Introduction

Patients with rheumatoid arthritis (RA), a common chronic inflammatory disease, have been shown to have risk of infection twice that of age- and sex-matched controls.¹ The most common infections reported in patients with established RA include urinary tract infections (UTI), respiratory tract infections (URI), skin and soft tissue infections, pneumonia, and joint infections.¹ Increased susceptibility to infections in RA is likely multifactorial resulting from the underlying immunologic disturbances associated with the disease, the use of immunosuppressive agents, and genetic predisposition. Population-based studies have identified clinical factors associated with infections in patients with RA, including increased age, comorbid diseases, corticosteroid use, and markers of disease severity such as serum rheumatoid factor (RF), the presence of rheumatoid nodules, elevated Westergren sedimentation rate (ESR), and reduced functional capacity.¹

Methotrexate (MTX) is one of the most commonly used and effective therapies for RA.² RA patients receiving MTX are more likely to have infections of the respiratory tract and skin and to receive prescriptions for antibiotics than RA patients receiving no disease-modifying antirheumatic drugs (DMARDS) or DMARDs other than MTX (not including biologic agents).³ In clinical trials, infection rates among subjects taking etanercept are no higher than that of MTX,⁴ although opportunistic infections have been reported (reviewed in Cunnane et al⁵). The use of biologic agents in RA has raised concern over risk of serious and opportunistic infections, but common infections have not been evaluated in large, prospective studies. Infections such as URI and UTI contribute substantially to health-care costs and patient morbidity.

The proinflammatory cytokines TNF and LTA play important roles in infectious and autoimmune diseases. TNF blockade ex vivo inhibits expression of Toll-like receptor 4 (TLR-4) on dendritic cells from RA patients and controls,⁶ which has important implications with regard to susceptibility to multiple infectious organisms. LTA shares many of the same biological and structural characteristics of TNF, and their genes lie in tandem in the human MHC region on chromosome 6p21. Murine studies have shown LTA to be important in tuberculosis, cerebral malaria, and cerebral Toxoplasmosis gondii.^{7, 8, 9}

Single nucleotide polymorphisms (SNPs) in TNF, and to a lesser extent LTA, are associated with susceptibility to, or severity of various autoimmune and infectious diseases. Since the report of association between the TNF -308 SNP and susceptibility to cerebral malaria,¹⁰ the role of TNF polymorphisms have been studied in many other diseases, including RA, psoriatic arthritis, leishmaniasis, hepatitis C, and ankylosing spondylitis.^{11, 12, 13, 14, 15} Although the roles of TNF and LTA SNPs in infections and in susceptibility to and severity of RA are controversial, their roles in susceptibility to infections may be accentuated in RA patients receiving MTX or etanercept.

In addition to cytokine-mediated immune processes, antibody-mediated pathways are integral to the pathogenesis of infectious and autoimmune diseases. The binding of IgG to Fc receptors can trigger numerous important effector responses including macrophage phagocytosis, natural killer cell antibody-dependent cellular cytoxicity, and neutrophil activation.¹⁶ Alteration in the function of Fc receptors has been associated with susceptibility to autoimmune and infectious diseases.¹⁷ In turn, allelic variants that influence receptor function have been identified in the FCGR2A, FCGR3A, and FCGR3B genes.¹⁸ These differences and their predicted effect on infections are shown in Table 1. Kimberly et al^{19, 20} have presented preliminary evidence for roles of Fc receptor and TNF polymorphisms in infection among subjects with RA.

Table 1 - Selected characteristics of human Fc receptors.

Full table

The goal of the current study was to delineate influences of clinical factors and polymorphisms in TNF, LTA, and FcR genes on infections in subjects with early RA treated with etanercept or MTX.

Results

Baseline characteristics, measures of disease activity, and number of subjects taking low dose prednisone were similar in the three treatment groups (Table 2). Approximately 40% of subjects were taking low-dose glucocorticoids (prednisone 10 mg/day) at study entry.⁴ Among subjects in the MTX group, the mean dose of MTX was 19 mg per week.⁴ The number of patients with one or more infections was similar among the treatment groups.⁴ In all, 61.5% of the subjects had at least one reported infection during the study period. The most common infections are listed in Table 3. The number of patients with one or more infections was similar in all treatment groups,⁴ except for increased frequency of URI in the MTX group compared to the ET10 and ET25 groups. Infections requiring hospitalization or intravenous antibiotics occurred in less than 3% of patients in each group.⁴ There were no documented opportunistic infections or deaths from infections during the study period.⁴ Importantly, neither the univariate nor multivariable analysis found association between the use of low-dose glucocorticoids and susceptibility to total infections, URI, or UTI.

