Treatment of Early Seropositive Rheumatoid Arthritis
Doxycycline Plus Methotrexate Versus Methotrexate Alone
James R. O’Dell,
1Jennifer R. Elliott,
2Jack A. Mallek,
3Ted R. Mikuls,
1Cynthia A. Weaver,
4Scott Glickstein,
5Kent M. Blakely,
6Raymond Hausch,
7and Rob D. Leff
7Objective.To compare the efficacy of doxycycline plus methotrexate (MTX) versus MTX alone in the treatment of early seropositive rheumatoid arthritis (RA), and to attempt to differentiate the antibacterial and antimetalloproteinase effects of doxycycline.
Methods. Sixty-six patients with seropositive RA of <1 year’s duration who had not been previously treated with disease-modifying antirheumatic drugs were randomized to receive 100 mg of doxycycline twice daily with MTX (high-dose doxycycline group), 20 mg of doxycycline twice daily with MTX (low-dose doxycycline group), or placebo with MTX (placebo group), in a 2-year double-blind study. Treatment was started with an MTX dosage of 7.5 mg/week, which was titrated every 3 months until remission was reached (maximum dos-age of 17.5 mg/week). The primary end point was an American College of Rheumatology 50% improvement (ACR50) response at 2 years.
Results.ACR50 responses were observed in 41.6%
of patients in the high-dose doxycycline group, 38.9% of those in the low-dose doxycycline group, and 12.5% of patients in the placebo group. Results of chi-square analysis of the ACR50 response in the high-dose doxy-cycline group versus that in the placebo group were significantly different (P ⴝ 0.02). Trend analysis vealed that the ACR20 response and the ACR50 re-sponse were significantly different between groups (Pⴝ
0.04 andPⴝ0.03, respectively). MTX doses at 2 years were not different among groups. Four patients in the high-dose doxycycline group, 2 patients in the low-dose doxycycline group, and 2 patients in the placebo group were withdrawn because of toxic reactions.
Conclusion. In patients with early seropositive RA, initial therapy with MTX plus doxycycline was superior (based on an ACR50 response) to treatment with MTX alone. The therapeutic responses to low-dose and high-dose doxycycline were similar, suggesting that the antimetalloproteinase effects were more important than the antibacterial effects. Further studies to evalu-ate the mechanism of action of tetracyclines in RA are indicated.
Methotrexate (MTX) has clearly been shown to be effective first-line therapy for rheumatoid arthritis (RA). Despite the efficacy of MTX as single-drug ther-apy, combination therapy has been shown to be effective for patients with early RA (1–3). Recent serial surveys of US rheumatologists revealed increasing trends toward the use of combination therapy and earlier therapy (4). Minocycline has also been used in early RA and has been shown to be superior to placebo as well as hydroxy-chloroquine (HCQ) (5,6). Hyperpigmentation and diz-ziness have been associated with minocycline use. Sec-ondary to these side effects and in an attempt to lower costs, an alternative tetracycline, doxycycline, was used in this study.
1James R. O’Dell, MD, Ted R. Mikuls, MD, MPH: University
of Nebraska Medical Center, Omaha; 2Jennifer R. Elliott, MD:
University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania;
3Jack A. Mallek, MD: Central Plains Clinic, Sioux Falls, South Dakota; 4Cynthia A. Weaver, MD: Black Hills Neurology and Rheumatology,
Rapid City, South Dakota;5Scott Glickstein, MD: Park Nicollet Clinic,
Minneapolis, Minnesota;6Kent M. Blakely, MD: Platte Valley
Medi-cal Group, Kearney, Nebraska;7Raymond Hausch, MD, Rob D. Leff,
MD: St. Mary’s Duluth Clinic Health System, Duluth, Minnesota. Dr. O’Dell has received consulting fees or honoraria (less than $10,000 per year) from Bristol-Myers Squibb, Amgen, Abbott, and Genentech. Dr. Mikuls has received consulting fees or honoraria (less than $10,000 per year) from TAP Pharmaceuticals. Dr. Glickstein owns stock in Amgen and Schering-Plough. Dr. Blakely has received consulting fees or honoraria from Merck, Amgen, and Wyeth.
