black square) or the rare allele (white square). The position of each polymorphic locus relative to the start site of translation is indicated at the top of the diagram. Each row indicates a unique haplotype inferred from genotype data. The clustering of similar haplotypes into clades is done as per the phylogenetic relation among haplotypes given by the MEGA2 software package. The positions chosen for genotyping (-607, -137, 8148 and 9545) are indicated in the figure by bold numbering of the relative position and letter designations for the alleles within the dia-gram. b) Haplotype diagram of the interleukin-18 gene in the patient population under study. Haplotypes were inferred from unphased geno-type data of 658 unrelated Caucasians making up the study cohort using the PHASE v2.0 program. Allele and clade designations within the diagram are as in Figure 1a. Haplotype frequency within the patient cohort is indicated to the right of each haplotype. Only haplotype clades with frequency Page 3 of 8(page number not for citation purposes)greater than 5% are represented.Critical Care    Vol 13 No 1    Shaw et al.and Thompson [18] and were all found to be in Hardy-Wein-berg equilibrium.Work by Long and Langley [19] indicates that on the order of500 individuals are sufficient to detect the presence of causa-tive polymorphisms having small effect on outcomes. Thus, wehave included 658 individuals in our study population.ResultsSelection of the 9545 T/G SNPUsing PHASE, we found four main haplotypes occurring at afrequency greater than 5% in our patient cohort based on thefour chosen htSNPs (Figure 1b). As an initial screen, we exam-ined the occurrence of our primary clinical outcome (pro-longed ICU stay) by PHASE-inferred haplotype in our studycohort. We found haplotypes 1 to 3 were associated withgreater occurrence of prolonged ICU stay (Figure 2). Haplo-types 1 to 3 are uniquely tagged by the T allele of the 9545 T/G polymorphism, so we focused on the 9545 T/G genotypefor further analyses. These data were best fit using a recessivemodel (TT genotype compared to TG plus GG genotypes),which was therefore used as the statistical analytic modelthroughout the study. The TT genotype occurred in 58% ofpatients. Genotypes of the 9545 T/G SNP were in HardyWeinberg equilibrium. No significant associations betweengenotype and any clinical or intermediate phenotype wereobserved for any of the three other htSNPs (data not shown).Baseline and peri-operative characteristicsThere were no significant differences between patientshomozygous for the T allele of the IL-18 9545 T/G SNP versuspatients having a GT or GG genotype in baseline characteris-tics (Table 1) except for body mass index (BMI) (28 kg/m2 forpatients with TT genotype, 28 kg/m2 for patients with GT gen-otype, 26 kg/m2 for patients with GG genotype, p = 0.022).There were no significant differences by genotype in peri-oper-ative management, including use of aprotinin (TT 15%,GT+GG 10%, p = 0.10), amicar (TT 39%, GT+GG 38%, p =0.99), or protamine (TT 0%, GT+GG 1%, p = 0.18) in ourstudy cohort. Similarly, there were no differences by genotypein the use of vasodilators dobutamine (TT 71%, GT+GG 75%,p = 0.25) or milrinone (TT 11%, GT+GG 13%, p = 0.47) inthe perioperative period. In addition there was no difference invasopressor use (receiving norepinephrine at 4, 12 and 24hours: TT 7.8%, 5.2%, 2.1%, GT+GG 7.2%, 5.4%, 1.1%,respectively).Primary clinical phenotypeIn our cohort of Caucasian patients who had CPB surgery,patients homozygous for the T allele of the IL-18 9545 T/GSNP had increased occurrence of prolonged ICU stay(greater than 72 hours) after CPB surgery: 32 of 383 (8.4%)for TT versus 10 of 275 (3.6%) for TG or GG (p = 0.015; Fig-ure 3). The overall occurrence of prolonged ICU stay was 42out of 658 patients (6.4%). These differences remained signif-icant (p = 0.012) using logistic regression adjusted for age,gender, duration of bypass and baseline organ function (ejec-tion fraction, diabetes, renal dysfunction; Table 2). The sensi-tivity of the TT genotype to predict prolonged ICU stay is (ortrue positive rate) 76% and the specificity (or true negativerate) is 