Table 2 - Baseline characteristics of 457 patients with early RA according to treatment group.

Full table

Table 3 - Number and type of most common infections in the parent clinical trial.

Full table

Genotype frequencies for the nine polymorphisms are shown in Table 4. There were no statistically significant differences in the allele frequencies between the Caucasian and non-Caucasian subjects, although the number of non-Caucasians was small.⁴ Univariate analysis revealed no significant association between any individual SNP allele and presence or absence of any infection in the entire group of subjects or in subjects in each treatment group. We then analyzed the three largest groups of specific infections (URI, influenza, and UTI) individually. There were no significant SNP associations with susceptibility to influenza, which was the second most common reported infection. Univariate analysis revealed significant associations between URI and age 65 years, treatment with MTX, and the presence of the FCGR3B NA2 allele (Table 5). Presence of serum RF and elevated ESR were of marginal significance (P=0.055 and 0.080, respectively), while corticosteroid use was not significant. The NA2 allele was associated with URI (odds ratio(OR) 1.34, 95% confidence interval(CI) 1.02–1.77). The proportions of subjects with at least one URI were: 52% (99/191) of those with the NA2/NA2 genotype; 42% (77/181) of those with NA1/NA2; and 39% (23/59) of those with NA1/NA1. In the multivariable model, age 65 years, and MTX treatment remained significant, and elevated baseline ESR achieved statistical significance (OR 1.70, 95% CI 1.12–2.58) (Table 5). The FCGR3B NA2 allele remained significantly associated with URI in this model (OR 1.34, 95% CI 1.01–1.78) (Table 5). We found no associations with FCGR haplotypes and URI, suggesting that NA2 allele is the biologically relevant polymorphism playing a role in susceptibility to URI.

Table 4 - SNP genotypes of all subjects and the subset of Caucasians.

Full table

Table 5 - Factors associated with URI in univariate and multivariable analyses.

Full table

When analyzing risk alleles for UTI, univariate analysis revealed statistically significant associations with the TNF -238 A, LTA +365 C, and FCGR3A F alleles, but with none of the clinical variables (Table 6). Seven of 42 (17%) of individuals with the TNF -238 AG genotype had at least one UTI during the study period, compared to 30 of 415 (7%) of subjects with the TNF -238 GG genotype, (OR 2.56, 95% CI 1.05–6.25) (Table 6). No individuals in the study had the TNF -238 AA genotype (Table 4). In all, 13% (17/132) of individuals with the LTA +365 CC genotype had at least one UTI during the study period, compared to 7% (14/209) of subjects with the LTA +365 CG genotype and 5% (6/116) of those with the LTA +365 GG genotype (OR 1.73, 95% CI 1.07–2.79) (Table 6). The FCGR3A F was marginally associated with UTI, as 10% (21/201) of those with the FF genotype, 7% (13/196) of those with FV genotype, and 3% (2/58) of those with the VV genotype had at least one UTI during the study period (OR 1.72, 95% CI 0.99–3.00) (Table 6). In the multivariable model, the TNF -238 A allele showed a trend towards increased UTI but no longer achieved statistical significance (OR 2.45, 95% CI 0.99–6.13) (Table 6). The LTA+365 C allele association with UTI remained significant in the multivariable analysis (OR 1.70, 95% CI 1.04–2.77), while the association with the FCGR3A F allele achieved significance (OR 1.76, 95% CI 1.01–3.10) (Table 6). There were no interactions between factors in the multivariable analysis.

Table 6 - Factors associated with UTI in the univariate and multivariable analyses.

