Antibiotic Therapy in Autoimmune Disorders

Review

Antibiotic therapy in autoimmune disorders

Yossi Rosman1, Merav Lidar2,3,4 & Yehuda Shoenfeld*3,4

Practice Points Infections may trigger autoimmunity through several mechanisms most notably; ‘molecular mimicry’, ‘epitope spreading’ and ‘bystander activation’. Minocycline, macrolides and fluoroquinolones may be recommended to patients with early and mild rheumatoid arthritis. Ciprofloxacin therapy may obviate chronic reactive arthritis in HLA-B27 patients, while antichlamydia-based antibiotic regimens may enhance chronicity. Trimethoprim/sulfamothoxazole should be part of the treatment protocol of granulomatosis with polyangitis. Helicobacter pylori eradication should be offered to all carriers with chronic thrombocytopenic purpura. Empiric antibiotic treatment should be considered as a standard part of the treatment protocol in catastrophic antiphosphlipid antibody syndrome.

SUMMARY: Antibiotics have been applied for the treatment of autoimmune diseases for over five decades, based on the premise that infections play a role in the initiation and propagation of these entities. The mechanisms by which an infection may trigger an autoimmune reaction include the so-called ‘molecular mimicry’, ‘epitope spreading’ or ‘bystander activation’. The association between infection and autoimmunity may be directly evident, as in cases of reactive arthritis, or in a more roundabout manner, as exemplified by the association between anaerobic bacterial infection of the gums and rheumatoid arthritis. Moreover, some antibiotics have, in addition to

1Department of Medicine D, Sheba Medical Center, Tel Hashomer, Israel 2Division of Rheumatology, Sheba Medical Center, Tel Hashomer, Israel 3Autoimmune Disease Center, Sheba Medical Center, Tel Hashomer, Israel 4Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel *Author for correspondence: [email protected] part of

their antibacterial effects, anti-inflammatory and immunomodulatory properties. In this review we focus on the rationale and possible benefits of antibiotic treatment in various autoimmune diseases, including rheumatoid arthritis, reactive arthritis, granulomatosis with polyangitis, immune thrombocytopenia purpura and the antiphospholipid syndrome.

Antibiotics have been applied for the treatment the association between anaerobic bacterial infec- of autoimmune diseases for over five decades, tion of the gums (periodontitis) and rheumatoid based on the premise that infections play a arthritis (RA) [5]. Epidemiological studies have role in the initiation and propagation of these shown that periodontal disease is more frequent entities. and more severe in patients with RA [6]. Also, the There are several proposed mechanisms by severity of periodontal disease was found to corre- which an infection may trigger an autoimmune late with RA severity [7], and treating periodontal reaction [1] . ‘Molecular mimicry’, in which the disease was demonstrated to improve clinical and pathogen and host share a common epitope laboratory parameters of the disease [8]. The pres- (i.e., proteins or DNA) resulting in the cross-acti- ence of peripathogenic bacteria in the synovium vation of autoreactive T or B cells by pathogen- of patients with RA suggests that joint seeding derived peptides is one of them [2,3]. The classic and localized inflammatory amplification may autoimmune disease rheumatic fever is thought initiate and propogate inflammation[9] . Nota- to be triggered by an infectious origin probably bly, the Gram-negative, anaerobic, coccobacillus via ‘molecular mimicry’ [4]. rheumatic fever is a Porphyromonas gingivalis has received consider- systemic disease, affecting the heart, joints, CNS able attention for its role in the development of and the skin, following Streptococcal infection periodontal disease and its association with RA. of the tonsils or the skin. In this disease, anti- It was shown that anti-P. gingivalis antibodies bodies against group A b-hemolytic Streptococ- were present in high titers in RA patients and cus are also active against host target, including that these antibodies were associated with the the heart and joints. Antibiotic treatment with presence of anticitrullinated cyclic peptides [10] . penicillin is known to prevent the autoimmune Moreover, in rat models, arthritis was induced by reaction when given during the acute infection, implanting killed P. gingivalis organisms subcu- or may prevent the continued deterioration when taneously, suggesting a role for molecular mim- given for a long period of time even after the icry [11] . Taken together, these findings suggest a infection was eliminated. pathological association between P. gingivalis and Other mechanisms by which infections may RA, hence the concept that antibiotics directed trigger an autoimmune reaction include ‘Epi- against P. gingivalis may be beneficial in the tope spreading’, where an epitope is switched treatment of RA, especially in early disease. from a dominant to a cryptic position result- Helicobacter pylori is an infectious agent note- ing in the creation of autoantibodies against worthy for its association with autoimmunity. the new epitope, is another possible initiator Anti-H. pylori antibodies have been associated of autoimmunity [1] . Additional mechanisms with antiphospholipid syndrome, giant cell arte- include ‘bystander activation’, in which tissue ritis, systemic sclerosis and primary biliary cir- damage results in the release of a new antigen, rhosis. Also, H. pylori infection has been impli- which activates autolymphocytes inducing an cated in the causation of various autoimmune autoinflammatory microenvironment, lead- diseases including immune thrombocytopenic ing in turn to the destruction of neighboring, purpura, autoimmune chronic gastritis and RA. uninfected cells [1] . The persistent presence of H. pylori in gas- The association between infection and auto- tric mucosa results in chronic immune system immunity may be directly evident, as in cases of activation with ongoing cytokine signaling, reactive arthritis (ReA), where a history of a recent infiltration of gastric mucosa by neutrophils, infection is elicited and in which bacterial prod- macrophages and lymphocytes, as well as pro- ucts may be detected in joints long after all signs duction of antibodies and effector T cells [12] . of infection have abated. Alternatively, infection Molecular mimicry between H. pylori antigen may contribute to the autoimmune cascade in and the H+/K+-ATPase (the ‘autoantigen’) leads a more roundabout manner, as exemplified by to autoimmune chronic gastritis, where the