Address correspondence and reprint requests to Jennifer R. Elliott, MD, University of Pittsburgh Medical Center, Division of Rheumatology and Clinical Immunology, S716 Biomedical Science Tower, 3500 Terrace Street, Pittsburgh, PA 15261. E-mail: elliottjr@upmc.edu.
Submitted for publication January 13, 2005; accepted in revised form October 31, 2005.
Although the mechanism of action of tetracy-clines in RA is poorly understood, tetracytetracy-clines, in addition to their antibacterial effects, inhibit matrix metalloproteinases (MMPs) and are immunomodula-tory. In an attempt to elucidate whether the antibacterial or the antimetalloproteinase effect is more important, 2 different dosages of doxycycline were used: the conven-tional antibiotic dosage of 100 mg twice daily and a low dosage of 20 mg twice daily. Doxycycline at a dosage of 20 mg twice daily is marketed to treat gingivitis and has been shown not to alter oral or gut flora.
PATIENTS AND METHODS
This study was conducted by investigators from the Rheumatoid Arthritis Investigational Network (RAIN), which brings rheumatologists at the University of Nebraska Medical Center together with rheumatologists in Nebraska, South Dakota, North Dakota, Minnesota, California, and Oregon who are interested in clinical studies in patients with RA. All physicians participating in this network were involved not only with patient enrollment and data collection but also in the development of the study protocols. Enrollment occurred at 9 centers, from March 1999 to October 2001.
Patient selection.Patients with RA who were followed up at rheumatology clinics at the University of Nebraska Medical Center, the Omaha Veterans Administration Medical Center, or in the private offices of network physicians were asked to participate in this study, if they met the study criteria. The Institutional Review Board at the University of Nebraska Medical Center and at other sites, where applicable, approved the protocol. All patients gave their informed, written consent. The eligibility criteria were as follows: age between 19 years and 75 years, RA fulfilling the American College of Rheumatology (ACR; formerly, the American Rheumatism Association) criteria (7), rheumatoid factor (RF) positivity, and duration of disease⬎6 weeks and⬍1 year. Patients were not eligible if they had received previous disease-modifying antirheumatic drug (DMARD) therapy, if they were allergic to or intolerant of tetracycline or MTX, if they had received dosages of oral steroids that were unstable or were ⬎7.5 mg/day, or if they had received intraarticular injections within the last few weeks. Other patients who were excluded were those with significant liver, renal, hematologic, pulmonary, cardiovascular, or active peptic ulcer disease, women who were pregnant, women who wished to become pregnant in the next 2 years, women who were not using successful contraceptive methods, patients who were not willing to abstain from alcohol consumption, and patients with diabetes who were receiving insulin or oral hypoglycemic agents.
Study design.We enrolled 66 patients in this 2-year, double-blind, randomized, controlled trial. The pharmacy han-dled the randomization; equal numbers of cards with each group assignment were mixed, drawn, and placed in sequen-tially numbered envelopes that were opened as the patients were enrolled. The randomization was done in blocks by the central pharmacy, thus ensuring that inequalities across cen-ters did not occur. Patients were randomized into the following
groups: 100 mg of doxycycline twice daily plus MTX (high-dose doxycycline group), 20 mg of doxycycline twice daily plus MTX (low-dose doxycycline group), or placebo plus MTX (placebo group) (Figure 1). During the last 6 months of enrollment, patients were not randomized into the low-dose doxycycline group. This modification of blinded randomization for the last 6 months occurred because of low numbers of patients being enrolled.
In all patients, treatment with MTX was started at a dosage of 7.5 mg/week, which was titrated up every 3 months until the ACR criteria for remission were met or a maximum dosage of 17.5 mg/week was reached. In addition, if a patient met the remission criteria at 12 months, the MTX dosage was tapered by 2.5 mg/month as long as the ACR 50% improve-ment (ACR50) response was met. Patients who had been receiving stable dosages of prednisone (⬍7.5 mg/day) before enrollment were allowed to continue receiving that dosage, and, starting at 9 months, the dosage was tapered off by 1 mg/month. If a patient required prednisone at 18 months and did not meet the ACR50 criteria, he or she was withdrawn from the study and considered an efficacy failure.