Full table

If these three polymorphisms each have independent biologic effects leading to increased susceptibility to UTI, there may be an additive or synergistic effect of having multiple risk alleles. Thus, we examined associations between UTI and the number of risk alleles, defined as the total number of the TNF -238 A, LTA +365 C, and FCGR3A F alleles. Since no individuals were homozygous for the TNF -238 A allele, subjects had a range of 0 to 5 UTI risk alleles. Although there was a normal distribution of UTI risk alleles among all individuals in the study (Figure 1a), there was a striking correlation between the number of risk alleles and presence of at least one UTI during the study period. As shown in Figure 1b, the percentage of subjects with at least one UTI were 0% (0 of 20 subjects), 1.7% (1 of 60 subjects), 6.7% (11 of 164 subjects), 8.5% (11 of 130 subjects), 15.7% (11 of 70 subjects), and 18.2% (2 of 11 subjects) for those with 0, 1, 2, 3, 4, and 5 risk alleles, respectively (Pearson correlation coefficient r=0.161, P< 0.001).

Figure 1.

(a) Distribution of 455 study subjects according to number of risk alleles. Of the original 457 subjects, two were excluded because of lack of complete genotypes. Risk alleles were defined as number of TNF -238 A, LTA +365 C, and FCGR3A 176 F alleles. Since no individuals were homozygous for the TNF -238 A allele, subjects could have a maximum of five risk alleles. (b) Percentage of subjects with given number of risk alleles who had at least one UTI during the study period.

Full figure and legend (21K)

There were no significant associations between UTI and various combinations of haplotypes defined by the three TNF SNPs, the three LTA SNPs, all six TNF/LTA SNPs, or the FcR SNPs. Again, these results suggest that associations with UTI are either specific to the TNF -238 A, LTA +365 C, and FCGR2A F alleles, or to haplotypes in linkage disequilibrium with these SNPs but not examined in this study.

Discussion

The current study is the largest reported analysis (457 subjects) of the impact of clinical and genetic risk factors for common infections. A major strength of the study is that infections were identified systematically as part of a prospective double-blind clinical trial. We identified several clinical and genetic factors that are associated with susceptibility to URI and UTI in RA patients receiving MTX or etanercept. Of the polymorphisms examined, the functional properties of the nonsynonymous alleles in the coding regions of the Fc receptor genes are best characterized. It is known that the isoform containing the FCGR3B NA1 allele produces a larger phagocytic response, oxidative burst, and degranulation response than the FCGR3B NA2 allele,¹⁸ which is consonant with our finding of increased URI in subjects with the NA2 allele. The FcIIIb receptor is expressed exclusively on neutrophils and eosinophils; depending on the causative organism, these cells may be critical for host defense against URI. The FCGR3B NA2 allele may play a role in susceptibility to other infections, as the combination of FCGR2A 131 R/R and FCGR3B NA2/NA2 genotypes is associated with meningococcal disease in subjects with late complement component (C5–C9) deficiency.²¹ The current study is the first to demonstrate a role for the FCGR3B NA2 allele in a common form of infection, URI.

FcIIIa, encoded by FCGR3A, is expressed predominantly on NK cells and macrophages. Previous reports of the functional significance of the FCGR3A F allele have shown decreased FCGR3A receptor activity compared to the V allele¹⁸ consistent with our finding of increased UTI in subjects with the FCGR3A F allele. Few reports exist regarding FcRIIIa polymorphisms and risk of infection. In one recent study, the FF genotype was associated with increased susceptibility to poliomyelitis.²² This effect has been hypothesized to be due to less effective clearance of the poliovirus by the FCGR3A F allele compared to the FCGR3A V allele. Thus, our findings of increased susceptibility to URI in relation to the FCGR3B NA2 allele and increased susceptibility to UTI in relation to the FCGR3A 176 F allele are consistent with the biologic effects of these polymorphisms.

Several studies have found associations of FCGR2A polymorphisms with susceptibility to meningococcal disease,²³ periodontal disease²⁴ recurrent bacterial respiratory tract infection in children with low IgG2 anti-carbohydrate antibodies²⁵ bacteremic pneumococcal pneumonia,²⁶ and invasive pneumococcal infections in patients with SLE.²⁷ The FCGR2A 131 H/R SNP strongly influences the ability of the receptor to bind human IgG₂, the isotype often made in response to encapsulated bacteria. In the current analysis, we did not find association between FCGR2A alleles and infections. One possible explanation for the absence of such an association is that most previous reports focused on serious infections requiring hospitalization, in which cases were identified in retrospective reviews of extended periods of hospital records. Our data were collected among outpatients over a 1-year period and there were few infections requiring hospitalization.