92 Clin. Pract. (2014) 11(1) future science group Antibiotic therapy in autoimmune disorders | Review

+ activated CD4 T lymphocytes cross react with joint damage and disability [18]. Antibiotics have + + H /K -ATPase and H. pylori antigens [13] . been administered sporadically in the past in That being said, evidence indicates an overall RA, in an attempt to retard disease progression, downregulation of the host immune response in with minimal benefit. The newly recognized H. pylori-infected individuals [12] and this bacte- importance of periodontal disease in the patho- ria is also thought to play a protective role in the genesis of RA has sparked new interest in the development of multiple sclerosis, systemic lupus use of antibiotics in the treatment and preven- erythematosus and inflammatory bowel disease. tion of this common disease. These studies are Whether these links are epiphenomenal summarized in Table 1. or H. pylori does play a causative role in the autoimmune diseases remains uncertain. The Tetracyclines negative associations could possibly support Tetracyclines are a group of broad-spectrum the notion that in susceptible individuals infec- antibiotics used in the treatment of infections tions that may protect from the development of of the respiratory tract, urinary tract and intes- autoimmune diseases. tines. They were first advocated as therapeutic While these examples highlight the potential candidates in RA by scientists who presumed use of antibiotics in curtailing infections which that mycoplasma infection triggers the disease may initiate an autoimmune cascade, antibiotics and proposed prolonged antibiotic therapy with are also utilized for their anti-inflammatory and the aim of eradicating the bacteria [19] . Anec- immunomodulatory properties. Tetracyclines, dotal use of tetracyclines in the treatment of RA for instance, were shown to inhibit the activity persisted during the 1970s and early 1980s yet of antiphospholipase A2, scavenge free radicals the rheumatology community remained reserved and inhibit various matrix metalloproteinases in judging their efficacy. [14,15] as well as impair lymphocyte [16] activity In the late 1980s, the anti-inflammatory prop- and impart chondroprotective properties [17] . erties of tetracyclines were discovered leading While single infectious agents promote our to renewed interest in these drugs in the treat- understanding of autoimmunity by clarifying ment for RA. Two small, but positive studies, the basic concepts of molecular mimicry, epit- published in the early 1990s, paved the way to ope spreading and bystander activation, as noted larger trials, suggesting that the tetracycline above, ‘real-life’ autoimmune disease is more derivate – minocycline, indeed exerts substantial likely the result of exposure to numerous infec- antirheumatic properties [20,21]. tious agents rather than a single inciter. More- The first of these trials was a multicenter over, the concept of ‘superorganism’ and ‘micro- double-blinded, randomized controlled study biome’, describing the ensemble of human and conducted in The Netherlands. A total of 80 nonhuman (microorganism) cells that constitute patients with a history of more than 10 years of the human body, which has been developed by active RA, who had failed more than one dis- Nobel laureate Joshua Lederberg over the last ease-modifying antirheumatic drug (DMARD), decade, further complicates our understanding were divided into two groups. The study group of autoimmunity. It seems that the intestinal received 200 mg of minocycline once daily on microbiota are able to shape the immune sys- top of their regular therapy, while the control tem to maintain homeostasis and healthy states group received placebo. At the end of 26 weeks, or promote inflammation when the composi- there was a pronounced improvement in labo- tion of the microbiota community becomes ratory parameters of disease activity in the imbalanced. Therefore the effect of antibiotics minocycline-treated patients. Alas, improvement in autoimmune diseases is undoubtedly more in clinical parameters was less impressive [22]. far-reaching than fathomed at this time point. In the second trial, 219 RA patients, who had The following review shall elaborate on the failed one or more DMARDs were allocated to rationale and possible benefits of antibiotic therapy with 200 mg/day of minocycline in treatment in various autoimmune diseases. place of the previous DMARD, or to placebo. A significant improvement in joint tenderness Rheumatoid arthritis and swelling as well as in laboratory parameters RA, the most common, chronic, inflammatory including hematocrit, erythrocyte sedimentation joint disease, may be complicated by substantial rate (ESR), platelet count and IgM rheumatoid

future science group www.futuremedicine.com 93 Review | Rosman, Lidar & Shoenfeld

Table 1. Antibiotic use in rheumatoid arthritis. Author (year) Antibiotic Patient Study design Patients (n) Duration of Outcome Ref. population treatment and follow up Kloppenburg Minocycline 200 mg Active RA for Multicenter, 80 26 weeks Improvement [22] et al. (1994) q.d. (on top of regular more than double-blinded, in laboratory treatment) 10 years, after randomized, parameters failure of one or placebo-controlled more DMARD Tilley et al. Minocycline 200 mg Active RA, after Multicenter, 219 1 year Improvement [23] (1995) q.d., instead of failure of one or double-blinded, in clinical and regular treatment more DMARD randomized, laboratory placebo-controlled parameters O’Dell et al. Minocycline 100 mg Newly diagnosed Double-blinded, 46 4 years Improvement [25] (1999) b.i.d. RA randomized, in clinical and placebo-controlled laboratory parameters O’Dell et al. Minocycline Newly diagnosed Double-blinded, 60 2 years Improvement in [28] (2001) 100 mg b.i.d. versus RA randomized clinical parameters hydroxychloroquine controlled 200 mg b.i.d. van der Laan Doxycycline 100 mg Active RA, after Double-blinded, 66 36 weeks No improvement [29] et al. (2001) q.d. failure of one randomized DMARD placebo-controlled O’Dell et al. Doxycycline Newly diagnosed Double-blinded, 66 2 years Improvement in [30] (2006) 100 mg q.d. versus RA randomized, clinical parameters doxycycline 20 mg placebo-controlled with no difference q.d. versus between high placebo. All with MTX or low dose doxyciline Ogrendik Clarithromycin Active RA, after Double-blinded, 81 6 months Improvement [32] (2007) 500 mg q.d. failure of one or randomized, in clinical and more DMARD placebo-controlled laboratory parameters Ogrendik Roxithromycin Active RA Double-blinded, 100 6 months Improvement in [33] & Karagoz 300 mg q.d. randomized, clinical parameters (2011) placebo-controlled Ogrendik Levofloxacine 500 mg Active RA despite Double-blinded, 66 6 months Improvement [34] (2007) q.d. treatment with randomized, in clinical and MTX placebo-controlled laboratory parameters Smith et al. Tetracycline 250 mg Active RA, after Double-blinded, 50 25 weeks No improvement [35] (2011) b.i.d., 3 times failure of one randomized, per week + iv. DMARD placebo-controlled clindamycin (tapering doses) b.i.d: Two times a day; DMARD: Disease-modifying antirheumatic drug; iv.: Intravenous; MTX: Methotrexate; q.d.: Once a day; RA: Rheumatoid arthritis.

factor levels were noted among patients treated of 3 months, 65% of the study population versus with minocycline [23]. 13% of the controls improved by 50% or more Finally, minocycline was given to 46 newly diag- in morning stiffness, joint tenderness, joint swell- nosed RA patients in a 6-month double-blinded, ing and ESR levels [24]. Moreover, after 4 years of randomized controlled trial. These patients, who follow-up, eight patients (40%) initially random- had not received DMARDs or steroids in the past, ized to minocycline were still receiving this com- were allocated into a 100 mg minocycline twice- pound and were in remission without additional daily (b.i.d.) group and a control group. At the end DMARD therapy [25].

