• ホーム>
  • 我々の研究成果 | 学会発表内容と結果

我々の研究成果

学会発表内容と結果

論文

1.全ゲノム網羅的SNP解析によるインフリキシマブあるいはエタネルセプト治療関節リウマチ患者の有効性、副作用発現を予測するSNPアルゴリズムの開発 (臨床リウマチ Vol.21/No.4 399-405,2009)

学会発表

[国内学会発表]

1.    演者:松原 司、児矢野聡、舟橋恵子、鳥山紗由美、中原邦彦、萩原敬史、三浦孝子、奥田康介、佐川 昭、桜井武男、松野博明、泉原智麿、生野英祐. 演題 名:多施設病院コホートによるインフリキシマブあるいはエタネルセプト治療関節リウマチ患者の有効性、副作用発現を予測するSNPアルゴリズムの開発及び その検証. 第53回日本リウマチ学会総会・学術集会. 東京 (2009)  目的:インフリキマブ及びエタネルセプトは関節リウマチに代表される炎症性疾患の治療に有効な生物学的製剤である。しかし、これら生物学的製剤の有効、無 効あるいは副作用発現を治療前に予測する方法はない。そこで本演題では多施設病院コホート集団検体を用いて、ゲノムワイドSNP解析による両生物学的製剤 治療の有効性、副作用発現を予測するアルゴリズムの開発を試みた。患者と方法:第一集団はインフリキシマブ投与患者95症例、エタネルセプト投与患者92 症例の計187症例を、第二集団はそれぞれ117症例、89症例の計206症例、総合計393症例を対象とした。薬剤有効性はEULAR基準に基づき、投 与後24-30週でのDAS28(CRP)値により決定した。副作用は投与期間中の発熱、皮膚障害、胃腸障害発現などを対象とした。ゲノムワイドSNP解 析はイルミナ社のHumanHap 300Kを用いておこなった。総計285,548 SNPsと有効性あるいは副作用発現との相関解析はカイ自乗検定を用いておこない、両集団の解析結果p < 0.01の相関のあった10 SNPsを選択した。次に各SNP部位に関して、レスポンダー群(あるいは副作用発現群)で頻度の高いアレルをホモで持つ患者を+1点、ヘテロで持つ患者 を0点、ノンレスポンダー群(あるいは副作用非発現群)で頻度の高いアレルをホモで持つ患者を-1点とし、計10 SNPsでの合計点を算出した。また、第一集団の解析で作成したアルゴリズムの第二集団での信頼性も検証した。結果:インフリキマブ有効性、副作用、エタ ネルセプト有効性、副作用発現に関して、90-95%の割合で合計点により両群を分けることが可能であった。第一集団の解析で作成したアルゴリズムの第二 集団での信頼性は検証中である。結論:このSNP解析によるアルゴリズムを用いることにより、インフリキマブあるいはエタネルセプト治療前に、その有効性 や副作用発現を高確率で予測可能であることが示唆され、将来的なテーラーメイドRA治療戦略に大変有用であることが考えられる。