43%.Secondary clinical phenotypeIn these Caucasian patients undergoing elective on-pump cor-onary bypass graft surgery (CPB), those patients homozygousfor the T allele of the IL-18 9545 T/G SNP had increased fre-quency of two consecutive SVRI measurements less than1800 dyne.s/cm5/m2. The TT genotype was associated withthe increased frequency of two SVRI measurements less than1800 (62%) compared with the GT+GG genotypes (53%)(Figure 3). These differences remained significant (p = 0.045)using logistic regression adjusted for age, gender and dura-tion of bypass (Table 3). The difference in SVRI was notaccounted for by differences in post-operative use of vasodi-lating inotropes or the vasopressor norepinephrine (above).The fraction of patients having two or more Systemic Inflam-matory Response Syndrome (SIRS) criteria at 24 hours was24% in TT patients and 20.7% in GT+GG patients. The heartrate component of SIRS scoring was significantly differentwith TT patients having a heart rate at 24 hours of 78.4 ± 0.7compared with 75.8 ± 0.7 for GT+GG patients (p = 0.016).A trend to increased numbers of TT patients remaining on theventilator at 24 hours (TT 10.2%, GT+GG 8.3%), increasednumbers of patients having a rise in creatinine of more than 50μmol/L at 24 hours (TT 4.1%, GT+GG 2.5%) and increasedFigure 2Prolonged intensive care unit (ICU) stay after cardiopulmonary bypass (CPB) by interleukin (IL)-18 9545 T/G genotype in Caucasi n patients who underw nt on-pump CPB surgeryl g d intens ve care unit (ICU) st y after cardiopulmon ry bypass (CPB) by interleukin (IL)-18 9545 T/G genotype in Cauca-sian patients who underwent on-pump CPB surgery. Percentage of patients having prolonged ICU stay (greater than 72 hours) are indi-cated on the Y-axis and IL-18 genotype on the X-axis. Patients homozygous for the T allele of the IL-18 9545 T/G SNP had increased occurrence of prolonged ICU stay, compared with patients having a TG Page 4 of 8(page number not for citation purposes)mortality (TT 1.8%, GT+GG 1.1%) was observed, althoughor GG genoytpe.Available online http://ccforum.com/content/13/1/R9the numbers of patients falling into these categories was toolow for these differences to be statistically significant.Intermediate phenotypeIn a subgroup of patients for whom serum cytokine measure-ments were available, a significant difference was found in theconcentration of IL-18 24 hours post-CPB surgery by geno-type of the IL-18 9545 T/G polymorphism. Patientshomozygous for the T allele had a significantly greater meanserum IL-18 concentration (372 ± 24 pg/mL) compared withthose with the GT or GG genotypes (260 ± 48 pg/mL, p =0.018; Figure 4).A significant difference in serum concentration of TNF-α by IL-18 9545 T/G genotype was observed 24 hours post-opera-tively. Patients homozygous for the T allele of the IL-18 9545T/G SNP had a significantly higher serum concentration ofTNF-α at this time point (314 ± 83 pg/mL) compared with allother patients (67 ± 37 pg/mL, p = 0.014; Figure 4).A significant difference in serum IL-10 levels was observed at24 hours post-CPB by IL-18 9545 T/G genotype. Patientshomozygous for the T allele had significantly lower serum IL-10levels 24 hours post-CPB (2.4 ± 1.0 pg/mL) compared withall other patients (9.1 ± 2.8 pg/mL, p = 0.018; Figure 4).Table 1Baseline characteristics by IL-18 9545 genotype9545 TT 9545 GT 9545 GG Total p valueNumber (%) 383 (58) 225 (34) 50 (8) 658Age (years) 65 ± 1 66 ± 1 66 ± 1 66 ± 1 0.655Male sex (%) 77 77 70 76 0.544Body mass index (kg/m2) 28 ± 1 28 ± 1 26 ± 1 28 ± 1 0.022Ejection fraction (%) 49.2 ± 0.7 48.8 ± 0.9 51.3 ± 2.1 49.3 ± 0.5 0.529Renal dysfunction (%) (Creatinine > 200 μmol/L) 3.6 2.2 0.0 2.9 0.266Diabetes (Types I & II) (%) 29 28 27 28 0.969Smoking (%) 32 41 36 32 0.129Anti-hypertensive use (%) 65 61 58 62 0.495Angiotensin-converting enzyme II inhibitor use (%) 48 54 50 52 0.447Beta-blocker use (%) 58 60 56 59 0.745Aspirin use (%) 59 59 60 59 0.992Duration of surgery (hours) 4.4 ± 0.1 4.5 ± 0.1 4.4 ± 0.1 4.5 ± 0.1 0.728Duration of bypass (hours) 1.8 ± 0.1 1.9 ± 0.1 1.8 ± 0.1 1.8 ± 0.1 0.544Cross clamp time (hours) 1.3 ± 0.1 1.4 ± 0.1 1.4 ± 0.1 1.4 ± 0.1 0.143Continuous