94 Clin. Pract. (2014) 11(1) future science group Antibiotic therapy in autoimmune disorders | Review

The sum of these studies prompted the expert [MTX]) it proved to have a beneficial effect on opinion published in 2000, supporting the use the ACR50 response [30]. of minocycline in early RA [26,27]. However, only in 2001 was the efficacy of minocycline Macrolides compared with that of a conventional DMARD, Macrolides are a group of antibiotics that are used hydroxychloroquine, utilizing the American in the treatment of infections caused by Gram- College of Rheumatology (ACR) response cri- positive, as well as oral anaerobic, bacteria in addi- teria to measure outcome. In a double-blinded tion to possessing inherent anti-inflammatory randomized controlled study, 60 patients with properties [31] . The discovery of the association newly diagnosed RA, who had not been pre- between RA and periodontitis, discussed previ- viously treated with DMARDs, were random- ously, prompted the 2007 study that tested the ized to receive minocycline, 100 mg b.i.d., or efficacy of 500 mg of clarithromycin in patients hydroxychloroquine, 200 mg b.i.d., in addition with active RA who had had an inadequate to low-dose prednisone. After 2 years of follow response to at least one previous DMARD [32]. up, patients in the minocycline group were more At 6 months, significantly more patients on clar- likely to achieve an ACR50 response compared ithromycin therapy achieved an ACR20 and/or with hydroxychloroquine-treated patients (60% 50 response, exhibited reductions in ESR and compared with 33%, respectively; p = 0.04) CRP levels and reported an improvement in and were also receiving less prednisone (mean quality of life. Similar results were demonstrated dose of 0.81 and 3.21 mg/day, in the minocy- in a study that tested the effects of a different cline and hydroxychloroquine groups, respec- macrolide – roxithromycine, 300 mg/day in RA tively; p < 0.01). In addition, patients treated patients [33]. with minocycline were more likely to have been completely tapered off prednisone (p = 0.03) [28]. Quinolones An additional double-blinded, placebo- Quinolones are antibiotics with activity against controlled crossover trial was performed in anaerobic bacteria. In total, 76 patients with 2001, studying the effect of another tetracy- active RA on MTX therapy were allocated to cline compound doxycycline, in RA [29]. In combination therapy with levofloxacin and this study, 66 patients with stable RA were MTX versus metothrexate and placebo in a ran- randomized to receive either low-dose (100 mg/ domized, double-blinded study. At 6 months, a day) doxycycline or placebo, and followed for significant improvement in all measures of dis- 36 weeks. Lamentably, doxycycline therapy had ease activity was noted including a significant no effect on any of the clinical, laboratory or reduction in the number of swollen joints in the radiographic parameters assessed. It was specu- study group. A significant improvement in sec- lated that the disparity in therapeutic efficacy ondary end points including a reduction in pain between minocycline and doxycycline was a and duration of morning stiffness, improvement result of distinct chemical properties, most nota- in the quality of life, physician’s and patient’s bly the enhanced lipophilicity of minocycline global assessments, as well as in objective labora- compared with doxycycline, which favors its tory measures was also noted [34]. The effect of distribution within the synovium. Presumably, levofloxacin was attributed to its activity against the enhanced lipophilicity of minocycline also oral anaerobic bacteria, as previously discussed. enables it to more readily penetrate the blood– These small trials with positive results using brain barrier and act on the parts of the CNS a macrolide or a quinolone raise the need for that influence the immune system, contributing more large-scale trials comparing the addition to its immunosuppressive effects. However, one of a macrolide or a quinolone to the conven- should bear in mind that most of the positive tional treatment protocol, including the use of effects attributed to minocycline were noted biological treatment, for severe active RA. in populations with early RA which were not exposed to prior DMARD therapy, whereas Combined antibiotic treatment the study with doxycycline was performed in The effect of combined therapy with intrave- patients with well-established RA. Indeed, nous clindamycin in a tapering dose and oral when doxycycline was tested in patients with tetracycline 250 mg b.i.d., 3 days a week, was early RA (in combination with methotrexate investigated in 50 patients with active RA who

future science group www.futuremedicine.com 95 Review | Rosman, Lidar & Shoenfeld

had failed therapy with MTX or sulphasalazine resolution within 6 months while the other half alone. The trial was stopped early owing to lack continue to suffer from chronic symptoms. of improvement in any of the designated out- On account of ReA being triggered by a preced- comes, which was ascribed, retrospectively, to ing bacterial infection and as bacteria or bacterial an insufficient dose of tetracycline[35] . products may be detected in patients joints by various techniques, the utilization of antibiotics in H. pylori eradication in RA this entity is clear. The studies regarding antibiotic As previously discussed, H. pylori infection has treatment for ReA are summarized in Table 2. been associated not only with chronic gastritis and peptic ulcers, but also with many extrain- Acute ReA testinal disorders and autoimmune diseases. The effect of a short course of antibiotic treat- Zentilin et al. studied the effect of H. pylori ment in the acute phase of postenteric ReA was eradication on the outcome of patients with RA studied by Fryden et al. in 1990 [37]. In this ran- [36]. A total of 58 patients with active RA and domized placebo-controlled study, 40 patients dyspepsia were divided into two groups. The were assigned to a 10–14 day course of antibiotic first, which consisted of 28 patients who had treatment or placebo. Choice of antibiotic ther- tested positive for a urea breath test for H. pylori, apy was determined according to the suspected received eradication therapy with amoxicillin, pathogen. No difference was noted between the whereas the second group, which suffered from groups at 18 months of follow-up. dyspepsia but was negative for H. pylori, received Sieper et al. assessed the effect of ciprofloxacin no therapy. At the end of a 2‑year follow-up, a administered for 3 months in 55 patients and significant reduction in the number of swollen showed a beneficial effect of ciprofloxacin over joints and laboratory markers was noted in the placebo in the subgroup who developed ReA first group, in whichH. pylori was eradicated. postchlamydia infection [38]. A similar study was This is a small preliminary study, regarding the conducted by Yli-Kerttula et al. in 2000 [39]. possible association between RA and H. pylori. A total of 62 patients with acute ReA received More large-scale randomized controlled studies either ciprofloxacin (500 mg b.i.d.) or placebo are necessary to better define this association. for 3 months. After 1 year of follow-up, clinical and laboratory parameters improved to a similar Summary of antibiotic therapy in RA extent in both groups. Interestingly, despite a The sum of evidence suggests that antibiotics similar outcome at 1 year, a significantly lower may have a role in DMARD-naive, early RA incidence of chronic arthritis was evident in patients (minocycline) as well as in addition patients originally allocated to treatment with to standard DMARD therapy in established ciprofloxacin after 4–7 years [40]. Early treat- disease (macrolides and quinolones). The ben- ment with ciprofloxacin was also associated with eficial effects of the antibiotic in these cases a lower incidence of ankylosing spondylitis and may be ascribed to immunomodulatory effects anterior uveitis in the future. in the former, or antibacterial effects, specifi- The evidence that antibiotic treatment may cally – antioral bacteria, in the latter. The data improve outcome in ReA developing post regarding H. pylori eradication in RA are not chlamydia infection, gave rise to studies using strong enough at this point in order to issue a specific antichlamydia antibiotics Putschky et al. recommendation. Antibiotic therapy has yet to for example, compared a long (100 mg b.i.d. be assessed in RA patients treated with biologics. for 4 months) with a short (100 mg b.i.d. for 10 days) course of doxycycline in the acute phase Reactive arthritis of postchlamydia ReA [41] . Surprisingly, clinical ReA is an inflammatory arthritis that typically and laboratory outcome were no different among develops 1–4 weeks following gastrointestinal or the two groups. Contrary to expectations, a genitourinary bacterial infections. Various bacte- favorable trend for a better clinical response was ria have been implicated in the cause of ReA, most discerned in the short course antibiotic group. notably, Chlamydia trachomatis, as well as Sal- The unfavorable outcome of long term antibiotic monella, Shigella, Campylobacter sp. and Yersinia treatment is supported by studies showing persis- entrocolitica. Approximately half of the individu- tence of chlamydia in affected joints after such als infected with ReA experience spontaneous treatment [42].

