This study will compare the efficacy of intra-articular betamethasone injection to intra-articular ketorolac injection for symptomatic treatment of knee osteoarthritis in an equivalence study using a double-blinded, randomized, controlled design. Patients will be recruited from orthopaedic clinics at St. Luke's University Health Network. A total of 448 patients will be recruited (224 in each group) to receive an administration of an intra-articular betamethasone or ketorolac injection. The primary outcome is change in pain as assessed by the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score. Secondary outcomes include Knee Injury and Osteoarthritis Outcome (KOOS) and Knee Outcome Survey - Activities of Daily Living (KOS-ADL) scores and physical exam findings. Statistical analyses include repeated measures analysis of variance (ANOVA) (primary outcome) and selected ANOVA and nonparametric tests as deemed appropriate (secondary outcomes), with p < .05 denoting significance for all comparisons, and no adjustment for multiple testing.

Introduction and Background

Min et al. performed a double blind study comparing triamcinolone to ketorolac in a total of 32 patients with subacromial impingement syndrome and found improved outcomes in University of California Los Angeles shoulder rating scale, improved active abduction, forward flexion strength and patient satisfaction scores. Shortcomings of this study included short-term follow up (4 weeks), lack of control of other forms of treatment of subacromial impingement including physical therapy and injection without ultrasound which limits the ability to determine the accuracy of injections. Many of the patients also had a lack of an MRI, thus being unable to determine if the patients had an undiagnosed rotator cuff tear or other pathology that may have influenced outcomes.

Karthikeyan et al. performed a similar trial analyzing tenoxicam (another non-steroidal anti-inflammatory medication) to methylprednisolone in 58 patients.Six weeks after injection, Constant-Murley shoulder scores, Disabilities of the Arm, Shoulder and Hand scores were significantly improved in the steroid group compared to tenoxicam injection.

Oztuna et al. randomized thirty patients with grade II to III osteoarthritis with an effusion to aspiration and oral tenoxicam for 10 days or aspiration and intra-articular tenoxicam. The patients were followed at 2, 4 and 8 weeks and then at 3 month intervals. Patients with intra-articular tenoxicam had significantly fewer effusions at 1 year and a more rapid relief in joint pain.

Unlu et al analyzed 69 patients with knee arthritis and compared weekly injections of tenoxicam for 3 weeks to PO tenoxicam for three weeks. They compared these two groups to patients who received only physical therapy. No significant difference was found between oral and intra-articular administration. Both oral and intra-articular tenoxicam had a significant improvement on knee pain

Several studies have noted safe intra-articular injection of ketorolac following knee arthroscopy.

Given the findings of the investigations described above, this study will compare the equivalence of intra-articular ketorolac injection to intra-articular betamethasone injection for symptomatic treatment of knee osteoarthritis. In all of the studies reviewed above for the study, the only complication that was listed was a syncopal episode that occurred with corticosteroid administration.

Research Questions and Hypotheses

This study will compare the efficacy of intra-articular betamethasone injection to intra-articular ketorolac injection for symptomatic treatment of knee osteoarthritis. The null hypothesis of this study is that intra-articular ketorolac injection is not equivalent to intra-articular corticosteroid in decreasing patient symptoms as measured by Western Ontario and McMaster University Index (WOMAC) scores (primary outcome). Secondary outcomes include changes in KOOS and KOS-ADL scores and physical exam findings

If equivalence is proven, this study may provide a safer alternative medication compared to betamethasone for intra-articular injections for osteoarthritic pain control.

Methods and Procedures

Patients will be recruited from all St. Luke's Orthopaedic Specialists (SLOS) offices where knee osteoarthritis is treated in associated with St. Luke's Health and University Network.

The study will be double-blinded, with two study arms: 1) 4cc of 1% lidocaine, 1cc of 0.9% normal saline and 1cc (6 mg) of betamethasone; and 2) 2cc (60 mg) of ketorolac and 4cc of 1% lidocaine. The medication will be prepared by a site-specific designee (essential personnel) who has the appropriate, up-to-date HIPPA (Healkth insurance portability and Accountability Act) and CITI (Collaborative Institutional Training Initiative) training required and prepares injections in the routine scope of SLOS practice. The contents of the injection will be blinded to the investigators by covering the syringe with a white tape. The medication will be generated at the time of patient enrollment at the office where the injection will be performed.

Patients will be randomized by the following procedure. A spreadsheet will be generated with study patient identifier numbers in numerical order (1, 2, 3, 4, etc.). This spreadsheet will be located on a password-protected database that can be continuously updated across the St. Luke's Network by multiple study personnel at the same time. A random number generator (http://randomizer.org) will be used to generate a series of binary numbers, either 0 or 1. The number 0 will represent betamethasone, while the number 1 will represent ketorolac. This number will then be attached to the study patient identifier number.