Our understanding of the biological relevance of the TNF -238 and LTA +365 SNPs is less clear than that of the FcR polymorphisms. The TNF -308 A allele, for example, has been reported to be associated with increased TNF production in vitro and increased transcription of TNF.²⁸ Other investigators, however, have found no significant effect of this polymorphism on TNF expression.^{11, 29} Studies using chromatin immunoprecipitation and mass spectroscopy showed, somewhat surprisingly, that the TNF -308 SNP influences transcription of the LTA gene, but not the TNF gene.³⁰

In vitro transfection assays indicate that TNF -238 A allele does not appear to have a direct effect on transcriptional activation of the TNF gene.³¹ A study of subjects with chronic active hepatitis C, however, found that the frequency of the TNF -238 A allele was significantly higher in subjects than controls.¹⁴ This finding, which was not be explained by linkage disequilibrium (LD) to HLA-B or HLA-DR genes, suggests that the TNF -238 A allele predisposes to less efficient host defense to this viral infection.

There are several potential explanations for the observed association of infections with individual TNF and LTA SNP alleles rather than SNP haplotypes. First, the number of subjects may be too small to allow for analysis of multiple haplotypes. Second, the individual TNF and LTA SNP alleles themselves may be more biologically relevant than haplotypes, as appears to be the case with the Fc receptor SNPs analyzed in this study. Finally, these genetic associations may reflect the effect of other SNPs in LD with the SNPs genotyped in this study. These SNPs may lie in the promoter or other regulatory regions and may influence gene transcription or mRNA stability.

It is unclear whether the results of our analysis can be extrapolated to individuals without RA, to subjects with RA on DMARDs other than MTX or etanercept, or to RA patients on no DMARDs. The majority of the subjects in the parent clinical trial were Caucasian,⁴ so the role of racial/ethnic differences in genetic influences on infection could not be examined. Because serious infections were rare in the 1-year study period, we were unable to analyze genetic influences on serious infections such as opportunistic infections, or those requiring hospitalization.

Our multivariable analysis found that three clinical variables, age >65 years, elevated baseline ESR, and MTX treatment, were associated with increased risk of URI. No clinical variables were associated with UTI or total infections. In a population-based study of 609 subjects with established RA, Doran et al¹ also found that increased age and elevated ESR were important clinical predictors of infection. In addition, they found that comorbid diseases, extra-articular manifestations of RA, positive RF, rheumatoid nodules, reduced functional capacity, and corticosteroid use were associated with infections. There are several factors that likely contribute to the lack of significance of these additional variables in our study. First, the patients in our study had early RA, in which extra-articular manifestations and reduced functional capacity are uncommon. In addition, patients with important concurrent illnesses were excluded from the parent clinical trial, whereas subjects with chronic lung disease, alcoholism, leukopenia, organic brain disease, and other serious conditions were included in the population-based study of Doran et al. Approximately 40% of our patients were taking low dose corticosteroids (prednisone 10 mg/day) at baseline. Since subjects had mean disease duration of 12 months, the duration of corticosteroid use was probably limited. In contrast, 47% of subjects in the study by Doran et al had received i.m. or i.v. corticosteroids, and the median number of days corticosteroids were received was 798 (2.2 years). Thus, major differences in study design probably account for the differences in results.

In summary, using prospectively collected data from 457 subjects with early RA treated with MTX or etanercept, we identified associations between common infections and polymorphisms in genes relevant to the pathogenesis of infectious and autoimmune diseases. The association between the FCGR3B NA 2 allele and susceptibility to URI, and between the FCGR3A F, TNF -238 A, and LTA +365 C alleles and UTI have important implications regarding the pathogenesis of infectious diseases and mechanisms underlying susceptibility to infection in subjects with RA and perhaps normal individuals.

Patients and methods

Study subjects

In the Immunex Early RA trial,⁴ a total of 632 patients with early RA (3 years) were randomized to receive either MTX (initial dose of 7.5 mg with rapid escalation to 20 mg per week by week 8), low-dose etanercept (10 mg twice weekly) (ET10), or standard dose etanercept (25 mg twice weekly) (ET25). DMARDs, including hydroxychloroquine and sulfasalazine, were discontinued at least 4 weeks before the study began. Stable doses of nonsteroidal anti-inflammatory drugs and prednisone (10 mg daily) were allowed. Of the 632 subjects, 457 (72.3%) consented to participate in the current genetic study.