96 Clin. Pract. (2014) 11(1) future science group Antibiotic therapy in autoimmune disorders | Review

Table 2. Antibiotic use in reactive arthritis. Author (year) Antibiotic Patient Study design Patients Duration Follow-up Outcome Ref. population (n) of therapy period Fryden et al. According to Acute post- Multicenter 40 10–14 days 18 months No clinical or [37] (1990) the suspected enteric ReA randomized, placebo- laboratory pathogen controlled improvement Sieper et al. Ciprofloxacin Acute ReA Double-blinded, 55 3 months 1 year A mild clinical [38] (1999) 500 mg b.i.d. randomized, placebo- improvement only controlled in postchlamydia Yli-Kerttula Ciprofloxacin Acute ReA Double-blinded, 62 3 months 1 year Both groups [39] et al. (2000) 500 mg b.i.d. randomized, placebo- improved the same controlled Yli-Kerttula Ciprofloxacin Acute ReA Double-blinded, 53 3 months 4–7 years Significantly less [40] et al. (2003) 500 mg b.i.d. randomized, placebo- chronic arthritis in controlled the study group Putschky et al. Doxycycline Acute Double-blinded, 32 10 days 4 months No difference [41] (2006) 100 mg b.i.d. postchlamydia randomized placebo- versus between the two ReA controlled 4 months groups Kvien et al. Azithromycin Acute ReA Double-blinded, 152 3 months 6 months No clinical or [43] (2004) 1 g once randomized, placebo- laboratory weekly controlled improvement Smieja et al. Doxycycline Chronic ReA Double-blinded, 37 3 months 3 months No clinical [44] (2001) 100 mg b.i.d. randomized, placebo- improvement controlled Carter et al. Doxycycline Post chlamydia Double-blinded, 42 6 months 9 months A clinical and [45] (2010) 100 mg chronic ReA randomized, placebo- laboratory b.i.d. versus controlled improvement in azithromycin both treatment 500 mg q.d. groups both on top of rifampin 300 mg q.d. b.i.d: Two times a day; q.d.: Once a day; ReA: Reactive arthritis.

A second study, using a different anti ReA (11 patients) did not show better results chlamydia regiment – azithromycin – begin- than the rest of the study population [44]. ning in the acute phase of ReA was conducted Carter et al. examined the effect of a long- by Kvien et al. In total, 152 patients with acute term combined antibiotic treatment in chronic ReA (less than 2 months duration) received ReA post chlamydia infection [45]. In this one dose of azithromycin followed by weekly double-blinded, triple placebo-controlled trial, azithromycin or placebo for 3 months. No dif- 42 patients with chronic (at least 6 months) ReA ferences in any clinical or laboratory parame- were divided into three groups. The first group ters were observed after 24 weeks, although the (12 patients) received long-term antibiotic treat- postchlamydia ReA subgroup was not analyzed ment with doxycycline and rifampin, the sec- separately due to its small size [43]. ond group received azithromycin and rifampin (15 patients) and the third group received Chronic ReA placebo. Treatment was given for 6 months The efficacy of doxycycline in chronic ReA was and assessments were performed at 9 months. assessed by Smieja et al. in a double-blinded Clinical and laboratory markers significantly placebo-controlled study, in which 37 patients improved in both treatment groups compared with ReA of more than 4 months duration, were with the placebo group, suggesting that long- treated with doxycycline (100 mg b.i.d.) or pla- term combined antibiotic treatment may have a cebo. At 3 months follow-up, no differences were role in treating chronic ReA due to chlamydia observed between the two treatment groups. infection. The difference between these two Moreover, the subgroup with postchlamydia small, but well-conducted, trials may be due

future science group www.futuremedicine.com 97 Review | Rosman, Lidar & Shoenfeld

to the addition of rifampin in the second trial. the effect of antibiotic treatment on the clinical More large-scale randomized controlled trials course of the disease. DeRemee et al. was first are needed for further recommendations. to demonstrate an improved clinical outcome in 11 patients with GPA who received trim- Summary of antibiotic therapy in ReA ethoprim/sulfamethoxazole (TMP‑SMX) [50]. Somewhat unpredictably, antibiotics seem to This observation led to a number of case reports have a limited and, in certain instances, even a and case series that reported a possible benefi- counterproductive role in ReA despite the rec- cial effect of TMP‑SMX in patients with GPA ognized infectious trigger of the disease. Aver- with or without cytotoxic treatment [51–53]. The sion of chronic disease, especially in HLAB27- impression from the sum of these reports was positive patients, was noted in only a single study that TMP‑SMX may be beneficial in patients in which ciprofloxacin therapy was instituted in with chronic nasal carriage of S. aureus and acute ReA. However, other studies have mostly a limited disease. Noncarriers of S. aureus, as shown a neutral effect on the acute and chronic well as those with a more generalized disease, course of the disease, whereas in chlamydia- tended not to achieve remission or to relapse induced ReA, early use of antibiotics was asso- often on this treatment [54]. However, to date, ciated with persistent disease. Until larger scale no randomized control trial studying the effect studies become available, it may be prudent to of TMP‑SMX in acute GPA has been conducted. withhold antibiotics in ReA. Conversely, the effect of maintenance prophylactic treatment with TMP‑SMX on relapses Systemic vasculitis in GPA was assessed in several controlled trials. The systemic vasculitides are characterized by Stegeman et al., conducted a double-blinded, inflammation of blood vessels occurring in a placebo-controlled, multicenter trial, in order variety of organ systems. Although the patho- to assess the efficacy of TMP‑SMX on the num- physiology and etiology of these diseases has not ber of relapses in patients with GPA in remis- been fully elucidated, bacterial infections have sion [55]. In total, 81 patients were randomly been recognized in both the induction and reac- assigned to receive TMP‑SMX (800 mg of SMX tivation of some of these disorders. ‘Granuloma- and 160 mg of TMP) or placebo twice daily for tosis with polyangitis’ (GPA, Wegener’s granulo- 24 months, on top of their usual medications. matosis) is a small–medium vessel sized vasculitis After 24 months, 82% of the study patients associated with PR3‑antineutrophil cytoplasmic were in remission as compared with 60% of antibodies manifested by granulomatous inflam- the controls (relative risk: 0.4; 95% CI: 0.17– mation in the upper and lower airways as well 0.98). Disease involving the upper airways was as necrotizing glomerulonephritis. Stegeman the most prominently reduced (10 vs 32.5% et al. described the association between chronic in the study and control groups, respectively). nasal carriage of S. aureus and a relapse of GPA Notwithstanding, nausea and anorexia led in their cohort of 57 patients with GPA followed to discontinuation of therapy in 20% of the for a median of three years. A total of 63% of the patients. The protective effect of TMP‑SMX patients had chronic nasal carriage of S. aureus, was ascribed to the elimination of S. aureus which was found to be an independent risk factor from the upper airways although no actual for relapse (adjusted relative risk: 7.16; 95% CI: data on carriage post-therapy was presented. 1.63–31.50) [46]. The association between GPA, A reduction in respiratory and nonrespiratory nasal carriage of S. aureus and disease relapse infections among the study group possibly was confirmed by others[47,48] . Presumably, attenuates potential triggers for relapse. The colonization with S. aureus induces the produc- utility of TMP‑SMX in maintaining remission tion of low-grade proinflamatory cytokines that in patients with GPA was assessed by Zycinska prime neutrophil/monocytes for antineutrophil et al. [56]. In a prospective, double-blinded, pla- cytoplasmic antibodies-induced activation, or cebo-controlled study, 31 patients with GPA in alternatively, a superantigen from the bacteria remission were assigned to TMP‑SMX (800 mg stimulates autoreactive lymphocytes [49]. of SMX and 160 mg of TMP, twice daily) or The evidence that chronic bacterial infec- placebo. At the end of 18 months follow-up, tion may play a role in the pathogenesis of GPA 75% of the study group was in remission versus prompted several trials which aimed to study 55% of the control group (hazard ratio: 0.8;

