Single nucleotide polymorphisms in genes for cytokines interleukin (IL)-2, IL-6 and TNFalpha influence severity of osteolysis after total hip arthroplasty F. Mrazek 1, J. Gallo 2, A. Arakelyan 1, Z. Kubistova 1, M. Petrek 1 1 Laboratory of Immunogenomics, Dept. of Immunology 2 Department of Orthopaedics Palacky University and University Hospital Olomouc, Czech Republic Supported by Czech Govt. Funding IGA MZ NR9490 and MSM

BACKGROUND Periprosthetic osteolysis (OL) is dominant long-term complication of the total hip arthroplasty (THA) which can result in prosthetic failure and re-operation. Pathogenesis of OL is complex – both biological and mechanical factors play an important role. Polyethylene wear particles liberated through the prosthetic surfaces stimulate an inflammatory response leading to osteolysis. Substantial interindividual variability in the severity of OL suggests implication of genetic factors.

CYTOKINES & OSTEOLYSIS  cytokines participate in bone resorption pathway (proliferation, differentiation and function of osteoclasts) - tumor necrosis factor (TNF) - interleukin (IL)-1 - macrophage-colony stimulating factor (M-CSF) - receptor activator of nuclear factor kappa beta ligand (RANKL)  production and regulation of cytokines is affected by particular variants of cytokine genes  cytokine / cytokine receptor polymorphisms – candidates for participation in genetic susceptibility to OL.

HYPOTHESIS & AIM Periprosthetic Osteolysis Cytokines Is there any association between the variants of cytokine genes and susceptibility to periprosthetic OL/ premature failure after total hip arthroplasty? Genetic component Individual susceptibility Cytokine gene polymorphisms ???

SUBJECTS & DESIGN 205 patients after total hip arthroplasty (THA) *Severe (types III-V) and mild (I+II) osteolysis according to Saleh et al classification, J Bone Joint Surg Am, 83: 1040, 2001.

METHODS 1. Genotyping of cytokine gene polymorphisms - 22 selected cytokine gene polymorphisms (Table 1) - Polymerase Chain Reaction with Sequence Specific Primers (PCR-SSP) - „The Cytokine Typing Tray kit“, University of Heidelberg 2. Statistics - conformity of genotype distribution to the Hardy-Weinberg equilibrium: Chi-square test - differences between allelic, genotype and phenotype („carriage rate“) frequencies: Chi-square test - risk assessment: odds ratio, population attributable risk (PAR) - survival analysis – Kaplan-Meier curves, log-rang test

Table 1: Overview of investigated cytokine single nucleotide polymorphisms Cytokine Gene location Gene polymorphisms (SNP) IL-1α 2q C/T IL-1β 2q C/T, C/T IL-1R 2q12 pstI 1970 C/T IL-1RA 2q14 Mspa T/C IL-4R  16p A/G IL-12 3q A/C  IFN 12q14 UTR 5644 A/T TGFβ 19q13 Codon 10 T/C, Codon 25 G/C TNFα 6p G/A, -238 G/A IL-2 4q T/G, +166 G/T IL-4 5q T/G, -590 C/T, -33 C/T IL-6 7p G/C, nt 565 G/A IL-10 1q A/G, -819 C/T, -592 C/A

RESULTS I Associations of cytokine gene polymorphisms with severity of osteolysis The proportion of TNF-238*A and IL-6-174*G allele carriers was higher in patients with severe OL than in those with mild OL in both study phases and in combined analysis (Fig. 1, 2). By contrast, combined analysis revealed that the allele IL *G was protective from severe osteolysis (Fig. 3).

RESULTS II Combined analysis of THA survival and risk of revision with cytokine gene variants Carriage of TNF-238*A allele was associated with worse long- term THA survival (Fig. 4). On the other hand, THA survived better in patients carrying IL-2-330*G allele (Fig. 5). Similarly, a risk for THA revision was higher in TNF-238*A allele carriers (p=0.017) and decreased in patients carrying IL-2-330*G allele (p=0.02).

Figure 1: Comparison of proportion of TNF-238*A allele carriers in patients with severe / mild OL and healthy control subjects (Controls) - combined analysis Severe OL versus mild OL: odds ratio = 6.6, p=0.005, PAR%=5.2* * PAR% - population attributable risk percentage – percentage of the incidence of a disease in the population that is due to a certain exposure

Severe OL versus mild OL: odds ratio = 2.5, p=0.007, PAR%=31.5 Figure 2: Comparison of proportion of IL-6-174*G allele carriers in patients with severe / mild OL and healthy control subjects (Controls) - combined analysis

Severe OL versus mild OL: odds ratio = 0.55, p=0.043 Figure 3: Comparison of proportion of IL-2-330*G allele carriers in patients with severe / mild OL and healthy control subjects (Controls) - combined analysis

Figure 4: Comparison of long-term THA survival between the carriers (blue curve) and non-carriers (green curve) of the TNF-238*A allele (log rank test: p=0.022) Mean survival: TNF-238*A carriers: 6.7 years TNF-238*A non carriers: 8.0 years

Figure 5: Comparison of long-term THA survival between the carriers (blue curve) and non-carriers (green curve) of IL-2-330*G allele (log rank test: p=0.018) Mean survival: IL-2-330*G carriers: 7.8 years IL-2-330*G non carriers: 6.7 years

CONCLUSIONS The alleles TNF-238*A and IL-6-174*G may predispose patients after total hip arthroplasty (THA) to the development of severe periprosthetic osteolysis (OL). By contrast, carriage of IL-2-330*G allele appeared to be protective from severe OL in our group of THA patients. Furthermore, certain variants of cytokine genes were associated with worse (TNF-238*A) or better (IL-2-330*G) overall THA outcome (long-term THA survival and risk for THA revision).