All patients underwent evaluations initially and then every 3 months, after which clinical decisions on medication dosages were made. The assessments consisted of a tender and swollen joint count, the Health Assessment Questionnaire (HAQ), physician’s and patient’s global assessment of disease activity, and patient’s assessment of pain. Laboratory data (complete blood cell count, aspartate aminotransferase [AST], creatinine, and albumin levels) were collected every 4 weeks, and the erythrocyte sedimentation rate (ESR) was determined initially and then every 6 months. The determination of RF and all other monitoring laboratory assessments were done locally. At 12 months, if patients did not meet ACR20 criteria and had been receiving the maximum dosage of MTX (17.5 mg/week) for 6 months, they were dropped from the study and
Figure 1. Flow diagram of patient enrollment and outcomes within the randomized groups. Secondary to low enrollment numbers, the final 12 patients were randomized to the doxycycline 100 mg group and the placebo group. ACR50⫽ American College of Rheumatology 50% improvement response.
considered efficacy failures. The blinded portion of the proto-col ended after the 2-year evaluation.
Evaluation criteria. The primary end point was an ACR50 response. The ACR core criteria for clinical response (8) require a 50% improvement in both the tender and swollen joint counts and a 50% improvement in 3 of the following 5 parameters: patient’s global assessment of disease activity, physician’s global assessment of disease activity, and patient’s assessment of pain on a 100-mm visual analog scale (VAS), HAQ score, and ESR. The primary end point (an ACR50 response at 2 years) was an intent-to-treat analysis. All patients who completed the study early as efficacy failures or because of toxicity were considered failures.
Joint counts were performed using a modified Ritchie Articular Index (9), in which 38 joints were scored as being involved or not involved (score of 0 or 1, respectively) for tenderness and swelling. The patient’s assessment of global status and overall pain, as well as the physician’s global assessment of disease activity, were scored using a VAS (0⫽ normal and 10⫽severe disease) (8). DNA, serum, and urine samples were collected and banked for future study. Individual components of the ACR core set of criteria for clinical response (8), the ACR20, ACR50, and ACR70 responses, and ending MTX dosages were also reported.
Toxicity monitoring. All patients were questioned about adverse reactions, and routine laboratory data were monitored every 4 weeks. If AST levels were elevated on 2 consecutive occasions, the dosage of MTX was decreased or MTX was discontinued. If AST values remained elevated, the patient was withdrawn from the study and considered a failure due to toxicity. Toxicity-related withdrawals occurred when the rheumatologist thought that a clinical event was severe enough to warrant unbinding and withdrawal of the patient from the study.
Concurrent therapy.Patients who were receiving non-steroidal antiinflammatory medications at stable doses were permitted to continue doing so, and no changes were made once the patient was enrolled. Similarly, if a patient was receiving a stable dosage of prednisone (⬍7.5 mg/day), that dosage was held constant and then was decreased by 1 mg/month starting at 9 months. A total of 2 intraarticular steroid injections were allowed during the study but not within 6 weeks prior to the 12-month or 24-month evaluation. Folic acid (at a maximum dosage of 4 mg/day) was permitted at the discretion of the treating rheumatologist.
Power calculations. The study was designed to have 80% power to detect the anticipated difference between
groups. Based on the assumption of a 75% response rate in the MTX plus doxycycline groups and a 40% response in the MTX plus placebo group, enrollment of 24 patients in each group was needed. When it became clear that enrollment was signif-icantly slower than anticipated, the decision was made to randomize patients to only the placebo group and the high-dose doxycycline group. This allowed us to reach the goal of 24 patients in each group over the next 6 months.
Statistical analysis. The primary end point was an ACR50 response at 2 years. This was an intent-to-treat analy-sis, and all randomized patients were included. The chi-square trend test was used for comparison among treatment groups meeting the ACR50 criteria. Individual components of the ACR core set were analyzed by Wilcoxon’s test. Final MTX dosages were compared with analysis of variance (ANOVA).
