Over 100 million Americans report chronic pain. Veterans are disproportionately affected for multiple reasons, including injuries and post-traumatic stress disorder. Treatment for chronic pain is a priority research area for the VA. One of the most common causes of chronic pain is osteoarthritis (OA). OA is attributable to "wear and tear," but reasons for pain are complex. Inflammatory arthritis (IA) includes multiple severe diseases that affect 2-3% of persons and require treatment with immune-suppressive drugs to prevent joint destruction. Pain often persists despite effective treatment. Pain in arthritis results from multiple sources: inflammation, perception of pain in the joint, and interpretation of pain by the brain. Unfortunately, management of pain in arthritis remains a challenge. Low dose naltrexone is a widely used but unproven "alternative" approach to chronic pain. It is attractive for study because it is safe and is proposed to work on all three pathways that contribute to pain. A small but high-quality clinical trial is needed to determine whether to invest in definitive studies.

Chronic pain affects over 100 million Americans, and arthritis is the most common cause. Existing treatments for chronic arthritic pain are only mildly effective, and risks of medications used to treat pain are numerous and continue to be discovered. Treatment of chronic is a high priority research area for VA CSR&D.

Naltrexone is an opioid antagonist that is FDA approved in an oral daily dose of 50 mg to prevent recidivism in alcoholics. At much lower doses of 4 - 4.5 mg daily, however, it has been shown in small, blinded, randomized trials to improve pain in fibromyalgia, gastrointestinal symptoms in Crohn's disease, and quality of life in multiple sclerosis. The only other published data are case reports in complex regional pain syndrome, low back pain, and scleroderma. However, advocacy of low-dose naltrexone (LDN) by internet-based MDs and patients is high, and since LDN can be prescribed off-label, its use greatly exceeds what is justified by evidence. The drug can be prescribed only via compounding pharmacies, so its use costs a patient ~$40/month.

Among the many unproven treatments that are widely used, LDN is of particular interest because results of surveys of patients are particularly impressive, because it is quite safe, and because its benefit is plausible pharmacologically. There is evidence both for modulation of central pain-processing pathways and for down-regulation of inflammatory pathways in microglia. Considering the diversity of conditions proposed to benefit from LDN and the unequivocal need for better approaches to pain relief in chronic conditions, high-quality clinical trials are needed in both inflammatory and non-inflammatory conditions. This small but placebo-controlled study, powered to detect an effect size as small as that seen with NSAIDs or the most beneficial non-pharmacologic approaches, is proposed as a prerequisite for considering a pivotal trial through the VA Cooperative Studies Program.

The proposed study is a randomized, double-blinded, cross-over, placebo-controlled trial in adults with osteoarthritis or inflammatory arthritis and persistent pain. Sixty patients will be enrolled for 16 weeks, during which they will receive LDN for 8 weeks and placebo for 8 weeks. Widely accepted patient-reported outcome measures will be used. The co-primary endpoints are reduction in pain severity or pain's interference with function during 8 weeks of LDN compared to 8 weeks placebo, using the Brief Pain Inventory. Other patient-reported data will be used both as secondary outcomes and as covariates in analyzing determinants of response to treatment.

Inclusion Criteria:

Patients must meet all of the following criteria in order to be eligible for enrollment:

• Veteran or otherwise eligible for VA benefits, able to travel to VA Boston

• One or more of the following chronic conditions:

  • osteoarthritis
  • rheumatoid arthritis
  • non-axial spondyloarthritis

• Average daily pain interference with function (average of the 7 parts of question 9 on the Brief Pain Inventory) rated at least 4 on a scale of 0-10, and no higher than 9

• No change in medication in the past 8 weeks made with the expectation of improving pain

• No plan to start another medication or a non-pharmacologic treatment regimen likely to affect pain during the next 16 weeks

• Age at least 18

• Registered for medical care in the VA Boston Healthcare System

• Capable of informed consent, and willingness to comply with study procedures, including receipt of weekly phone calls from the study coordinator

Exclusion Criteria:

Any of the following requires exclusion from participation:

• Current use of opioids including tramadol
• Pregnant, breast feeding, or unwilling to engage in contraceptive practices if sexually active and capable of conceiving
• Schizophrenia, bipolar disorder, or poorly controlled depression or anxiety
• Previous use of low-dose naltrexone
• Back pain described by the patient as greater in severity than arthritic pain in a non-axial location
• Significant kidney disease, defined as glomerular filtration rate < 30 ml/min
• Liver cirrhosis. There is no specific screening procedure to exclude cirrhosis.
• Painful peripheral neuropathy. There is no specific screening procedure.
• Plan to have surgery during the next 16 weeks
• Inconsistency in self-reporting at the screening visit. BPI, PainDETECT, WOMAC, and PROMIS-29 all contain 0-10 scales of average pain intensity, although the times listed vary from 1-4 weeks. The severity reported on these three scales cannot differ by more than 1.
• Other qualitative circumstances that the investigator feels would make the patient a poor candidate for this clinical trial, such as an unstable social situation or unreliable transportation

A de-identified, anonymized dataset containing all the primary study data will be created and shared per VA policies. This dataset will be included as a supplementary file attached to the published manuscript, which in turn will be available through PubMed Central per VA rules. In the event that the study has not been published, the dataset will be made available by other means within 3 years of study completion. The publicly available dataset will not include any identifiers, e.g. age will be included but not birthdate.

ICF to be posted within 3 months per VA policy. De-identified dataset to be included as supplementary material with publication, anticipated to be within 12 months of study completion.

Patients who completed the study

16 weeks (8 weeks each naltrexone and placebo)