During the trial, clinical and laboratory assessments were performed at baseline, week 2, and at 1, 6, 8, 10, and 12 months. All infections were recorded by study coordinators using a standardized protocol that characterized the site and intensity of the infection, infection onset, treatment given, requirement for hospitalization, and resolution of the infection.

Genetic polymorphisms

Candidate genes for this study and associated polymorphisms were carefully chosen based on their possible association with susceptibility to infections and their likelihood of being affected by immunomodulatory agents. We analyzed SNPs at positions -308, -238, and +488 of the TNF gene; and three SNPs at positions +249, +365, and +720 of the lymphotoxin-alpha (LTA) gene, which have been reported to define haplotypes.³² Because of the importance of Fc receptors in infectious diseases, and because etanercept–TNF complexes are thought to be degraded through the Fc receptor (FcR) pathway, we genotyped each subject for the following FcR polymorphisms: FCGR2A 131 H/R, FCGR3A 176 F/V, and FCGR3B NA 1/2.

Genotyping methods

TNF/LTA SNPs. TNF and LTA SNPs were genotyped by PCR amplification of genomic DNA and restriction fragment length polymorphism (RFLP) analysis based in part on the methods described by Mullighan et al.³²

Fc receptor polymorphisms. The 131 H/R alleles of FCGR2A, the 176 F/V alleles of FCGR3A, and the NA1/NA2 alleles of FCGR3B were determined as previously described.³³ The genotyping approach included both allele-specific PCR and direct sequencing optimized for heterozygote detection on an ABI 377 automated sequencer.

Haplotypes defined by the six TNF-LTA polymorphisms and by the three FcR alleles were constructed using the PHASE program.³⁴

Statistical methods

Baseline clinical characteristics potentially affecting susceptibility to infections and SNP alleles were analyzed for association with (1) presence or absence of any infection and (2) presence or absence of the three most common reported specific infections (URI, influenza, and UTI). Each potential predictor variable was analyzed using univariate logistic regression and the OR, 95% CI, and P-values were reported. A stepwise selection process was used to create a multivariable model of potential predictors of infection, including examination of two-way interaction among all variables in the model. The following baseline clinical and laboratory variables were included in the analyses: (1) advanced age (65 years of age vs < 65 years of age); (2) elevated ESR (females > 30 mm/hr, males >13 mm/hr), presence of rheumatoid factor (baseline RF 20 IU/ml, use of corticosteroids (prednisone 10 mg/day at baseline), and treatment group (MTX vs ET10 vs ET25). Because UTI is more common in women than in men, sex (female vs male) was included in the model used to analyze UTI. Specific SNP alleles were analyzed using an additive mode of inheritance. All analyses were performed using SPSS version 11.0 statistical software.³⁵

Although the genes examined in the current study were chosen a priori based on evidence of association with infection, the number of comparisons increases the likelihood of false-positive results. It is difficult to determine the appropriate level of statistical adjustment for these analyses since many of the polymorphisms (eg, those corresponding to the same gene or genomic region) are not independent. Therefore, all P-values shown are nomimal (ie, uncorrected) and P-values <0.05 were considered significant.