2.    演者:松原 司、児矢野聡、鳥山紗由美、中原邦彦、萩原敬史、三浦孝子、奥田康介、舟橋恵子. 演題名:ゲノムワイドSNP解析によるRA患者の関節破壊 進行度とADL障害を予測するアルゴリズムの開発.  第53回日本リウマチ学会総会・学術集会. 東京 (2009)  目的:RAにおいて関節破壊進行度やADL障害を治療前に予測することは治療戦略を立てる上で極めて価値が高いが、これらを予測する方法はない。そこで本 演題ではゲノムワイドSNP解析を用いた関節破壊進行度とADL障害を予測するアルゴリズムの開発を試みた。患者と方法:解析に用いた136症例は関節破 壊進行度の指標であるSteinbrokerのstageあるいはADL障害のclassに分類した。全ての症例はDMARDs無効でTNF療法に移行し たものであった。発症5年未満の48症例では、stage I及びII(low stage群)が22症例、stage III及びIV(high stage群)が26症例であり、class I及びII(low class群)が41症例、class II及びIV(high class群)が7症例であった。ゲノムワイドSNP解析はイルミナ社のHumanHap 300Kを用いておこなった。総計285,548 SNPsと関節破壊進行度あるいはADL障害との相関解析はカイ自乗検定を用いておこない、p < 0.001の非常に強い相関のあった10 SNPsを選択した。次に各SNP部位に関して、high stage群(あるいはhigh class群)で頻度の高いアレルをホモで持つ患者を+1点、ヘテロで持つ患者を0点、low stage群(あるいはlow class群)で頻度の高いアレルをホモで持つ患者を-1点とし、計10 SNPsでの合計点を算出した。結果:発症5年未満の48症例の関節破壊進行度に関してはhigh stage群の約96%が合計点2点以上であり、low stage群の約96%が1点以下であった。さらにADL障害に関してはhigh class群の約86%が-2点以上であり、low class群の100%が-3点以下であった。結論:このSNP解析によるアルゴリズムを用いることにより、患者個人の関節破壊進行度やADL障害を治療 前に高精度で予測することが可能であることが示唆され、将来的なテーラーメイドRA治療戦略に大変有用であることが考えられる。

3.    演者:松原 司、児矢野聡、鳥山紗由美、中原邦彦、萩原敬史、三浦孝子、奥田康介、舟橋恵子. 演題名:ゲノムワイドSNP解析によるインフリキシマブ (IFX)とエタネルセプト (ETN) 治療RA患者における有効例 (R)、無効例 (NR) と副作用発現を予測するアルゴリズムの開発. 第53回日本リウマチ学会総会・学術集会. 東京 (2009)  目的:IFX及びETNの有効、無効あるいは副作用発現を治療前に予測する方法はなく、そのためのSNP解析を用いたアルゴリズムを作成した。患者と方 法:IFX投与患者80名、ETN投与患者69名の計149名を対象とした。p < 0.001の10 SNPs部位に関して、R群(あるいは副作用発現群)で頻度の高いアレルをホモで持つ患者を1点、ヘテロで持つ患者を0点、NR群(あるいは副作用非発現 群)で頻度の高いアレルをホモで持つ患者を-1点とし、計10 SNPsでの合計点を算出した。結果:IFX有効性に関してはR群の95%が合計点5点以上であり、NR群の92%が4点以下であった。さらにIFX副作 用、ETN有効性、副作用に関しても90-98%の割合で合計点により両群を分けることが可能であった。結論:SNPアルゴリズムを用いることにより、治 療前にその有効性や副作用発現を高確率で予測可能であることが示唆された。

4.    演者:松原 司、舟橋恵子、鳥山紗由美、中原邦彦、萩原敬史、三浦孝子、奥田康介、児矢野聡. 演題名:ゲノムワイドSNP解析によるアダリムマブ (ADA) とトシリズマブ (TOC) 治療RA患者における有効例 (R)、無効例 (NR) と副作用発現を予測するアルゴリズムの開発. 第53回日本リウマチ学会総会・学術集会. 東京 (2009)  目的:ADA及びTOCの有効、無効あるいは副作用発現を治療前に予測する方法はなく、そのためのSNPアルゴリズムの開発を試みた。患者と方法:ADA 投与患者21名、TOC投与患者30名を対象とした。p < 0.001の10 SNPs部位に関して、R群(あるいは副作用発現群)で頻度の高いアレルをホモで持つ患者を1点、ヘテロで持つ患者を0点、NR群(あるいは副作用非発現 群)で頻度の高いアレルをホモで持つ患者を-1点とし、計10 SNPsでの合計点を算出した。結果:ADA有効性に関してはR群の100%が合計点7点以上であり、NR群の100%が-1点以下であった。さらに ADA副作用、HAHA発現、TOC有効性、副作用に関しても95-100%の割合で合計点により両群を分けることが可能であった。結論:SNPアルゴリ ズムを用いることにより、治療前にADAおよびTOCの有効性や副作用発現を高確率で予測可能であることが示唆された。