98 Clin. Pract. (2014) 11(1) future science group Antibiotic therapy in autoimmune disorders | Review

95% CI: 0.21–1.2). TMP‑SMX treatment was to higher platelets counts after 4 months. As an independent predictor of disease-free interval mentioned previously, it was speculated that the (0.4; 95% CI: 0.12–0.69), whereas S. aureus association between H. pylori and ITP may be chronic nasal carriage was an independent a product of the immunologic response to the risk factor for disease relapse (6.15; 95% CI: bacteria resulting in the release of cytokines 2.15 –33.2). and autoantibodies. In a prospective open study The beneficial effect of TMP‑SMX in GPA, conducted in 2001, H. pylori was found by 13C hitherto delineated, was questioned by de Groot urea breath test in 43.3% (13/30) ITP patients et al. [57,58]. In their first study, the efficacy of that were assessed [60]. Eradication was success- TMP‑SMX (800 mg of SMX and 160 mg of ful in 12 of the 13 patients, after which com- TMP, twice daily) compared with MTX, in plete response was noted in four patients and maintaining remission in 65 patients with partial response in two more. The only patient GPA, was assessed. MTX alone maintained that failed eradication remained with low platelet remission in 86% of the patients versus 58% counts. The favorable outcome was achieved over with TMP‑SMX (p < 0.05). Importantly, all of 8 months of follow-up. In contrast to these favor- the patients relapsed when low-dose prednisone able results, other prospective open studies failed was added to TMP‑SMX. In the second study, to demonstrate a beneficial effect of eradication TMP‑SMX was shown to be effective in induc- of H. pylori in patients with ITP [61,62]. To our ing remission when given to patients in the ini- knowledge, the only randomized control trial was tial phase of GPA (58% success rate), but was carried out in Japan in 2004 by Suzuki et al. [63]. ineffective in maintaining remission in patients In this randomized controlled trial, 25 patients with generalized GPA (42% relapse rate versus with chronic ITP on no current treatment and 29% in the untreated group). with a positive breath test for H. pylori, were randomly assigned to eradication therapy versus Antibiotics in systemic vasculitis observation. In total, 46% (six/13) of the study Studies conducted on antibiotic use in systemic group increased their platelet counts at 6 months vasculitis have centered on the role of TMP‑SMX follow-up, while no change was noted among the in GPA. Taken together it seems that TMP‑SMX control group. After 6 months, the control group is effective in inducing and maintaining remis- received eradication therapy yielding a 30% sion in GPA, more so in mild and limited dis- response (four/12) exemplified by elevation of ease, possibly through elimination of chronic platelet counts. There was no difference in clini- S. aureus nasal carriage. cal characteristics or H. pylori virulence factors between the responders and nonresponders. In Idiopathic thrombocytopenic purpura 2007, Franchini et al. published a meta-analysis Idiopathic thrombocytopenic purpura (ITP) on the effect of H. pylori eradication in chronic is an autoimmune disorder, characterized by ITP [64]. Data on 788 ITP patients was collected antibodies to platelets, resulting in thrombo from 17 studies (16 with a prospective cohort cytopenia. Generally this disorder is highly design and one randomized trial). A statistically responsive to steroids, Rho(D) immunoglobulin significant difference in the increase in platelet or intravenous immunoglobulin. Some cases are count was documented in patients in whom refractory, requiring more extreme therapeutic H. pylori eradication was successful compared measures such as splenectomy or chronic steroid with controls: 40.77 platelets (95% CI: 20.92– treatment. As indicated by its name, the cause 60.63) compared with untreated patients; 52.16 for ITP is unknown. Gasbarrini et al. were first platelets (95% CI: 34.26–70.05) compared with to report a possible association between ITP and patients who failed eradication and 46.35 platelets H. pylori, a bacteria that has been implicated in (95% CI: 27.79–64.91) compared with H. pylori- gastritis and peptic disease [59]. In their report, negative patients. Moreover, in the meta-regres- H. pylori was identified by 13‘ C urea breath test’ sion model, the success of H. pylori eradication in 11 patients with ITP. Eradication of the bac- was highly significant as an explanatory variable teria employing amoxicillin (1000 mg b.i.d.), for platelet count increase. Not surprisingly, the clarithromycin (250 mg three times a day) and beneficial effect of H. pylori eradication was pentoprazole (40 mg once daily) led to resolu- only demonstrated in patients with chronic ITP tion of the ITP in eight cases (73%) as well as that were found to be positive for H. pylori. As

future science group www.futuremedicine.com 99 Review | Rosman, Lidar & Shoenfeld