RESULTS
Treatment groups. We randomly assigned 66 patients to 1 of 3 treatment groups: 100 mg of doxycy-cline twice daily plus MTX (n⫽ 24), 20 mg of doxycy-cline twice daily plus MTX (n ⫽ 18), or placebo plus MTX (n ⫽ 24). The demographic characteristics of patients in the 3 groups are shown in Table 1, and baseline disease characteristics are shown in Table 2. During the final year of enrollment, secondary to a low number of enrolled patients, the final 12 patients were randomized to the high-dose doxycycline group and the placebo group, thus explaining the different numbers of patients in the groups. As per protocol, all patients were RF positive, had a mean disease duration of 5 months, and had not received prior DMARD therapy. Despite having early disease, patients did have significant disease activity, with an average of 16–18 tender and swollen joints, a mean ESR of⬎35 mm/hour, and HAQ scores between 1.0 and 1.2. The baseline characteristics of the patients were similar, with no significant difference between age, disease duration, the number of tender joint and swollen joints, the ESR, and HAQ scores. Interestingly, patients in the placebo group were initially receiving higher dosages of prednisone (statistically
sig-Table 1. Patient characteristics at baseline*
Characteristic MTX⫹doxy 100 mg (n⫽24) MTX⫹doxy 20 mg (n⫽18) MTX⫹placebo (n⫽24)
Age, mean (range) years 49.5 (35–74) 49.9 (27–74) 55.7 (41–74)
Female sex, % 67 89 79
Caucasian race, % 88 100 88
Disease duration, mean⫾SD months 5.0⫾3.1 5.4⫾2.9 4.8⫾2.7
Rheumatoid factor positive, % 100 100 100
Starting dosage of prednisone, mean⫾SD mg/day
3.0⫾3.5 1.3⫾2.6 4.0⫾3.2
nificantly higher than the dosage received by patients in the low-dose doxycycline group); however, there was no difference between the dosages received by patient in the doxycycline-treated groups.
As shown in Figure 1, 12 (50%) of 24 patients in the high-dose doxycycline group, 7 (39%) of 18 patients in the low-dose doxycycline group, and 8 (33%) of 24 patients in the placebo group completed the full 2 years of the study; the remaining patients completed early as efficacy failures or withdrew due to adverse events, toxicities, the patient’s preference, or protocol viola-tions. Four patients in the high-dose doxycycline group were considered early efficacy failures, because 3 did not meet the ACR20 criteria at 12 months, and 1 failed to show a treatment response despite maximal MTX ther-apy. Four other patients in the high-dose doxycycline group were withdrawn secondary to patient desire (n⫽ 2) and protocol violation (n ⫽ 2). Five patients in the low-dose doxycycline group did not meet the ACR20 criteria (efficacy failure), 1 patient expressed the desire to be withdrawn, and 3 patients violated protocol. Nine patients in the placebo group were considered efficacy failures, with 8 not meeting the ACR20 criteria and 1 not responding to maximal MTX therapy. Five other pa-tients in the placebo group were dropped secondary to patient desire (n⫽3) and protocol violations (n⫽2), as shown in Figure 1.
Treatment-related toxicity. Four patients in the high-dose doxycycline group, 2 patients in the low-dose doxycycline group, and 2 patients in the placebo group withdrew because of toxic reactions. In the high-dose doxycycline group, withdrawals occurred because of an increased alanine aminotransferase (ALT) level (n⫽1), myocardial infarction (n⫽1), and photosensitivity rash (n⫽2). Two patients in the low-dose doxycycline group withdrew because of increased ALT levels. In the
pla-cebo group, 1 patient withdrew because of multiple medical problems (not thought to be secondary to the trial medication), and 1 patient withdrew because of a rash.