5.    演者:児矢野聡、舟橋恵子、松原 司. 演題名:RANK関連遺伝子SNPsと関節リウマチ (RA) 患者の関節破壊進行度との関連性. 第53回日本リウマチ学会総会・学術集会. 東京 (2009)  目的:本演題では破骨細胞形成に深く関わるRANK関連遺伝子SNPsと関節破壊進行度との関連性を調べた。患者と方法:RA患者136症例は関節破壊進 行度の指標であるSteinbrokerのstage I及びII(mild)、stage III及びIV(severe)に分類した。続いてRANK関連遺伝子SNPs型と関節破壊進行度との関連をカイ自乗解析で調べた。結果:発症5年未満の 48症例の解析では、RANK, RANKL及びERKのSNPs において、mildとsevereの間で出現頻度差が見られた(p = 0.03-0.04)。機能障害では、OPG, VDR及びNFATC1のSNPsにおいて両群間で出現頻度差が見られた(p = 0.02-0.04)。結論:関節破壊進行度の個人差にRANK関連遺伝子SNPsが深く関与している可能性が大きい。

6.    演者:児矢野聡、舟橋恵子、松原 司. 演題名:FAS/CD95関連遺伝子SNPsと関節リウマチ (RA) 患者の関節破壊進行度との関連性. 第53回日本リウマチ学会総会・学術集会. 東京 (2009)  目的:本演題ではアポトーシス経路に深く関わるFAS関連遺伝子SNPsと関節破壊進行度との関連性を調べた。患者と方法:RA患者136症例は関節破壊 進行度の指標であるSteinbrokerのstage I及びII(mild)、stage III及びIV(severe)に分類した。続いてFAS関連遺伝子SNPs型と関節破壊進行度との関連をカイ自乗解析で調べた。結果:発症5年未満の 48症例の解析では、β-1-integrin, NOS3, ERK, CASP-8, CASP-9のSNPs において、mildとsevereの間で出現頻度差が見られた(p = 0.002-0.05)。機能障害では、CD44とCASP-3のSNPsにおいて両群間で出現頻度差が見られた(p = 0.008-0.04)。結論:FAS関連遺伝子SNPsの出現頻度の違いによってRAの関節破壊進行度の個人差が生じる可能性が大きい。

7.    演者:舟橋恵子、鳥山紗由美、児矢野聡、松原 司. 演題名:抗TNFα製剤投与関節リウマチ患者 (RA) に対するゲノム薬理学的検索 -効果と副作用に対する検討-. 第52回日本リウマチ学会総会・学術集会. 札幌 (2008)

8.    演者:松原 司、舟橋恵子、鳥山紗由美、児矢野聡. 演題名:R A疾病重症度に関する広域SNPs解析を用いた遺伝子学的解析. 第52回日本リウマチ学会総会・学術集会. 札幌 (2008)


[国際学会発表]