variables are reported as mean ± standard error of the mean.Figure 3Prolonged intensive care unit (ICU) stay and low Systemic Vascular Resistance Index (SVRI) after cardiopulmonary bypass (CPB) by inter-leuk n (IL)-18 9545 T/G ge otype n Caucasian patients who unde -w nt on-pump CPB su geryl ged i tensive car  unit (ICU) st y nd low Sy temic Vascular Resistance Index (SVRI) after cardiopulmonary bypass (CPB) by interleukin (IL)-18 9545 T/G genotype in Caucasian patients who underwent on-pump CPB surgery. Percentage of patients hav-ing prolonged ICU stay (greater than 72 hours) are indicated on the first Y-axis and percentage of patients having low SVRI after CPB sur-gery are indicated on the second Y-axis. IL-18 9545 T/G genotypes are indicated on the X-axis. Patients homozygous for the T allele of the IL-18 9545 T/G SNP had significantly greater occurrence of prolonged ICU stay and significantly increased occurrence of low SVRI after CPB compared with all others (GT and GG genotypes).Page 5 of 8(page number not for citation purposes)Critical Care    Vol 13 No 1    Shaw et al.DiscussionWe have identified an association between the TT genotype ofthe novel 9545 T/G polymorphism of the IL-18 gene andadverse outcome after CPB surgery. The TT genotype wasfound to be associated with increased occurrence of pro-longed ICU stay and of two consecutive SVRI measurementsless than 1800 dyne.s/cm5/m2. A biologically plausible expla-nation for these findings is that the TT genotype of the 9545T/G htSNP was also associated with higher serum concentra-tion of IL-18, higher serum TNF-α concentration and lowerserum concentration of IL-10 in our cardiac surgical popula-tion. Prolonged ICU stay is reported to be correlated with riskfactors for poor outcome after CPB surgery [11]. Low sys-temic vascular resistance, defined as two consecutive SVRImeasurements less that 1800 dyne.s/cm5/m2, is a marker ofpoor vascular tone after CPB, related to the systemic inflam-matory response [13]. Greater serum concentrations of IL-18[2] and TNF-α [1] and lower serum levels of IL-10 [1] havebeen associated with increased prevalence of complicationsafter CPB, and may be indicative of a prominent pro-inflamma-tory state.Production of humoral inflammatory mediators and priming ofneutrophils by exposure to the CPB apparatus enables a'post-pump' syndrome characterised by a systemic inflamma-tory response syndrome and its anti-inflammatory counterpart,termed the compensatory anti-inflammatory response syn-drome [20]. IL-18 plays a central role in regulating and balanc-ing these responses. IL-18 regulates the expression of thepotent pro- and anti-inflammatory mediators TNF-α [4] and IL-10 [5]. In accord with this, we found that the TT genotype wasassociated with an increased serum IL-18 concentration andalso with increased serum TNF-α and decreased serum IL-10.The increased serum TNF-α and decreased serum IL-10 levelsare associated with increased organ dysfunction [10]. There-fore, our results are consistent with the hypothesis that the IL-18 9545 TT genotype leads to increased production of serumIL-18, subsequently leading to increased serum TNF-α anddecreased serum IL-10, causing increased organ dysfunctionand increased occurrence of prolonged ICU stay.Polymorphisms in the IL-18 gene have been studied for asso-ciation with inflammatory conditions such as type 1 diabetes[3] and sepsis [21]. The -607 C/A and -137 G/C promoterTable 2Logistic regression parameters. The outcome of logistic regression is shown for a model where intensive care unit (ICU) stay more than three days is the dependent variable (primary outcome). The covariates entered into the model were IL-18 9545 genotype, age, gender, duration of bypass, pre-operative ejection fraction and whether the patient had diabetes or renal dysfunction. Renal dysfunction was defined as baseline creatinine more than 200 μmol/LCovariate Relative riskof ICU stay > three days95% confidence interval p valueIL-18 9545 TT versus TG/GG genotype 3.01 1.28 to 7.10 0.012Age (per year) 1.01 0.98 to 1.04 0.608Gender (female) 0.95 0.40 to 2.27 0.905Duration of bypass (per hour) 2.09 1.49 to 2.93 0.000Ejection