demonstrated by Arnold et al. in their systemic resulted in the disappearance of antiphospho- review, H. pylori eradication therapy was found to lipid antibodies with improvement of all clinical be of little benefit forH. pylori-negative patients signs and symptoms [71] . with chronic ITP [65]. Summary of antibiotic therapy in APS Summary of antibiotic therapy in ITP Despite the evidence that links infections to An increased association of H. pylori carriage has APS, data proving the efficacy of antibiotic been noted in numerous autoimmune diseases treatment in these acute situations are lacking. suggesting a causal, if yet unproved, relation- Notwithstanding, due to the severity of the ship. Based on the meta-analysis cited, provi- clinical situation, it is our common practice to sion of antibiotic therapy to H. pylori-positive administer antibiotic therapy to all patients with patients with chronic ITP should be routinely catastrophic APS, even when an infectious cause initiated prior to more expensive and more is obscure. immune-suppressive therapy. Undifferentiated connective tissue disease Antiphospholipid antibody syndrome Undifferentiated connective tissue disease The classic antiphosphlipid antibody syn- (UCTD) is an autoimmune disease with symp- drome (APS) is characterized by the presence toms/signs of connective tissue disease, and the of antiphospholipid antibodies acting mainly presence of one or more positive autoimmune dis- through B2GP1, recurrent fetal losses and ease tests, but with insufficient criteria to make thromboembolic events. ‘Catastrophic APS’ is a definitive diagnosis. Moskowitz et al. studied an unusual and life-threatening manifestation the effect of a 12-week course of clarithromycin of the APS defined by Asherson et al. in 1992 in seven patients suffering from UCTD [72]. In [66]. The syndrome is characterized by the rapid this open label study, clinical improvement was development of multiple blood clots in several noted in six of the seven patients at the end of organs; most commonly the heart, lungs, ner- the treatment course. vous system and kidneys. An association between When this limited amount of data are taken APS and infections was documented by Cervera together with the recognition that UCTD is a et al. who described 100 cases of APS that were mixed entity which tends to develop over time preceded by an infection, typically pneumonia, into an established disease, and as antibiotic cellulitis, urinary tract infections and HIV [67]. therapy has shown a limited benefit in most Others have described an association of APS autoimmune disease entities, we see no role for with Streptococcus spp. M. pneumonia, Coxiella therapeutic antibiotics in UCTD at this point. burnetti and Mycobacterium leprae [67]. The association between catastrophic APS and infection Conclusion is even more remarkable as Asherson et al. had The potential beneficial role of antibiotics in documented that 24% of cases of catastrophic the treatment of autoimmune diseases is based APS develop in these clinical situations [68]. on two main premises. The first is the assump- The proposed mechanisms by which infection tion that eradication of an infectious process, induces APS include molecular mimicry between which may both initiate and propagate autoim- the pathogen and the B2GP1 molecule, poly- munity, will curtail an ongoing stimulus and clonal activation of lymphocytes and stimulation lead to disease amelioration. Second, several of cytokines by proteins of the infectious agent antibiotics have shown to have inherent anti- [69]. Blank et al. showed that infecting mice with inflammatory properties, and as such were uti- various agents including Haemophilus influenza, lized as a DMARD. In some instances, as in the Neisseria gonorrhea and Tetanus toxoid produce case of RA, antibiotics were assessed in small, a hexapeptide which is in turn recognized by a but well-designed trials showing substantial pathogenic anti-B2GP1 antibody. The induction effects in early and mild disease. Given the newly of this hexapeptide leads to clinical and labora- recognized importance of periodontal disease tory manifestations of APS in association with in RA as well as the cost of the newer biologic elevated titers of anti-B2GP1 [70]. and nonbiologic DMARDS, we suggest a revival Cicconi et al. described a 33 year old woman of minocycline, macrolide and quinolone use in with APS in which eradication of H. pylori early and mild RA, as well as in cases where

100 Clin. Pract. (2014) 11(1) future science group Antibiotic therapy in autoimmune disorders | Review

periodontal disease is evident. Finer tuning of Trials should be conducted in life-threatening these recommendations deserves larger scale diseases for which effective therapies are lacking clinical studies. Similarly, the recommendation and which have not yet been intensively inves- for TMP‑SMX therapy in GPA is well founded tigated, such as catastrophic APS, acute and and is the mainstay of treatment in mild disease. chronic GPA, and chronic ITP. A less recognized association that is highlighted Controlled trials to establish the efficacy of in the present review is the beneficial effect of antibiotics in the treatment of RA, in which a H. pylori eradication in chronic ITP, which is bacterial culprit has been identified (i.e., perio consolidated by a meta-analysis. One should dontitis), and in which uncontrolled studies in consider testing for H. pylori carriage in all early disease have shown promise, should be such patients, most definitely before immuno performed. suppressive agents are added and splenectomy is performed. As for catastrophic APS, despite Financial & competing interests disclosure the lack of published cases on the role of anti The authors have no relevant affiliations or financial biotics in this clinical scenario, we believe that involvement with any organization or entity with a finan- the evidence pointing to an infectious process in cial interest in or financial conflict with the subject matter the initiation of this calamity is reason enough to or materials discussed in the manuscript. This includes institute prompt antibiotic therapy in all cases. employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or Future perspective pending, or royalties. Controlled trials studying the effect of antibiotics No writing assistance was utilized in the production of in autoimmune diseases are needed. this manuscript.

systemic inflammation in rheumatoid arthritis 16 Gabler WL, Creamer HR. Suppression of References patients. Oral Dis. 19(4), 394–400 (2013). human neutrophil functions by tetracyclines. 1 Kivity S, Agmon-Levin N, Blank M, 9 Moen K, Brun JG, Valen M et al. Synovial J. Periodontal Res. 26(1), 52–58 (1991). Shoenfeld Y. Infections and inflammation in active rheumatoid arthritis 17 Yu LP Jr, Smith GN Jr, Brandt KD, Myers autoimmunity – friends or foes? Trends and psoriatic arthritis facilitates trapping of a SL, O’Connor BL, Brandt DA. Reduction of Immunol. 30(8), 409–414 (2009). variety of oral bacterial DNAs. Clin. Exp. the severity of canine osteoarthritis by 2 Rashid T, Ebringer A. Autoimmunity in Rheumatol. 24(6), 656–663 (2006). prophylactic treatment with oral doxycycline. rheumatic diseases is induced by microbial 10 Mikuls TR, Payne JB, Reinhardt RA et al. Arthritis Rheum. 35(10), 1150–1159 (1992). infections via crossreactivity or molecular Antibody responses to Porphyromonas 18 Harris ED Jr. Rheumatoid arthritis. mimicry. Autoimmune Dis. 2012, 539282 gingivalis (P. gingivalis) in subjects with Pathophysiology and implications for therapy. (2012). rheumatoid arthritis and periodontitis. Int. N. Engl. J. Med. 322(18), 1277–1289 (1990). 3 Proal AD, Albert PJ, Marshall TG. The Immunopharmacol. 9(1), 38–42 (2009). 19 Williams MH, Brostoff J, Roitt IM. Possible human microbiome and autoimmunity . Curr. 11 Bartold PM, Marino V, Cantley M, Haynes role of Mycoplasma fermentans in pathogenesis Opin. Rheumatol. 25(2), 234–240 (2013). DR. Effect of Porphyromonas gingivalis- of rheumatoid arthritis. Lancet 2(7667), 4 Cunningham MW. Streptococcus and induced inflammation on the development of 277–280 (1970). rheumatic fever . Curr. Opin. Rheumatol. rheumatoid arthritis. J. Clin. Periodontol. 20 Langevitz P, Bank I, Zemer D, Book M, Pras 24(4), 408–416 (2012). 37(5), 405–411 (2010). M. Treatment of resistant rheumatoid 5 Bingham CO 3rd, Moni M. Periodontal 12 Blaser MJ, Atherton JC. Helicobacter pylori arthritis with minocycline: an open study. disease and rheumatoid arthritis: the evidence persistence: biology and disease. J. Clin. J. Rheumatol. 19(10), 1502–1504 (1992). accumulates for complex pathobiologic Investig. 113(3), 321–333 (2004). 21 Breedveld FC, Dijkmans BA, Mattie H. interactions 25(3), . Curr. Opin. Rheumatol. 13 Hasni SA. Role of Helicobacter pylori infection Minocycline treatment for rheumatoid 345–353 (2013). in autoimmune diseases. Curr. Opin. arthritis: an open dose finding study. 6 Dissick A, Redman RS, Jones M et al. Rheumatol. 24(4), 429–434 (2012). J. Rheumatol. 17(1), 43–46 (1990). Association of periodontitis with rheumatoid 14 Alarcon GS, Mikhail IS. Antimicrobials in the 22 Kloppenburg M, Breedveld FC, Terwiel JP, arthritis: a pilot study. J. Periodontol. 81(2), treatment of rheumatoid arthritis and other Mallee C, Dijkmans BA. Minocycline in 223–230 (2010). arthritides: a clinical perspective. Am. J. Med. active rheumatoid arthritis. A double-blind, 7 Abou-Raya S, Abou-Raya A, Naim A, Sci. 308(3), 201–209 (1994). placebo-controlled trial. Arthritis Rheum. Abuelkheir H. Rheumatoid arthritis, 15 Greenwald RA, Moak SA, Ramamurthy NS, 37(5), 629–636 (1994). periodontal disease and coronary artery Golub LM. Tetracyclines suppress matrix 23 Tilley BC, Alarcon GS, Heyse SP et al. disease. Clin. Rheumatol. 27(4), 421–427 metalloproteinase activity in adjuvant arthritis Minocycline in rheumatoid arthritis. A (2008). and in combination with flurbiprofen, 48‑week, double-blind, placebo-controlled 8 Erciyas K, Sezer U, Ustun K et al. Effects of ameliorate bone damage. J. Rheumatol. 19(6), trial. MIRA Trial Group. Ann. Intern. Med. periodontal therapy on disease activity and 927–938 (1992). 122(2), 81–89 (1995).