References

1. Doran MF, Crowson CS, Pond GR, O'Fallon WM, Gabriel SE. Predictors of infection in rheumatoid arthritis. Arthritis Rheum 2002; 46: 2294–2300. | Article | PubMed |
2. Kremer JM. Rational use of new and existing disease-modifying agents in rheumatoid arthritis. Ann Intern Med 2001; 134: 695–706. | PubMed |
3. van der Veen MJ, van der Heide A, Kruize AA, Bijlsma JW. Infection rate and use of antibiotics in patients with rheumatoid arthritis treated with methotrexate. Ann Rheum Dis 1994; 53: 224–228. | PubMed |
4. Bathon JM, Martin RW, Fleischmann RM et al. A comparison of etanercept and methotrexate in patients with early rheumatoid arthritis. N Engl J Med 2000; 343: 1586–1593. | Article | PubMed | ISI | ChemPort |
5. Cunnane G, Doran M, Bresnihan B. Infections and biological therapy in rheumatoid arthritis. Best Pract Res Clin Rheumatol 2003; 17: 345–363. | Article | PubMed |
6. Netea MG, Radstake T, Joosten LA, van der Meer JW, Barrera P, Kullberg BJ. Salmonella septicemia in rheumatoid arthritis patients receiving anti-tumor necrosis factor therapy: association with decreased interferon-gamma production and Toll-like receptor 4 expression. Arthritis Rheum 2003; 48: 1853–1857. | Article | PubMed | ISI | ChemPort |
7. Engwerda CR, Mynott TL, Sawhney S, De Souza JB, Bickle QD, Kaye PM. Locally up-regulated lymphotoxin alpha, not systemic tumor necrosis factor alpha, is the principle mediator of murine cerebral malaria. J Exp Med 2002; 195: 1371–1377. | Article | PubMed | ISI | ChemPort |
8. Roach DR, Briscoe H, Saunders B, France MP, Riminton S, Britton WJ. Secreted lymphotoxin-alpha is essential for the control of an intracellular bacterial infection. J Exp Med 2001; 193: 239–246. | Article | PubMed | ISI | ChemPort |
9. Schluter D, Kwok LY, Lutjen S et al. Both lymphotoxin-alpha and TNF are crucial for control of Toxoplasma gondii in the central nervous system. J Immunol 2003; 170: 6172–6182. | PubMed |
10. McGuire W, Hill AV, Allsopp CE, Greenwood BM, Kwiatkowski D. Variation in the TNF-alpha promoter region associated with susceptibility to cerebral malaria. Nature 1994; 371: 508–510. | Article | PubMed | ISI | ChemPort |
11. Brinkman BM, Huizinga TW, Kurban SS et al. Tumour necrosis factor alpha gene polymorphisms in rheumatoid arthritis: association with susceptibility to, or severity of, disease? Br J Rheumatol 1997; 36: 516–521. | Article | PubMed | ISI | ChemPort |
12. Cabrera M, Shaw MA, Sharples C, Williams H, Castes M, Convit J et al. Polymorphism in tumor necrosis factor genes associated with mucocutaneous leishmaniasis. J Exp Med 1995; 182: 1259–1264. | Article | PubMed | ISI | ChemPort |
13. Hohler T, Schaper T, Schneider PM, Meyer zum Buschenfelde KH, Marker-Hermann E. Association of different tumor necrosis factor alpha promoter allele frequencies with ankylosing spondylitis in HLA-B27 positive individuals. Arthritis Rheum 1998; 41: 1489–1492. | Article | PubMed | ISI | ChemPort |
14. Hohler T, Kruger A, Gerken G, Schneider PM, Meyer zBK, Rittner C. Tumor necrosis factor alpha promoter polymorphism at position -238 is associated with chronic active hepatitis C infection. J Med Virol 1998; 54: 173–177. | Article | PubMed |
15. Wilson AG, Gordon C, di Giovine FS et al. A genetic association between systemic lupus erythematosus and tumor necrosis factor alpha. Eur J Immunol 1994; 24: 191–195. | Article | PubMed | ISI | ChemPort |
16. Kimberly RP, Salmon JE, Edberg JC. Receptors for immunoglobulin G. Molecular diversity and implications for disease. Arthritis Rheum 1995; 38: 306–314. | PubMed | ISI | ChemPort |
17. Takai T. Roles of Fc receptors in autoimmunity. Nat Rev Immunol 2002; 2: 580–592. | Article | PubMed | ISI | ChemPort |
18. Kimberly RP, Wu J, Gibson AW et al. Diversity and duplicity: human Fcgamma receptors in host defense and autoimmunity. Immunol Res 2002; 26: 177–189. | Article | PubMed |