Primary outcome. The trial’s primary outcome was an ACR50 response. Ten of 24 patients (41.6%) receiving 100 mg of doxycycline twice daily, 7 of 18 patients (38.9%) receiving 20 mg of doxycycline twice daily, and 3 of 24 patients (12.5%) in the placebo group met the criteria for an ACR50 response (Figure 2). The trend test among groups meeting the ACR50 criteria (primary outcome) showed significant differences (P⫽ 0.03) favoring the doxycycline-treated groups. Results of chi-square analysis of the ACR50 response in the high-dose doxycycline group versus that in the placebo group were also significant (P⫽ 0.02).
Table 2. Disease characteristics at baseline*
Characteristic MTX⫹ doxy 100 mg MTX⫹ doxy 20 mg MTX⫹ placebo
Tender joint count† 16⫾7 16⫾7 18⫾9
Swollen joint count† 16⫾7 15⫾7 17⫾9
ESR, mm/hour 38⫾27 35⫾14 40⫾32
HAQ score 1.08⫾0.69 1.02⫾0.47 1.26⫾0.62
Physician global assessment of disease activity (0–10 VAS) 5⫾2 5⫾2 6⫾2 Patient global assessment of disease activity (0–10 VAS) 5⫾3 5⫾2 6⫾3
Patient assessment of pain (0–10 VAS) 5⫾2 5⫾3 6⫾2
Morning stiffness, minutes 151⫾90 151⫾84 134⫾76
Fatigue (0–10 VAS) 6⫾2 6⫾2 5⫾3
* Values are the mean ⫾ SD. MTX ⫽ methotrexate; doxy ⫽ doxycycline; ESR ⫽ erythrocyte sedimentation rate; HAQ⫽Health Assessment Questionnaire; VAS⫽visual analog scale.
† Modified 38-joint count (9), graded as involved or not involved.
Figure 2. American College of Rheumatology 20% improvement response (ACR20), ACR50, and ACR70 in each treatment group. All patients received concomitant methotrexate therapy. Results of trend test analysis of the ACR50 response (primary outcome) among treatment groups were statistically significant (P⫽0.03).
Other treatment outcomes. ACR20 and ACR70 responses are shown in Figure 2. Trend test analysis of the ACR20 responses among the groups showed signif-icant differences (P ⫽ 0.04). ACR70 response rates in the 3 groups were not significantly different (P⫽0.25); however, the trend favored the doxycycline-treated groups.
The individual components of the ACR core set were analyzed. At baseline, there were no significant differences between the 3 groups. Within all 3 groups, each core set component improved significantly from baseline. When comparing the data for the difference between baseline and 2 years, the high-dose and low-dose doxycycline groups were not significantly different. Table 3 shows doxycycline-treated patients compared with patients in the placebo group. For each core variable, the response to treatment was better in doxycycline-treated patients than in patients in the pla-cebo group, although only the difference for morning stiffness was statistically significant (by Wilcoxon’s test). The final mean MTX dosages at 2 years were 14.09 mg/week, 15.69 mg/week, and 14.78 mg/week in the high-dose doxycycline group, the low-dose doxycy-cline group, and the placebo group, respectively. ANOVA of the final MTX dosages revealed no statisti-cally significant difference among the groups.
DISCUSSION
In this trial, we demonstrated that in patients with early seropositive RA, initial therapy with MTX
plus doxycycline is superior (based on the ACR50 response) to treatment with MTX alone. Within each core component analyzed, the response between base-line and 2 years was better in the doxycycbase-line-treated patients than in the placebo group. The superior re-sponses are confirmed by the fact that there was no statistically significant difference between the final mean dosages of MTX in the 3 study groups.
Results of trials of minocycline in patients with established RA have revealed statistically significant improvement in minocycline-treated patients compared with those receiving placebo, especially with regard to laboratory parameters (10,11). The Minocycline in Rheumatoid Arthritis Trial (11) revealed modest im-provement in tender and swollen joints in minocycline-treated patients compared with those receiving placebo for established RA. Our network, RAIN, has completed 2 trials of minocycline in patients with early seropositive RA. Our first trial in patients with early RA showed that 65% of minocycline-treated patients versus 13% of placebo-treated patients met the Paulus criteria for 50% improvement (primary end point) at 6 months (5), and patients who were treated early with minocycline had more frequent remissions and less need for DMARD therapy at 4-year followup (12). In a second trial in patients with early RA, we compared minocycline plus active DMARD therapy with HCQ therapy and ob-served a 60% ACR50 response (primary end point) among the minocycline-treated patients compared with 33% of HCQ-treated patients; in addition, minocycline-treated patients required less prednisone therapy (6).