1. Authors: Matsubara T, Koyano S, Funahashi K, Toriyama S, Nakahara K, Hagiwara T, Miura T, Okuda K, Sagawa A, Sakurai T, Matsuno H, Izumihara T and Shono E: Title: An algorithm using genome-wide SNP analysis for prediction of responders and non-responders, and adverse events in tocilizumab-treated RA patients. The 73th annual meeting of the American College of Rheumatology (ACR). Philadelphia, PA, USA (2009) Purpose: Tocilizumab, a human anti-IL-6 receptor antibody, is an efficient biologic agent for inflammatory diseases such as RA. However, there is no method for prediction of responders, non-responders, and adverse events which can occur during the treatment. We established an SNP algorithm for prediction of responders or non-responders, and adverse events among tocilizumab-treated RA patients. Patients and Methods: One hundred RA patients treated with tocilizumab were included in this study. The efficacy was determined by Clinical Disease Activity Index (CDAI) within 24-30 weeks after the initial treatment. The efficacy of tocilizumab was judged by the scores of CDAI (remission and low disease activity group- 'responders', moderate and high disease activity group- 'nonresponders'). Adverse events such as leukopenia, high total cholesterol, fever, and skin manifestations were documented. Genome-wide SNP genotyping was performed by Illumina HumanHap300K chip technology. Case-control analyses between 285,548 SNPs and CDAI were examined by Fisher’s exact tests. We selected 10 SNPs strongly associated with tocilizumab- responsiveness, or adverse events (p < 0.001). Results: Accuracy ((true positive+true negative)/total), specificity (true negative/(false positive+true negative)) and sensitivity (true positive/(true positive+false negative)) of the algorithm for responsiveness of tocilizumab ranged 92-97%. For adverse events, accuracy, specificity and sensitivity of the algorithm ranged 90-97%. It is, therefore, suggested that the SNP algorithm predict responders and adverse events prior to the initiation of treatment with this biologic agent. Conclusion: The highly accurate algorithm using SNP analysis may be useful in the prediction of responsiveness and adverse events before treatment of tocilizumab, and in this way can contribute to future tailor-made treatment with biologic agents.

2. Authors: Matsubara T, Koyano S, Funahashi K, Toriyama S, Nakahara K, Hagiwara T, Miura T, Okuda K, Sagawa A, Sakurai T, Matsuno H, Izumihara T and Shono E: Title: Validation of an algorithm using genome-wide SNP analysis for prediction of responders and non-responders, and adverse events of infliximab- or etanercept-treated RA patients by using two population samples from multiple medical cohorts. The 73th annual meeting of the American College of Rheumatology (ACR). Philadelphia, PA, USA (2009) Purpose: Infliximab (IFX) and etanercept (ETN) are efficient biologic agents for inflammatory diseases such as RA. However, there is no method for prediction of responders or non-responders, and which patients are prone to adverse events. We established and validated an SNP algorithm for prediction of responders or non-responders, and adverse events among IFX- or ETN-treated RA patients by using multiple medical cohorts. Patients and Methods: The first population samples included 187 RA patients and the second population samples included 206 patients, total 393 patients from 6 hospitals in different regions of Japan. Efficacy was determined by DAS28(CRP) within 24-30 weeks after the initial treatment with the biologics according to EULAR criteria (good and moderate response group- 'responders', poor response group- 'nonresponders'), and adverse events such as fever, skin manifestations, and GI tract symptoms were documented. Genome-wide SNP genotyping was performed by HumanHap300K chip. Case-control analyses between 285,548 SNPs and efficacy or adverse events were examined by Fisher’s exact tests. We selected 10 SNPs associated with IFX- or ETN- responsiveness, or adverse events which are common in both analyses of the first and second populations (p < 0.02). Results: For IFX responsiveness, accuracy ((true positive+true negative)/total) of the algorithm of the first, second and total populations were 85.1%, 80.4% and 82.8%. For IFX adverse events, accuracy of the algorithm of the first, second and total populations were 86.3%, 88.0% and 87.3%. In responsiveness of ETN, accuracy of the algorithm of the first, second and total populations were 95.5%, 96.3% and 95.9%. The accuracy of the algorithm of ETN adverse events in the first, second and total populations were 78.3%, 88.8% and 83.4%. An endocrine hormone-acting gene was revealed in the associated genes, linking association of TNF with the endocrine hormone. Conclusion: The highly accurate algorithm using SNP analysis may prove useful in the prediction of responsiveness or determining those patients prone to adverse events before treatment of IFX or ETN, and may facilitate future tailor-made treatment of anti-TNF biologic agents.