fraction (per %) 0.97 0.95 to 1.0 0.026Diabetes 1.10 0.50 to 2.43 0.806Renal dysfunction 1.77 0.36 to 8.63 0.482Table 3Logistic regression parameters. The outcome of logistic regression is shown for a model where low post-cardiopulmonary bypass (CPB) Systemic Vascular Resistance Index (SVRI) is the dependent variable (secondary outcome). The covariates entered into the model were IL-18 9545 genotype, age, gender and duration of bypass.Covariate Relative riskof low post-CPB SVRI95% confidence interval p valueIL-18 9545 (TT versus TG/GG genotype) 1.57 1.01 to 2.44 0.045Age (per year) 1.01 0.98 to 1.03 0.480Gender (female) 0.70 0.39 to 1.27 0.703Duration of bypass (per hour) 1.03 0.66 to 1.62 0.889Page 6 of 8(page number not for citation purposes)Available online http://ccforum.com/content/13/1/R9SNPs have been found to be associated with a susceptibilityto type 1 diabetes such that the C allele of -137 was found tobe a risk allele and the A allele of -607 was found to be a pro-tective allele [3]. Polymorphisms in the IL-18 gene have alsobeen associated with alterations in serum concentrations of IL-18 [22]. None of the previous literature reports an associationbetween the 9545 T/G SNP and outcomes, or between thisSNP and serum cytokine levels. The previous SNPs, -607 C/A and -137 G/C are not in significant linkage disequilibriumwith the 9545 T/G SNP (Figure 1). Neither of these previouslyreported SNPs were in association with our primary or sec-ondary clinical or intermediate phenotypes.The 9545 T/G SNP of the IL-18 gene is located within intron2 and therefore may not be the causal SNP with regard to clin-ical outcomes or intermediate phenotypes. Strong linkage dis-equilibrium exists between the 9545 T/G SNP and severalother SNPs within the IL-18 gene (Figure 1): -2163 C/A(rs5744222) in the promoter; 790 G/T (n/a) and 8936 A/G(rs4988359) in intron 1; 11024 G/C (rs1834481), 12003 T/C (rs5744256) and 13084 G/C (rs5744258) in intron 3; and17980 G/C (rs5744276) in intron 5. Promoter SNPs are likelyto have an effect by disrupting or creating binding sites fortranscription factors, thus altering levels of intracellular tran-script and/or extracellular protein [23]. The -2163 C/A SNP,however, is not found to be located within a putative transcrip-tion factor binding site, nor does the rare allele create one,based on a scan of the promoter region of the IL-18 geneusing TESS (transcriptional element search software). IntronicSNPs can affect splicing of the pre-mRNA, resulting in alterna-tive splice variants of proteins; however, most intronic SNPsthought to have these effects are observed to lie within aboutitself does not lie so close to such a boundary, nor do any ofthe SNPs in LD with 9545 mentioned above. Linkage disequi-librium can carry over the putative gene boundary, and mayallow the 9545 T/G SNP to be tightly correlated with a causalSNP up- or downstream of the IL-18 gene, one which mayhave a causal effect on clinical and intermediate phenotypes.Haplotypes represent a powerful method of selecting SNPsfor genotyping based on linkage disequilibrium. The IL-18gene has 56 polymorphic loci in Caucasians according tosequencing data from the University of Arizona's Innate Immu-nity Program in Genomic Applications website. The SNPs atpositions -607 and -137 relative to the start of translation areliterature SNPs, and so make good choices for genotyping inan inflammatory-related population such as CPB patients. Thepositions 8148 and 9545 are not previously reported SNPs,however, and so would not have been chosen for genotypingwithout some method of selecting SNPs. By selecting htSNPsto maximise information while minimising the number of SNPsto be genotyped, we have queried the underlying haplotypestructure while using relatively few polymorphic loci. Althoughthis marker-style approach to disease association is useful, itdoes not address mechanism or functionality of the polymor-phism itself.Our study has several strengths. The use of haplotypes tochoose htSNPs has the benefit of not being restricted to liter-ature SNPs for disease association studies. The large samplesize allows for statistical power to detect associations