Results of a double-blind, placebo-controlled trial of doxycycline (50 mg orally twice daily by mouth) in patients with RA showed that doxycycline had no therapeutic benefit (13). Two small trials of intravenous doxycycline in patients with established RA also did not reveal any benefits of doxycycline in these patients (14,15). A small single-blind prospective trial demon-strated that treatment of RA patients with doxycycline or MTX resulted in similar outcomes (16). A recent meta-analysis of randomized controlled trials of tetracy-cline in patients with RA revealed a small reduction in the tender and swollen joint counts, with a larger reduction in the levels of acute-phase reactants; in addition, subgroup analysis revealed minocycline to be more effective than doxycycline (17).
Tetracycline was initially used in the treatment of RA becauseMycoplasmawas thought to be the cause of RA. We now know that tetracyclines do not treat an infectious component of RA but have biologic effects on the inflammatory and immunologic cascade. In the
Table 3. Changes in individual disease parameters, baseline to 2 years* Treatment group MTX⫹ doxycycline MTX⫹ placebo
Tender joint count† ⫺9.0 ⫺7.7
Swollen joint count† ⫺8.9 ⫺8.1
ESR, mm/hour ⫺14.6 ⫺13.1
HAQ score ⫺0.47 ⫺0.28
Physician global assessment of disease activity (0–10 VAS)
⫺2.67 ⫺1.78
Patient global assessment of disease activity (0–10 VAS)
⫺2.18 ⫺1.16
Patient assessment of pain (0–10 VAS)
⫺1.68 ⫺0.85
Morning stiffness, minutes ⫺108 ⫺27
Fatigue (0–10 VAS) ⫺1.4 ⫺0.5
* MTX ⫽ methotrexate; ESR ⫽ erythrocyte sedimentation rate; HAQ ⫽ Health Assessment Questionnaire; VAS ⫽ visual analog scale.
1980s, periodontal research revealed that tetracyclines inhibited collagenase activity (18,19). Collagenase is an enzyme within a large family of MMPs that break down macromolecules in the connective tissue, leading to the pathologic changes of RA. The tetracyclines, including minocycline and doxycycline, have proven anti-MMP properties (including MMP-1, MMP-8, and MMP-11), which have been well documented in multiple tissues (for review, see refs. 20–23). Tetracycline has been shown to inhibit the production of nitric oxide, the level of which is increased in patients with RA (24). Inhibition of leukotaxis and phagocytosis is also seen in the setting of tetracycline therapy (25,26). Yu and colleagues re-vealed that doxycycline given preoperatively reduced collagenase activity in human osteoarthritic cartilage (27). In addition to these antiinflammatory properties, minocycline (but not doxycycline) has been shown to up-regulate interleukin-10, an antiinflammatory cyto-kine, which may explain its ability to cause drug-induced lupus (20,28).
The initial design of this trial involved 3 arms: placebo plus MTX, doxycycline 100 mg twice daily plus MTX, and an anti-MMP agent plus MTX. Prior to the start of the trial, that anti-MMP agent was pulled from testing, and doxycycline 20 mg twice daily was used for its anti-MMP effects, as documented in the periodontal literature (19). Our results showing apparently similar efficacy of the 2 different dosages of doxycycline suggest that the antimetalloproteinase effects of doxycycline rather than its antibacterial effects are critical for its efficacy, because doxycycline at a dosage of 20 mg twice daily does not alter microbial flora in the mouth or gut. A surprising result from our study was the low ACR response rates in the MTX plus placebo group. By comparison, the Early Rheumatoid Arthritis trial re-vealed an ACR20 response rate at 1 year of 65% in patients with seropositive or erosive disease (29). Our patient population was unique to our trial, in that all of our patients had early disease (mean duration 5 months), all were seropositive, and all were followed up for 2 years. In addition, we chose the ACR50 response rather than the ACR20 response as our primary out-come.