future science group www.futuremedicine.com 101 Review | Rosman, Lidar & Shoenfeld

24 O’Dell JR, Haire CE, Palmer W et al. 37 Fryden A, Bengtsson A, Foberg U et al. Early granulomatosis as compared with rheumatoid Treatment of early rheumatoid arthritis with antibiotic treatment of reactive arthritis arthritis and chronic Rhinosinusitis with minocycline or placebo: results of a associated with enteric infections: clinical and nasal polyps. Clin. Exp. Rheumatol. randomized, double-blind, placebo-controlled serological study. BMJ 301(6764), 1299–1302 28(1 Suppl. 57), S51–S55 (2010). trial. Arthritis Rheum. 40(5), 842–848 (1997). (1990). 48 Zycinska K, Wardyn KA, Zielonka TM, 25 O’Dell JR, Paulsen G, Haire CE et al. 38 Sieper J, Fendler C, Laitko S et al. No benefit Demkow U, Traburzynski MS. Chronic Treatment of early seropositive rheumatoid of long-term ciprofloxacin treatment in crusting, nasal carriage of Staphylococcus arthritis with minocycline: four-year followup patients with reactive arthritis and aureus and relapse rate in pulmonary of a double-blind, placebo-controlled trial. undifferentiated oligoarthritis: a three-month, Wegener’s granulomatosis. J. Physiol. Arthritis Rheum. 42(8), 1691–1695 (1999). multicenter, double-blind, randomized, Pharmacol. 59(Suppl. 6), S825–S831 (2008). 26 Langevitz P, Livneh A, Bank I, Pras M. placebo-controlled study. Arthritis Rheum. 49 Popa ER, Stegeman CA, Abdulahad WH Benefits and risks of minocycline in 42(7), 1386–1396 (1999). et al. Staphylococcal toxic-shock-syndrome- rheumatoid arthritis. Drug Saf. 22(5), 39 Yli-Kerttula T, Luukkainen R, Yli-Kerttula U toxin-1 as a risk factor for disease relapse in 405–414 (2000). et al. Effect of a three month course of Wegener’s granulomatosis. Rheumatology 27 Alarcon GS. Tetracyclines for the treatment of ciprofloxacin on the outcome of reactive (Oxford) 46(6), 1029–1033 (2007). rheumatoid arthritis. Expert Opin. Investig. arthritis. Ann. Rheum. Dis. 59(7), 565–570 50 DeRemee RA, McDonald TJ, Weiland LH. Drugs 9(7), 1491–1498 (2000). (2000). Wegener’s granulomatosis: observations on 28 O’Dell JR, Blakely KW, Mallek JA et al. 40 Yli-Kerttula T, Luukkainen R, Yli-Kerttula U treatment with antimicrobial agents. Mayo Treatment of early seropositive rheumatoid et al. Effect of a three month course of Clin. Proc. Mayo Clin. 60(1), 27–32 (1985). arthritis: a two-year, double-blind comparison ciprofloxacin on the late prognosis of reactive 51 Valeriano-Marcet J, Spiera H. Treatment of of minocycline and hydroxychloroquine. arthritis. Ann. Rheum. Dis. 62(9), 880–884 Wegener’s granulomatosis with Arthritis Rheum. 44(10), 2235–2241 (2001). (2003). sulfamethoxazole-trimethoprim. Arch. Intern. 29 van der Laan W, Molenaar E, Ronday K et al. 41 Putschky N, Pott HG, Kuipers JG, Zeidler H, Med. 151(8), 1649–1652 (1991). Lack of effect of doxycycline on disease Hammer M, Wollenhaupt J. Comparing 52 Ohtake T, Kobayashi S, Honjou Y et al. activity and joint damage in patients with 10‑day and 4‑month doxycycline courses for Generalized Wegener’s granulomatosis rheumatoid arthritis. A double blind, placebo treatment of Chlamydia trachomatis-reactive responding to sulfamethoxazole-trimethoprim controlled trial. J. Rheumatol. 28(9), arthritis: a prospective, double-blind trial. monotherapy. Intern. Med. 40(7), 666–670 1967–1974 (2001). Ann. Rheum. Dis. 65(11), 1521–1524 (2006). (2001). 30 O’Dell JR, Elliott JR, Mallek JA et al. 42 Dreses-Werringloer U, Padubrin I, Jurgens- 53 Israel HL. Sulfamethoxazole-trimethoprim Treatment of early seropositive rheumatoid Saathoff B, Hudson AP, Zeidler H, Kohler L. therapy for Wegener’s granulomatosis. Arch. arthritis: doxycycline plus methotrexate versus Persistence of Chlamydia trachomatis is Intern. Med. 148(10), 2293–2295 (1988). induced by ciprofloxacin and ofloxacin methotrexate alone. Arthritis Rheum. 54(2), 54 Kallenberg CG. What is the evidence for in vitro. Antimicrob. Agents Chemother. 621–627 (2006). prophylactic antibiotic treatment in patients 44(12), 3288–3297 (2000). 31 Ianaro A, Ialenti A, Maffia P et al. Anti- with systemic vasculitides? Curr. Opin. inflammatory activity of macrolide antibiotics. 43 Kvien TK, Gaston JS, Bardin T et al. Three Rheumatol. 23(3), 311–316 (2011). month treatment of reactive arthritis with J. Pharmacol. Exp. Ther. 292(1), 156–163 55 Stegeman CA, Tervaert JW, de Jong PE, azithromycin: a EULAR double blind, (2000). Kallenberg CG. Trimethoprim- placebo controlled study. Ann. Rheum. Dis. 32 Ogrendik M. Effects of clarithromycin in sulfamethoxazole (co-trimoxazole) for the 63(9), 1113–1119 (2004). patients with active rheumatoid arthritis. prevention of relapses of Wegener’s Curr. Med. Res. Opin. 23(3), 515–522 (2007). 44 Smieja M, MacPherson DW, Kean W et al. granulomatosis. Dutch Co-Trimoxazole Randomised, blinded, placebo controlled trial 33 Ogrendik M, Karagoz N. Treatment of Wegener Study Group. N. Engl. J. Med. of doxycycline for chronic seronegative rheumatoid arthritis with roxithromycin: 335(1), 16–20 (1996). arthritis. Ann. Rheum. Dis. 60(12), a randomized trial. Postgrad. Med. 123(5), 56 Zycinska K, Wardyn KA, Zielonka TM, 1088–1094 (2001). 220–227 (2011). Krupa R, Lukas W. Co-trimoxazole and 45 Carter JD, Espinoza LR, Inman RD et al. 34 Ogrendik M. Levofloxacin treatment in prevention of relapses of PR3‑ANCA positive Combination antibiotics as a treatment for patients with rheumatoid arthritis receiving vasculitis with pulmonary involvement. Eur. chronic chlamydia-induced reactive arthritis: methotrexate. Southern Med. J. 100(2), J. Med. Res. 14(Suppl. 4), S265–S267 (2009). a double-blind, placebo-controlled, 135–139 (2007). 57 de Groot K, Reinhold-Keller E, Tatsis E et al. prospective trial. Arthritis Rheum. 62(5), Therapy for the maintenance of remission in 35 Smith A, Dore C, Charles P, Vallance A, 1298–1307 (2010). Potier T, Mackworth-Young C. Randomised sixty-five patients with generalized Wegener’s 46 Stegeman CA, Tervaert JW, Sluiter WJ, double-blind trial of combination antibiotic granulomatosis. Methotrexate versus Manson WL, De Jong PE, Kallenberg CG. therapy in rheumatoid arthritis. Int. trimethoprim/sulfamethoxazole. Arthritis Association of chronic nasal carriage of J. Rheumatol. 2011, 585497 (2011). Rheum. 39(12), 2052–2061 (1996). Staphylococcus aureus and higher relapse rates 58 Reinhold-Keller E, de Groot K, Rudert H, 36 Zentilin P, Seriolo B, Dulbecco P et al. in Wegener granulomatosis. Ann. Intern. Med. Nolle B, Heller M, Gross WL. Response to Eradication of Helicobacter pylori may reduce 120(1), 12–17 (1994). disease severity in rheumatoid arthritis. trimethoprim/sulfamethoxazole in Wegener’s 47 Laudien M, Gadola SD, Podschun R et al. Aliment. Pharmacol. Ther. 16(7), 1291–1299 granulomatosis depends on the phase of Nasal carriage of Staphylococcus aureus and (2002). disease. Q JM 89(1), 15–23 (1996). endonasal activity in Wegener s

