Insulin Resistance, Rheumatoid Arthritis
Conditions
Keywords
rheumatoid arthritis, insulin resistance, hydroxychloroquine, cholesterol
Brief summary
The purpose of this study is to determine whether hydroxychloroquine (HCQ) reduces insulin resistance in non-diabetic subjects with rheumatoid arthritis (RA). The investigators will conduct a double-blind randomized crossover trial in subjects with RA to test the hypothesis that HCQ improves insulin sensitivity. The investigators will also use data from the trial to identify determinants of insulin resistance in RA. The investigators hypothesize that RA will be associated with an increased risk of insulin resistance and that independent risk factors for increased insulin resistance in RA include higher BMI, elevated acute phase reactants, greater fat to muscle ratio, and less physical activity.
Detailed description
Our ability to better control the pain and disability of rheumatoid arthritis (RA) now focuses attention on reducing the impact of RA-associated comorbidities. The most common cause of death in RA is cardiovascular (CV) disease, and the risk of myocardial infarction and stroke are approximately doubled in RA. The determinants of CV risk in RA include traditional CV risk factors as well as aspects of the inflammatory process defining RA. It is likely that RA-associated inflammation accelerates atherosclerosis through direct effects on the endothelium as well as indirect effects on insulin metabolism. Several studies report an increased prevalence of insulin resistance among persons with RA. However, it is not clear whether the inflammation of RA causes insulin resistance. Corticosteroids and abnormalities in the hypothalamic-pituitary axis may also contribute to abnormal glucose metabolism. Little information is available to guide management of a pre-diabetic insulin resistance state in RA. Hydroxychloroquine (HCQ), a commonly used medicine early in RA, may play a role in improving insulin resistance. Several previous trials demonstrated the ability of HCQ to reduce blood glucose levels in diabetics, and a large epidemiologic study found that subjects with RA using HCQ were less likely to develop diabetes. In animal models, anti-malarials lower blood glucose through slowing insulin metabolism. With CV disease a major comorbidity in RA and insulin resistance possibly a major determinant of CV risk, intervention studies need to begin to translate prior work into clinical therapeutics. Relevance: If this study demonstrates a beneficial effect of HCQ on insulin resistance among the randomized subjects, this would provide strong evidence that HCQ has benefits beyond RA and SLE disease activity. Currently, HCQ is stopped in many patients as they step-up to more aggressive DMARD treatments, or HCQ may never be tried in some patients who present with RA carrying with poor prognosis. If HCQ improves insulin sensitivity, there may be rationale for continuing HCQ chronically in patients with RA. As well, a larger clinical endpoint study would be strongly considered.
Interventions
DRUGHydroxychloroquine
Hydroxychloroquine comes in 200 mg tablets and is taken orally. The dose provided will be based upon a calculation of 6.5 mg/kg (subject's weight), which is the dose range commonly used to treat rheumatoid arthritis and lupus. Dosages will be rounded to the nearest 100 mg.
Sponsors
National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS)
Brigham and Women's Hospital
Study design
Allocation
RANDOMIZED
Intervention model
CROSSOVER
Primary purpose
TREATMENT
Masking
QUADRUPLE (Subject, Caregiver, Investigator, Outcomes Assessor)
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Age 18 or older * Able to provide informed consent and comply with study visits * Hemoglobin ≥ 10 g/dL (within last two months) * WBC ≥ 4 K/uL (within last two months) * Platelet count ≥ 150 ≤ 450 K/uL (within last two months) * (GFR) Creatinine clearance ≥ 70 ml/min (MDRD) (within last two months) * SGOT, SGPT ≤ 1.5 times upper limits of normal (within last two months) * Normal eye exam within 12 months of study entry (copy of letter from subject's ophthalmologist or optometrist stating that the subject has no evidence of macular pathology) * Diagnosis of rheumatoid arthritis