Two potential photosensitivity rashes were noted in the high-dose doxycycline group. Therefore, clinicians should be aware of the potential for a significant in-crease in the severity of photosensitivity reactions when MTX and doxycycline are combined.
In conclusion, our study demonstrated that in patients with early seropositive RA, treatment with doxycycline and MTX is superior (based on the ACR50
response) to treatment with MTX alone. We did not observe significant clinical differences between the 2 doxycycline-treated groups, suggesting that the antimet-alloproteinase effects of doxycycline are more important than its antibacterial effects. Additional trials are needed to corroborate these findings and further eluci-date the mechanism of action of tetracycline in RA.
ACKNOWLEDGMENTS
We acknowledge the contributions and dedication of the following RAIN investigators: Drs. Ana Fernandez, Anita Goel, Lynell Klassen, Eric Schned, and John Schousboe. We especially extend our gratitude to the patients with RA, without whom this research would not be possible.
REFERENCES
1. Boers M, Verhoeven AC, Markusse HM, van de Laar MA, Westhovens R, van Denderen JC, et al. Randomized comparison of combination step-down prednisolone, methotrexate and sul-phasalazine with sulsul-phasalazine alone in early rheumatoid arthri-tis. Lancet 1997;350:309–18.
2. Mottonen T, Hannonen P, Leirisalo-Repo M, Nissila M, Kauti-ainen H, Korpela M, et al, and the FIN-RACo trial group. Comparison of combination therapy with single-drug therapy in early rheumatoid arthritis: a randomized trial. Lancet 1999;353: 1568–73.
3. Calguneri M, Pay S, Caliskaner Z, Apras S, Kiraz S, Ertenli I, et al. Combination therapy versus monotherapy for the treatment of patients with rheumatoid arthritis. Clin Exp Rheumatol 1999;17: 699–704.
4. Mikuls TR, O’Dell J. The changing face of rheumatoid arthritis therapy: results of serial surveys [concise communication]. Arthri-tis Rheum 2000;43:464–5.
5. O’Dell JR, Haire CE, Palmer W, Drymalski W, Wees S, Blakely K, et al. Treatment of early rheumatoid arthritis with minocycline or placebo: results of a randomized, double-blind, placebo-controlled trial. Arthritis Rheum 1997;40:842–8.
6. O’Dell JR, Blakely KW, Mallek JA, Eckhoff PJ, Leff RD, Wees SJ, et al. Treatment of early seropositive rheumatoid arthritis: a two-year, double-blind comparison of minocycline and hydroxy-chloroquine. Arthritis Rheum 2001;44:2235–41.
7. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24.
8. Felson DT, Anderson JJ, Boers M, Bombardier C, Furst D, Goldsmith C, et al. American College of Rheumatology prelimi-nary definition of improvement in rheumatoid arthritis. Arthritis Rheum 1995;38:727–35.
9. Egger MJ, Huth DA, Ward JR, Reading JC, Williams HJ. Reduced joint count indices in the evaluation of rheumatoid arthritis. Arthritis Rheum 1985;28:613–9.
10. Kloppenburg M, Breedveld FC, Terwiel JP, Mallee C, Dijkmans BA. Minocycline in active rheumatoid arthritis: a double-blind, placebo-controlled trial. Arthritis Rheum 1994;37:629–36. 11. Tilley BC, Alarcon GS, Heyse SP, Trentham DE, Neuner R,
Kaplan DA, et al, and the MIRA Trial Group. Minocycline in rheumatoid arthritis: a 48-week, double-blind, placebo-controlled trial. Ann Intern Med 1995;122:81–9.
12. O’Dell JR, Paulsen G, Haire CE, Blakely K, Palmer W, Wees SJ, et al. Treatment of early seropositive rheumatoid arthritis with
minocycline: four-year followup of a double-blind, placebo-con-trolled trial. Arthritis Rheum 1999;42:1691–5.