of mod-est effect and the limitation of sampling to Caucasians patientsreduces the likelihood of type I error due to population admix-ture.One of the main weaknesses of our study design is that we donot identify the causative SNP for worse clinical outcome afterCPB. Linkage disequilibrium existing within and possiblybeyond the IL-18 gene suggests that polymorphic loci in link-age disequilibrium with the T allele of the 9545 T/G SNP couldcontribute to detrimental effects after CPB surgery. We havelimited our study to a single cohort; therefore this arisinghypothesis should be tested in other cohorts to ensure repro-ducibility.ConclusionsIn the present study the TT genotype of a novel polymorphismof the IL-18 gene, 9545 T/G, was associated with greateroccurrence of prolonged ICU stay after CPB surgery, greaterfrequency of low SVRI (two consecutive SVRI measurements< 1800 dyne.s/cm5/m2), higher serum concentrations ofcytokine IL-18, higher serum concentrations of the cytokineTNF-α and lower serum concentrations of the anti-inflamma-tory cytokine IL-10. These widely varied markers of intensity ofrecovery post-CPB indicate this genotype is potentially a riskfactor for patients undergoing CPB surgery.Figure 4Serum concentrations of interleukin (IL)-18, tumour necrosis factor (TNF)-α a d IL-10 by IL-18 9545 T/G genotype in Cauca an patients who underw  on-pump cardiopulmonary b ass (CPB) urgery concentrations of int rleukin (IL)-18, tumour necrosis fac-tor (TNF)-α and IL-10 by IL-18 9545 T/G genotype in Caucasian patients who underwent on-pump cardiopulmonary bypass (CPB) surgery. Serum concentrations of cytokines (pg/mL) are indicated on the y-axes, with IL-18 9545 T/G genotype indicated in the legend. Serum cytokine concentrations were measured 24 hours post-CPB. Patients homozygous for the T allele of the IL-18 9545 T/G SNP had significantly higher serum IL-18 and TNF-α concentrations and signifi-cantly lower serum IL-10 concentrations 24 hours post-CPB.Page 7 of 8(page number not for citation purposes)20 bp of intron/exon boundaries [24]. The 9545 T/G SNPCritical Care    Vol 13 No 1    Shaw et al.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsDS contributed to experimental design, data collection, geno-typing and protein measurement, conducted the primary anal-ysis of data and wrote the initial draft of the manuscript. ASassisted in data collection, in genotyping and in protein meas-urement. JR and SL contributed to experimental design anddata collection. KRW contributed to experimental design, datacollection and data analysis. All authors read, approved andcontributed to the final draft of the manuscript.AcknowledgementsSupported by Canadian Institutes of Health Research and by the Heart and Stroke Foundation of BC & Yukon. DS and AS are the recipients of Michael Smith Foundation for Health Research Graduate Studentships. KRW is a Michael Smith Foundation for Health Research Distinguished Scholar.References1. Wan S, LeClerc JL, Schmartz D, Barvais L, Huynh CH, Deviere J,DeSmet JM, Vincent JL: Hepatic release of interleukin-10 duringcardiopulmonary bypass in steroid-pretreated patients.  AmHeart J 1997, 133:335-339.2. Sablotzki A, Dehne MG, Friedrich I, Grond S, Zickmann B, MuhlingJ, Silber RE, Czeslick EG: Different expression of cytokines insurvivors and non-survivors from MODS following cardiovas-cular surgery.  Eur J Med Res 2003, 8:71-76.3. Kretowski A, Mironczuk K, Karpinska A, Bojaryn U, Kinalski M,Puchalski Z, Kinalska I: Interleukin-18 promoter polymorphismsin type 1 diabetes.  Diabetes 2002, 51:3347-3349.4. Chandrasekar B, Vemula K, Surabhi RM, Li-Weber M, Owen-Schaub LB, Jensen LE, Mummidi S: Activation of intrinsic andextrinsic proapoptotic signaling pathways in interleukin-18-mediated human cardiac endothelial cell death.  J Biol Chem2004, 279:20221-20233.5. Takeuchi D, Yoshidome H, Kato A, Ito H, Kimura F, Shimizu H, Oht-suka M, Morita Y, Miyazaki M: Interleukin 18 causes hepaticischemia/reperfusion injury by suppressing anti-inflammatorycytokine expression in mice.  