Exclusion criteria
* History of any neuromuscular disease including muscular dystrophy, metabolic myopathies, peripheral neuropathy, multiple sclerosis, and other myopathies or myositides * History of diabetes or fasting plasma glucose of 126 mg/dl or greater * History of any untoward reaction to antimalarials * Uncontrolled hypertension (\>140/90) * History of any ophthalmologic disease except for glaucoma or cataracts * Planned elective surgery during the study period * Digoxin therapy * Treatment with corticosteroids (\> 5 mg) for any disorder * History of psoriasis * Any chronic disease that in the opinion of the investigator warrants exclusion (e.g. inflammatory bowel disease, malignancy other than basal cell carcinoma, chronic liver disease) * History of chronic intestinal disorders (Crohn's disease, ulcerative colitis, celiac sprue, collagenous colitis, eosinophilic enteritis) * Creatinine clearance ≤ 60 ml/min (MDRD) (within last two months) * Hemoglobin ≤ 10 g/dL (within last two months) * WBC ≤ 4 K/uL (within last two months) * Platelet count ≤ 150 ≥ 450 K/uL (within last two months) * SGOT, SGPT ≥ 1.5 times upper limits of normal (within last two months) * Women who are pregnant or breastfeeding
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Insulin Sensitivity Index | Baseline and Week 8 | We will examine the effect of HCQ on the Matsuda Insulin Sensitivity Index (ISI) during the active treatment phase compared with placebo phase. ISI is based on insulin and glucose levels in a fasting state during an oral glucose tolerance test (OGTT) and is calculated as follows: ISI (Matsuda) = 10000/√ G0 X I0 X Gmean X Imean G0 - fasting plasma glucose (mg/dL) I0 - fasting plasma insulin (mIU/L) Gmean - mean plasma glucose during OGTT (mg/dL) Imean - mean plasma insulin during OGTT (mIU/L) |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| HOMA-B | Baseline and Week 8 | HOMA-B = (360 x Insulin)/(Glucose - 63) |
| Total Cholesterol | Baseline and Week 8 | mg/dL |
| HOMA-IR | Baseline and Week 8 | We will examine the effect of HCQ on HOMA-IR during the active treatment phase compared with placebo phase. HOMA-IR = (Glucose x insulin)/405 |
| HDL Cholesterol | Baseline and Week 8 | mg/dL |
| Triglycerides | Baseline and Week 8 | mg/dL |
| LDL Cholesterol | Baseline and Week 8 | mg/dL |
Countries
United States
Participant flow
Recruitment details
107 potentially eligible RA patients were screened during 2010-2011. 7 subjects did not meet inclusion/exclusion criteria at screening and 63 declined participation.
Pre-assignment details
37 subjects were initially consented, 7 withdrew consent before randomization. Thirty patients were randomized. Five patients withdrew consent after randomization and two patients stopped participation in the study due to non-serious adverse events.
Participants by arm
| Arm | Count |
|---|---|
| HCQ Then Placebo This arm of the study contains half the study population after randomization. The participants in this arm received hydroxychloroquine for 8 weeks and then crossed over to a placebo for 8 weeks. Study staff was blinded to which order they were taking the hydroxychloroquine and placebo in.
Hydroxychloroquine: Hydroxychloroquine comes in 200 mg tablets and is taken orally. The dose provided will be based upon a calculation of 6.5 mg/kg (subject's weight), which is the dose range commonly used to treat rheumatoid arthritis and lupus. Dosages will be rounded to the nearest 100 mg. | 15 |
| Placebo Then HCQ This arm of the study contains half the study population after randomization. The participants in this arm received hydroxychloroquine for 8 weeks and then crossed over to a placebo for 8 weeks. Study staff was blinded to which order they were taking the hydroxychloroquine and placebo in.