13. Van der Laan W, Molenaar E, Ronday K, Verheijen J, Breedveld F, Greenwald R, et al. Lack of effect of doxycycline on disease activity and joint damage in patients with rheumatoid arthritis: a double-blind, placebo controlled trial. J Rheumatol 2001;28: 1967–74.
14. St.Clair EW, Wilkinson WE, Pisetsky DS, Sexton DJ, Drew R, Krause VB, et al. The effects of intravenous doxycycline therapy for rheumatoid arthritis: a randomized, double-blind, placebo-controlled trial. Arthritis Rheum 2001;44:1043–7.
15. Pillemer S, Gulko P, Ligier S, Yarboro C, Gourley M, Goldbach-Mansky R, et al. Pilot clinical trial of intravenous doxycycline versus placebo for rheumatoid arthritis. J Rheumatol 2003;30: 41–3.
16. Sreekanth VR, Handa R, Wali JP, Aggarwal P, Dwivedi SN. Doxycycline in the treatment of rheumatoid arthritis: a pilot study. J Assoc Physicians India 2000;48:804–7.
17. Stone M, Fortin PR, Pacheco-Tena C, Inman RD. Should tetra-cycline treatment be used more extensively for rheumatoid arthri-tis? Metaanalysis demonstrates clinical benefit with reduction in disease activity. J Rheumatol 2003;30:2112–22.
18. Golub LM, Lee HM, Lehrer G, Nemiroff A, McNamara TF, Kaplan R, et al. Minocycline reduces gingival collagenolytic activ-ity with diabetes: preliminary observations and a proposed new mechanism of action. J Periodontal Res 1983;18:516–26. 19. Golub LM, Ramamurthy NS, McNamara TF, Gomes B, Wolff M,
Casino A. Tetracyclines inhibit tissue collagenase activity: a new mechanism in the treatment of periodontal disease. J Periodontal Res 1984;19:651–5.
20. Ritchlin CT, Haas-Smith SA, Schwarz EM, Greenwald RA.
Mi-nocycline but not doxycycline upregulates IL-10 production in human synoviocytes, mononuclear cells and synovial explants [abstract]. Arthritis Rheum 2000;43 Suppl 9:S345.
21. Golub LM, Ramamurthy NS, McNamara TF, Greenwald RA, Rifkin BR. Tetracyclines inhibit connective tissue breakdown: new therapeutic implications for an old family of drugs. Crit Rev Oral Biol Med 1991;2:297–321.
22. Vernillo AT, Ramamurthy NS, Golub LM, Rifkin BR. The nonantimicrobial properties of tetracycline for the treatment of periodontal diseases. Curr Opin Periodontol 1994;2:111–8. 23. Greenwald RA. Tetracyclines may be therapeutically beneficial in
rheumatoid arthritis, but not for the reasons that you might think. J Clin Rheumatol 1995;1:185–9.
24. Amin AR, Attur MG, Thakker GD, Patel PD, Vyas PR, Patel RN, et al. A novel mechanism of action of tetracyclines: effects on nitric oxide synthases. Proc Natl Acad Sci U S A 1996;93:14014–9. 25. Martin RR, Warr GA, Couch RB, Yeager H, Knight V. Effects of
tetracycline on leukotaxis. J Infect Dis 1974;129:110–6.
26. Forsgen A, Schmeling D, Quie PG. Effect of tetracycline on the phagocytic function of human leukocytes. J Infect Dis 1974;130: 412–5.
27. Yu LP Jr, Smith GN Jr, Brandt KD, Capello WN. Preoperative oral administration of doxycycline reduces collagenase activity in extracts of human osteoarthritis cartilage [abstract]. Arthritis Rheum 1994;37 Suppl 9:S362.
28. Schlienger RG, Bircher AJ, Meier CR. Minocycline-induced lu-pus: a systematic review. Dermatology 2000;200:223–31. 29. Bathon JM, Martin RW, Fleischmann RM, Tesser JR, Schiff MH,
Keystone EC, et al. A comparison of etanercept and methotrexate in patients with early rheumatoid arthritis. N Engl J Med 2000; 343:1586–93.