3. Authors: Koyano S, Funahashi K and Matsubara T: Title: Involvement of RANK and RANKL gene SNPs in joint degradation in rheumatoid arthritis. The 72th annual meeting of the American College of Rheumatology (ACR). San Francisco, CA, USA (2008) Purpose: Receptor activator of NF-kappaB (RANK) cascade system has been reported to be essential for osteoclastogenesis. Osteoclasts,activated by a variety of cytokines, are critically associated with RA joint degradation. We examined the association of SNPs on RANK and RANKL genes in RA joint destruction. Patients and Methods: 136 RA patients were enrolled in this study. Genotypings were performed by Illumina HumanHap 300K chip. RA joint destruction of all patients was estimated by Steinbroker's stage classification. Forty eight of 136 patients were transferred to anti-TNF therapy within 5 years of onset, after at least two failed DMARD therapies. Twenty two patients were in Steinbroker's stage I, II (mild joint destruction) and 26 in stage III, IV (severe joint destruction) at the initiation of anti-TNF therapy. Genotypes of SNPs on RANK and RANKL genes were extracted from HumanHap 300K chip genotyping results. Association of these SNPs with joint destruction was examined by the case-control analysis (mild joint destruction group vs. severe group). Results: In the patient group of less than 5 years' duration, there was a marked difference in frequency in 2 SNPS on the RANK gene between the mild group and the severe group (p = 0.02-0.03). In the patient group of more than 5 years' duration, a significant difference in frequency was observed in 2 SNPs on the RANKL gene as compared between the 2 cohorts (p = 0.01-0.02). In all patients,the frequency of a further 5 SNPs on the RANK gene differed between the mild and severe cohorts (p = 0.01-0.05). Conclusion: The results suggested that SNPs on RANK and RANKL genes correlate with RA joint destruction and we believe that individual differences in osteoclastogenesis caused by these SNPs frequency may be responsible, at least in part, for progress of the disease.

4. Authors: Matsubara T, Funahashi K, Toriyama S and Koyano S: Title: Algorithm using genome-wide SNP analysis for prediction of responder and non-responder, and adverse events of Infliximab- or Etanercept-treated RA patients. The 72th annual meeting of the American College of Rheumatology (ACR). San Francisco, CA, USA (2008) Purpose: Infliximab and Etanercept are efficient biologic agents for inflammatory diseases such as RA. However, there are no method for prediction of responder or non-responder, and adverse events for Infliximab- or Etanercept-treated patients. Here, we examined prediction of responder or non-responder, and adverse events among Infliximab- or Etanercept-treated RA patients by SNP analysis. Patients and Methods: 150 RA patients with treatment with Infliximab or Etanercept were included in this study. The efficacy was determined by DAS28(CRP) values within 24-30 weeks after the initial treatment of the biologics according to EULAR criteria (good and moderate response group: responder, poor response group: nonresponder), and adverse events such as fever, skin manifestations, and GI tract reactions were documented. Genome-wide SNP genotyping was performed by Illumina HumanHap300K chip technology. Case-control analyses between 285,548 SNPs and clinical determination were examined by chi-square tests. We selected 10 SNPs with strongly associated with Infliximab- or Etanercept- responsiveness, or their adverse events (p < 0.01). Then, we scored relationship between each SNP and the responsiveness, estimated total score of 10 SNPs (estimated scoring in each SNP was as follows: homo allele in the majority in responder: +1 point, hetero allele: 0 point, and homo allele in the majority in non-responder: -1 point), and examined relationships between responder and non-responder, or adverse events plus or minus, and the total score. Results: Approximately 93% of the responder group has more than score 2 points, and ~95% of the non-responder group has less than score 1 point in Infliximab-responsiveness using this algorithm. For adverse events in Inflximab-treated group, ~94% of the plus group has more than score 4 points, and ~98% of the minus group has less than score 3 points. Similarly, for Etanercept-treated group, 90-95% of the responder, non-responder, adverse events plus or minus group could be judged by the algorithm. Conclusion: More patients are needed for more accurate prediction. Forward the future, this algorithm could be useful for prediction of responsiveness or adverse events before treatment of Infliximab or Etanercept.