19. Kimberly RP, Moreland LW, Wu J, Edberg JC, Weinblatt ME, Blosch C. Occurrence of infection varies with Fc receptor genotype. Arthritis Rheum 1998; 41 (Suppl): S273 (Abstract).
20. Kimberly RP, Moreland LW, Gibson A, Weinblatt M, Bosch C. Susceptibility to infection may vary with TNF promoter genotype. Arthritis Rheum 1998; 41 (Suppl): S273 (Abstract).
21. Fijen CA, Bredius RG, Kuijper EJ et al. The role of Fcgamma receptor polymorphisms and C3 in the immune defence against Neisseria meningitidis in complement-deficient individuals. Clin Exp Immunol 2000; 120: 338–345. | Article | PubMed |
22. Rekand T, Langeland N, Aarli JA, Vedeler CA. Fcgamma receptor IIIA polymorphism as a risk factor for acute poliomyelitis. J Infect Dis 2002; 186: 1840–1843. | Article | PubMed |
23. Domingo P, Muniz-Diaz E, Baraldes MA et al. Associations between Fc gamma receptor IIA polymorphisms and the risk and prognosis of meningococcal disease. Am J Med 2002; 112: 19–25. | Article | PubMed | ChemPort |
24. Meisel P, Carlsson LE, Sawaf H, Fanghaenel J, Greinacher A, Kocher T. Polymorphisms of Fc gamma-receptors RIIa, RIIIa, and RIIIb in patients with adult periodontal diseases. Genes Immun 2001; 2: 258–262. | Article | PubMed |
25. Sanders LA, van de Winkel JG, Rijkers GT et al. Fc gamma receptor IIa (CD32) heterogeneity in patients with recurrent bacterial respiratory tract infections. J Infect Dis 1994; 170: 854–861. | PubMed | ChemPort |
26. Yee AM, Phan HM, Zuniga R, Salmon JE, Musher DM. Association between FcgammaRIIa-R131 allotype and bacteremic pneumococcal pneumonia. Clin Infect Dis 2000; 30: 25–28. | Article | PubMed | ChemPort |
27. Yee AM, Ng SC, Sobel RE, Salmon JE. Fc gammaRIIA polymorphism as a risk factor for invasive pneumococcal infections in systemic lupus erythematosus. Arthritis Rheum 1997; 40: 1180–1182. | PubMed |
28. Wilson AG, Symons JA, McDowell TL, McDevitt HO, Duff GW. Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation. Proc Natl Acad Sci, USA 1997; 94: 3195–3199. | Article | PubMed | ChemPort |
29. Stüber F, Udalovaa IA, Book M et al. 308 tumor necrosis factor (TNF) polymorphism is not associated with survival in severe sepsis and is unrelated to lipopolysaccharide inducibility of the human TNF promoter. J Inflamm 1995; 46: 42–50. | PubMed | ISI |
30. Knight JC, Keating BJ, Rockett KA, Kwiatkowski DP. In vivo characterization of regulatory polymorphisms by allele-specific quantification of RNA polymerase loading. Nat Genet 2003; 33: 469–475. | Article | PubMed | ISI | ChemPort |
31. Kaijzel EL, van Krugten MV et al. Functional analysis of a human tumor necrosis factor alpha (TNF-alpha) promoter polymorphism related to joint damage in rheumatoid arthritis. Mol Med 1998; 4: 724–733. | PubMed | ISI | ChemPort |
32. Mullighan CG, Fanning GC, Chapel HM, Welsh KI. TNF and lymphotoxin-alpha polymorphisms associated with common variable immunodeficiency: role in the pathogenesis of granulomatous disease. J Immunol 1997; 159: 6236–6241. | PubMed | ISI | ChemPort |
33. Edberg JC, Langefeld CD, Wu J et al. Genetic linkage and association of Fcgamma receptor IIIA (CD16A) on chromosome 1q23 with human systemic lupus erythematosus. Arthritis Rheum 2002; 46: 2132–2140. | Article | PubMed | ISI | ChemPort |
34. Stephens M, Smith NJ, Donnelly P. A new statistical method for haplotype reconstruction from population data. Am J Hum Genet 2001; 68: 978–989. | Article | PubMed | ISI | ChemPort |
35. SPSS for Windows. Rc1.11.01.2001, Chicago: SPSS Inc.

Acknowledgements

This work was supported by NIH R01 AR47224, P60 AR48095, the UAB Multidisciplinary Clinical Research Center Methodology Core; K24 AR-02175 and the Rosalind Russell Medical Research Center for Arthritis at UCSF. The technical assistance of Kevin N Turner, Blanche Woehl, Yuanqing Zhu, and Jinyi Wang is gratefully acknowledged.