Hepatology 2004, 39:699-710.6. Oberholzer A, Steckholzer U, Kurimoto M, Trentz O, Ertel W: Inter-leukin-18 plasma levels are increased in patients with sepsis7. Hanifi-Moghaddam P, Schloot NC, Kappler S, Seissler J, Kolb H:An association of autoantibody status and serum cytokine lev-els in type 1 diabetes.  Diabetes 2003, 52:1137-1142.8. Schroeder S, Borger N, Wrigge H, Welz A, Putensen C, Hoeft A,Stuber F: A tumor necrosis factor gene polymorphism influ-ences the inflammatory response after cardiac operation.  AnnThorac Surg 2003, 75:534-537.9. Giedraitis V, He B, Huang WX, Hillert J: Cloning and mutationanalysis of the human IL-18 promoter: a possible role of poly-morphisms in expression regulation.  J Neuroimmunol 2001,112:146-152.10. Daemen MA, Ven MW van de, Heineman E, Buurman WA:Involvement of endogenous interleukin-10 and tumor necro-sis factor-alpha in renal ischemia-reperfusion injury.  Trans-plantation 1999, 67:792-800.11. Nakasuji M, Matsushita M, Asada A: Risk factors for prolongedICU stay in patients following coronary artery bypass graftingwith a long duration of cardiopulmonary bypass.  J Anesth2005, 19:118-123.12. Lawrence DR, Valencia O, Smith EE, Murday A, Treasure T: Par-sonnet score is a good predictor of the duration of intensivecare unit stay following cardiac surgery.  Heart 2000,83:429-432.13. Kristof AS, Magder S: Low systemic vascular resistance state inpatients undergoing cardiopulmonary bypass.  Crit Care Med1999, 27:1121-1127.14. Stephens M, Donnelly P: A comparison of bayesian methods forhaplotype reconstruction from population genotype data.  AmJ Hum Genet 2003, 73:1162-1169.15. Kumar S, Tamura K, Jakobsen IB, Nei M: MEGA2: molecular evo-lutionary genetics analysis software.  Bioinformatics 2001,17:1244-1245.16. Carlson CS, Eberle MA, Rieder MJ, Yi Q, Kruglyak L, NickersonDA: Selecting a maximally informative set of single-nucleotidepolymorphisms for association analyses using linkage dise-quilibrium.  Am J Hum Genet 2004, 74:106-120.17. Livak KJ: Allelic discrimination using fluorogenic probes andthe 5' nuclease assay.  Genet Anal 1999, 14:143-149.18. Guo SW, Thompson EA: Performing the exact test of Hardy-Weinberg proportion for multiple alleles.  Biometrics 1992,48:361-372.19. Long AD, Langley CH: The power of association studies todetect the contribution of candidate genetic loci to variation incomplex traits.  Genome Res 1999, 9:720-731.20. Picone AL, Lutz CJ, Finck C, Carney D, Gatto LA, Paskanik A,Searles B, Snyder K, Nieman G: Multiple sequential insultscause post-pump syndrome.  Ann Thorac Surg 1999,67:978-985.21. Stassen NA, Breit CM, Norfleet LA, Polk HC Jr: IL-18 promoterpolymorphisms correlate with the development of post-injurysepsis.  Surgery 2003, 134:351-356.22. Kretowski A, Kinalska I: Serum levels of interleukin-18 – apotential marker of cardiovascular death – could be deter-mined by genetic predisposition.  Circulation 2003,107:e206-207. author reply e206–207.23. Szalai AJ, Wu J, Lange EM, McCrory MA, Langefeld CD, WilliamsA, Zakharkin SO, George V, Allison DB, Cooper GS, Xie F, Fan Z,Edberg JC, Kimberly RP: Single-nucleotide polymorphisms inthe C-reactive protein (CRP) gene promoter that affect tran-scription factor binding, alter transcriptional activity, and asso-ciate with differences in baseline serum CRP level.  J Mol Med2005, 83:440-447.24. Majewski J, Ott J: Distribution and characterization of regula-tory elements in the human genome.  Genome Res 2002,12:1827-1836.Key messages• The TT genotype of the IL-18 9545 T/G polymorphism is associated with an increased occurrence of pro-longed stay in the ICU post-surgery.• The same genotype was associated with increased IL-18 levels.• The increase in IL-18 levels associated with the TT gen-otype appeared to result in increased pro-inflammatory TNF-α levels and decreased anti-inflammatory IL-10 lev-els; TNF-α and IL-10 having previously been shown to be regulated by IL-18 in this way.• The pro-inflammatory balance (as indicated by increased TNF-α and decreased IL-10) may account for the adverse clinical outcomes associated with the TT genotype post surgery.Page 8 of 8(page number not for citation purposes)compared to severely injured patients.  Shock 2001,16:411-414.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items