5. Authors: Matsubara T, Funahashi K, Toriyama S, Itoh H, Muramatsu M, Emi M and Koyano S: Title: Genome-wide phararmacogenomics in anti-TNF therapy of rheumatoid arthritis: Implications for their efficacy and adverse reactions. The 71th annual meeting of the American College of Rheumatology (ACR). Boston, MA, USA (2007) Purpose: Infliximab, etanercept and adalimumab have shown clinical benefit in immune-mediated inflammatory diseases such as RA. However, there remains the problem of responders and non-responders to treatment with these anti-TNF-biologics, with no method available to predict adverse events. To discover gene(s) associated with efficacy and adverse reactions, we performed an association study of RA patients by genome-wide SNP analysis. Methods: 130 RA patients with treatment with anti-TNF-biologics were included in this study. The efficacy was determined by DAS28(ESR) or DAS28(CRP) values within 24-30 weeks after the initial treatment of the biologics, and adverse reactions such as fever, skin manifestations, and GI tract reactions were documented. Genome-wide SNP genotyping was performed by Illumina HumanHap300K chip technology. After evaluations of genotyping success rate > 90% and minor allele frequency > 1%, 285,548 SNPs were eligible for the study, out of a total of 316,994 SNPs. Case-control analyses between 285,548 SNPs and clinical determination were examined by chi-square tests. Results: 150 SNPs showed strong association with efficacy (p < 0.0005), and 88 SNPs, with adverse reactions (p < 0.0005). Especially, 12 and 6 genes included at least 2 SNPs with strong association with the efficacy and adverse reactions respectively (p < 0.0005). Furthermore, the significantly associated SNPs of these genes were in strong linkage disequilibrium according to HapMap data, indicating that the SNPs consisted of certain haplotypes. These genes resided in pathways in signal transduction, cell adhesion or metabolism. Conclusion: The associated SNPs are potential candidates for predicting the efficacy or adverse reactions in anti-TNF therapy of RA. Further replication and biological studies are warranted to validate these results. Genome-wide SNP analysis could be useful for the RA-related pharmacogenomics.

6. Authors: Matsubara T, Funahashi K, Toriyama S, Muramatsu M, Emi M and Koyano S: Title: A genome-wide SNP chip analysis revealed genes regulating the severity of rheumatoid arthritis. The 71th annual meeting of the American College of Rheumatology (ACR). Boston, MA, USA (2007) Purpose: Genome-wide association analysis using high throughput SNP typing technology has been a powerful tool for identifying disease-related genes and pharmacogenomics studies. The pathogenesis of many autoimmune diseases including RA remains unknown. To discover gene(s) regulating the severity of RA, we performed an association study of RA patients by genome-wide SNP analysis. Methods: 136 RA patients with various degrees of clinical severity were included in the study. DAS28(ESR) or DAS28(CRP) values for the RA patients before the initial treatment of the biologics including infliximab, etanercept, adalimumab or tocilizumab were used as clinical parameters (DAS28(ESR)(0 week) or DAS28(CRP)(0 week). Genome-wide SNP genotyping was performed by Illumina HumanHap300K chip technology. After evaluations of genotyping success rate > 90% and minor allele frequency > 1%, 285,548 SNPs were eligible for the study, out of a total of 316,994 SNPs. Association of 285,548 SNPs with the clinical parameter was examined by regression analysis. Results: SNPs in four genes showed strong association with both of DAS28(ESR)(0 week) and DAS(28CRP)(0week) values (p < 0.0005), and one gene was significantly associated with DAS28(CRP)(0 week) value (p < 0.0005). All of the five associated genes included at least 3 SNPs with strong association p-values (p < 0.0005). Furthermore, the significantly associated SNPs of the five genes were in strong linkage disequilibrium according to HapMap data, indicating that the SNPs consisted of certain haplotypes. The five genes resided in pathway in intracellular transport, gene transcription or metabolism. Conclusion: The associated genes are potential candidates for regulating the severity of RA. Further biological studies are warranted to investigate the mechanism by which these genes are involved in the pathogenesis of RA. Genome-wide SNP analysis could be useful for the RA-related gene search.