102 Clin. Pract. (2014) 11(1) future science group Antibiotic therapy in autoimmune disorders | Review

59 Gasbarrini A, Franceschi F, Tartaglione R, thrombocytopenic purpura-a randomized characteristics of 100 patients. Ann. Rheum. Landolfi R, Pola P, Gasbarrini G. Regression controlled trial. Am. J. Gastroenterol. Dis. 63(10), 1312–1317 (2004). of autoimmune thrombocytopenia after 100(6), 1265–1270 (2005). 68 Asherson RA, Shoenfeld Y. The role of eradication of Helicobacter pylori. Lancet 64 Franchini M, Cruciani M, Mengoli C, infection in the pathogenesis of catastrophic 352(9131), 878 (1998). Pizzolo G, Veneri D. Effect of Helicobacter antiphospholipid syndrome – molecular 60 Emilia G, Longo G, Luppi M et al. pylori eradication on platelet count in mimicry? J. Rheumatol. 27(1), 12–14 (2000). Helicobacter pylori eradication can induce idiopathic thrombocytopenic purpura: a 69 Shoenfeld Y, Blank M, Cervera R, Font J, platelet recovery in idiopathic systematic review and meta-analysis. Raschi E, Meroni PL. Infectious origin of the thrombocytopenic purpura. Blood 97(3), J. Antimicrob. Chemother. 60(2), 237–246 antiphospholipid syndrome. Ann. Rheum. Dis. 812–814 (2001). (2007). 65(1), 2–6 (2006). 61 Jarque I, Andreu R, Llopis I et al. Absence of 65 Arnold DM, Bernotas A, Nazi I et al. 70 Blank M, Krause I, Fridkin M et al. Bacterial platelet response after eradication of Platelet count response to H. pylori induction of autoantibodies to Helicobacter pylori infection in patients with treatment in patients with immune b2‑glycoprotein-I accounts for the infectious chronic idiopathic thrombocytopenic thrombocytopenic purpura with and etiology of antiphospholipid syndrome. purpura. Br. J. Haematol. 115(4), without H. pylori infection: a systematic J. Clin. Investig. 109(6), 797–804 (2002). 1002–1003 (2001). review. Haematologica 94(6), 850–856 71 Cicconi V, Carloni E, Franceschi F et al. (2009). 62 Michel M, Khellaf M, Desforges L et al. Disappearance of antiphospholipid antibodies Autoimmune thrombocytopenic Purpura 66 Asherson RA. The catastrophic syndrome after Helicobacter pylori eradication. and Helicobacter pylori infection. Arch. antiphospholipid syndrome. J. Rheumatol. Am. J. Med. 111(2), 163–164 (2001). Intern. Med. 162(9), 1033–1036 (2002). 19(4), 508–512 (1992). 72 Moskowitz RW, Lesko M, Hooper M. Open- 63 Suzuki T, Matsushima M, Masui A et al. 67 Cervera R, Asherson RA, Acevedo ML et al. label study of clarithromycin in patients with Effect of Helicobacter pylori eradication in Antiphospholipid syndrome associated with undifferentiated connective tissue disease. patients with chronic idiopathic infections: clinical and microbiological Semin. Arthritis Rheum. 36(2), 82–87 (2006).

future science group www.futuremedicine.com 103