Hydroxychloroquine: Hydroxychloroquine comes in 200 mg tablets and is taken orally. The dose provided will be based upon a calculation of 6.5 mg/kg (subject's weight), which is the dose range commonly used to treat rheumatoid arthritis and lupus. Dosages will be rounded to the nearest 100 mg. | 15 |
| Total | 30 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| First Intervention (8 Weeks) | Adverse Event | 1 | 0 |
| First Intervention (8 Weeks) | Withdrawal by Subject | 2 | 3 |
| Second Intervention (8 Weeks) | Adverse Event | 0 | 1 |
Baseline characteristics
| Characteristic | HCQ Then Placebo | Placebo Then HCQ | Total |
|---|---|---|---|
| Age, Continuous | 56 years STANDARD_DEVIATION 14 | 57 years STANDARD_DEVIATION 14 | 56.5 years STANDARD_DEVIATION 11 |
| Sex: Female, Male Female | 13 Participants | 14 Participants | 27 Participants |
| Sex: Female, Male Male | 2 Participants | 1 Participants | 3 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | — / — | — / — |
| other Total, other adverse events | 2 / 15 | 0 / 15 |
| serious Total, serious adverse events | 0 / 15 | 0 / 15 |
Outcome results
Primary
Insulin Sensitivity Index
We will examine the effect of HCQ on the Matsuda Insulin Sensitivity Index (ISI) during the active treatment phase compared with placebo phase. ISI is based on insulin and glucose levels in a fasting state during an oral glucose tolerance test (OGTT) and is calculated as follows: ISI (Matsuda) = 10000/√ G0 X I0 X Gmean X Imean G0 - fasting plasma glucose (mg/dL) I0 - fasting plasma insulin (mIU/L) Gmean - mean plasma glucose during OGTT (mg/dL) Imean - mean plasma insulin during OGTT (mIU/L)
Time frame: Baseline and Week 8
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Hydroxychloroquine | Insulin Sensitivity Index | Baseline | 7.7 (dL x L)/(mg x mIU) | Standard Deviation 4.8 |
| Hydroxychloroquine | Insulin Sensitivity Index | Period Value | 8.1 (dL x L)/(mg x mIU) | Standard Deviation 4.5 |
| Hydroxychloroquine | Insulin Sensitivity Index | Change | 0.4 (dL x L)/(mg x mIU) | Standard Deviation 2.9 |
| Placebo | Insulin Sensitivity Index | Baseline | 7.7 (dL x L)/(mg x mIU) | Standard Deviation 4.8 |
| Placebo | Insulin Sensitivity Index | Period Value | 7.8 (dL x L)/(mg x mIU) | Standard Deviation 4 |
| Placebo | Insulin Sensitivity Index | Change | 0.14 (dL x L)/(mg x mIU) | Standard Deviation 3.1 |
Comparison: The P-value calculated was for the difference betweeen the change during hydroxychloroquine versus placebo using Wilcoxon signed-rank tests.p-value: 0.93Wilcoxon (Mann-Whitney)
Comparison: The P-value was calculated for the difference between the change during hydroxychloroquine versus placebo from a linear regression model.p-value: 0.785Regression, Linear
Secondary
HDL Cholesterol
mg/dL
Time frame: Baseline and Week 8
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Hydroxychloroquine | HDL Cholesterol | Baseline | 58.1 mg/dL | Standard Deviation 17.9 |
| Hydroxychloroquine | HDL Cholesterol | Period | 59.4 mg/dL | Standard Deviation 18.2 |
| Hydroxychloroquine | HDL Cholesterol | Change | 1.3 mg/dL | Standard Deviation 6.7 |
| Placebo | HDL Cholesterol | Baseline | 58.1 mg/dL | Standard Deviation 17.9 |
| Placebo | HDL Cholesterol | Period | 60.3 mg/dL | Standard Deviation 17.8 |
| Placebo | HDL Cholesterol | Change | 2.2 mg/dL | Standard Deviation 9.8 |
Comparison: P-value for the difference between the change during hydroxychloroquine versus placebo using Wilcoxon signed-rank tests.p-value: 0.73Wilcoxon (Mann-Whitney)
Comparison: P-value for the difference between the change during hydroxychloroquine versus placebo from a linear regression model.p-value: 0.208Regression, Linear
Secondary
HOMA-B
HOMA-B = (360 x Insulin)/(Glucose - 63)
Time frame: Baseline and Week 8
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Hydroxychloroquine | HOMA-B | Baseline | 116.5 (mIU x dL)/(L x mg) | Standard Deviation 100.4 |
| Hydroxychloroquine | HOMA-B | Change | -5.8 (mIU x dL)/(L x mg) | Standard Deviation 72.9 |
| Hydroxychloroquine | HOMA-B | Period | 110.8 (mIU x dL)/(L x mg) | Standard Deviation 81.6 |
| Placebo | HOMA-B | Baseline | 116.5 (mIU x dL)/(L x mg) | Standard Deviation 100.4 |
| Placebo | HOMA-B | Change | -6.8 (mIU x dL)/(L x mg) | Standard Deviation 78 |
| Placebo | HOMA-B | Period | 109.7 (mIU x dL)/(L x mg) | Standard Deviation 82.6 |
Comparison: P-value for the difference betweeen change during hydroxychloroquine versus placebo suing Wilcoxon signed-rank tests.p-value: 0.468Wilcoxon (Mann-Whitney)
Comparison: P-value for the difference between the change during hydroxychloroquine versus placebo from a linear regression model.p-value: 0.902Regression, Linear
Secondary
HOMA-IR
We will examine the effect of HCQ on HOMA-IR during the active treatment phase compared with placebo phase. HOMA-IR = (Glucose x insulin)/405
Time frame: Baseline and Week 8
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Hydroxychloroquine | HOMA-IR | Baseline | 2.0 (mg x mIU)/(dL*L) | Standard Deviation 1.7 |
| Hydroxychloroquine | HOMA-IR | Period | 1.7 (mg x mIU)/(dL*L) | Standard Deviation 1 |
| Hydroxychloroquine | HOMA-IR | Change | -0.3 (mg x mIU)/(dL*L) | Standard Deviation 1.5 |
| Placebo | HOMA-IR | Baseline | 2.0 (mg x mIU)/(dL*L) | Standard Deviation 1.7 |
| Placebo | HOMA-IR | Period | 1.6 (mg x mIU)/(dL*L) | Standard Deviation 0.79 |
| Placebo | HOMA-IR | Change | -0.42 (mg x mIU)/(dL*L) | Standard Deviation 1.4 |
Comparison: P-value for the difference between change during hydroxychloroquine versus placebo using Wilcoxon signed-rank tests.p-value: 0.575Wilcoxon (Mann-Whitney)
Comparison: P-value for the difference between the change during hydroxychloroquine versus placebo from a linear regression model.p-value: 0.308Regression, Linear
Secondary
LDL Cholesterol
mg/dL
Time frame: Baseline and Week 8
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Hydroxychloroquine | LDL Cholesterol | Baseline | 114.1 mg/dL | Standard Deviation 29.8 |
| Hydroxychloroquine | LDL Cholesterol | Period | 101.7 mg/dL | Standard Deviation 36.7 |
| Hydroxychloroquine | LDL Cholesterol | Change | -12.4 mg/dL | Standard Deviation 20.1 |
| Placebo | LDL Cholesterol | Baseline | 114.1 mg/dL | Standard Deviation 29.8 |
| Placebo | LDL Cholesterol | Period | 109.9 mg/dL | Standard Deviation 33.1 |
| Placebo | LDL Cholesterol | Change | -4.2 mg/dL | Standard Deviation 17.7 |
Comparison: P-value for the difference between the change during hydroxychloroquine versus placebo using Wilcoxon signed-rank tests.p-value: 0.009Wilcoxon (Mann-Whitney)
Comparison: P-value for the difference between the change during hydroxychloroquine versus placebo from a linear regression model.p-value: 0.011Regression, Linear
Secondary
Total Cholesterol
mg/dL
Time frame: Baseline and Week 8
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Hydroxychloroquine | Total Cholesterol | Baseline | 192.4 mg/dL | Standard Deviation 37.5 |
| Hydroxychloroquine | Total Cholesterol | Period | 179.7 mg/dL | Standard Deviation 44.3 |
| Hydroxychloroquine | Total Cholesterol | Change | -12.7 mg/dL | Standard Deviation 20 |
| Placebo | Total Cholesterol | Baseline | 192.4 mg/dL | Standard Deviation 37.5 |
| Placebo | Total Cholesterol | Period | 189.4 mg/dL | Standard Deviation 37.9 |
| Placebo | Total Cholesterol | Change | -3.0 mg/dL | Standard Deviation 22.4 |
Comparison: P-value for the difference between the change during hydroxychloroquine versus placebo using Wilcoxon signed-rank tests.p-value: 0.004Wilcoxon (Mann-Whitney)
Comparison: P-value for the difference between the change during hydroxychloroquine versus placebo from a linear regression model.p-value: 0.004Regression, Linear
Secondary
Triglycerides
mg/dL
Time frame: Baseline and Week 8
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Hydroxychloroquine | Triglycerides | Period | 92.4 mg/dL | Standard Deviation 41.7 |
| Hydroxychloroquine | Triglycerides | Baseline | 100.6 mg/dL | Standard Deviation 44 |
| Hydroxychloroquine | Triglycerides | Change | -8.2 mg/dL | Standard Deviation 45 |
| Placebo | Triglycerides | Baseline | 100.6 mg/dL | Standard Deviation 44 |
| Placebo | Triglycerides | Period | 95.6 mg/dL | Standard Deviation 44.9 |
| Placebo | Triglycerides | Change | -5.0 mg/dL | Standard Deviation 20.1 |
Comparison: P-value for the difference between the change during hydroxychloroquine versus placebo using Wilcoxon signed-rank tests.p-value: 0.487Wilcoxon (Mann-Whitney)
Comparison: P-value for the difference between the change during hydroxycholorquine versus placebo from a linear regression model.p-value: 0.884Regression, Linear