When a patient is determined to be eligible for enrollment into the study, their medical record number will be entered into the password-protected spreadsheet database as the next patient by a site-specific designee. At that point the site-specific designee will provide the medication as described earlier to the injector. The injector and patient will remain unaware as to the contents of the syringe.

Prior to injection, patients will receive the following study questionnaires: WOMAC, Knee Injury and Osteoarthritis Outcome Score Physical Function Short Form (KOOS-PS), and Knee Outcome Survey Activities of Daily Living Scale (KOS-ADL). These study questionnaires were created with Google Forms and be stored on the same spreadsheet as described earlier. Patients will be instructed to return the completed questionnaires to the office during an office visit, by mail or fill them out online prior to study injection as well as 2 weeks, 1 month, 3 months and 6 months.

Patients' medical history will be taken and a clinical exam performed before the injection is given. Demographic information, including body mass index (BMI), duration of disease, and gender, will be obtained. Number and type of prior injections will also be documented if known by the patient or documented in the patient's chart. The clinical exam will include joint laxity, passive range of motion, presence of contractures, and tenderness on a scale of 0-3 (0 = no tenderness, 1 = patient winces and 3 = withdrawal). A three-view x-ray will also be obtained if there is no recent image on file. Other treatment modalities, including physical therapy, knee off-loading brace, acetaminophen, and current oral non-steroidal treatment, will be noted. Finally, Kellgren-Lawrence Staging will be performed for the medial, lateral and patellofemoral joints. This information will be documented in the password-protected Google Spread Sheet.

A standard superolateral injection site will be used during medication administration. This site is considered to be the most accurate, especially among experienced hands. The administration site will be cleaned with an ethanol swab and allowed to dry. An 18 gauge needle connected to a syringe filled with the study drug will be administered. If there is no effusion, a 22 gauge needle with 6cc of the study drug will be injected into the administration site. All of the study drug will be administered in an intra-articular fashion. The needle will be removed and a bandage applied. It will be documented if patient felt relief immediately following injection. If the patient has a history of diabetes, they will record morning fasting levels of glucose for 10 days following administration of study drug and then documented into the password-protected Google spreadsheet.

All patients will be provided with the option for a home or outpatient physical therapy regimen to follow for treatment of their arthritis. During follow up they will be asked about compliance with therapy and will be documented into the password-protected Google spreadsheet.

Clinical follow up will performed at 1, 3 and 6 months after injection. A repeat examination and questionnaire will be performed and documented at that time. The investigators will allow patients to receive a second corticosteroid administration on or after the three month visit as is current standard of care at St Luke's University Health Network.

If a recent MRI has been obtained within the year prior to or during treatment, this information will also be documented.

In knee osteoarthritis patients, common concerns with corticosteroid injections include skin color changes, systemic elevations in blood sugar, and may place the patient at a higher risk for infection in immunocompromised individuals. Typically injections of corticosteroids are limited and often not given more than once every three months. Ketorolac is a non-steroidal anti-inflammatory medication that has a more limited mechanism of action and may be a safer alternative to corticosteroid injections. Expected adverse effects of ketorolac include localized bruising, compensatory effusion, minor localized bleeding, and a flare reaction. Both of these medications are currently used by practitioners at St. Luke's University and Health Network for osteoarthritis. The goal of this study is to directly compare ketorolac's efficacy and safety to betamethasone

Statistics

Sample size calculation to assess equivalence was based on research by Unlu et al (2005) that compared WOMAC scores in patients with knee osteoarthritis who received intra-articular tenoxicam, oral tenoxicam or exercise only. Three was identified as the smallest clinically meaningful margin to determine equivalence of intra-articular ketorolac to intra-articular betamethasone as the active comparator. At α = 0.5 and β= 0.90, 204 patients per group (N=408) are required. To guard against missing or otherwise unusable data, 10% was added to that minimum sample size, for a total of 448 patients.

The primary outcome (change in WOMAC scores) will be analyzed via repeated measures analysis of variance (ANOVA), and the secondary outcomes will be analyzed via selected ANOVA and nonparametric tests as deemed appropriate. For all analyses, p < .05 will denote statistical significance, with no adjustment for the multiple comparisons.

Plan for the Dissemination of Project Results

This project will be submitted to the annual residents' and fellows' Research Symposium at the Bethlehem campus in the spring of 2016, as well as apply to present at appropriate national meetings. This project will also be submitted to orthopaedic and related journals.

This study will be registered with ClinicalTrials.Gov.

Personnel Qualifications

Primary personnel responsible for the compilation and publication of the results would include Gregory F Carolan, MD (principle investigator), Paul Morton, MD, Shane McGowan, MD and John Anagostakos. Personnel administering the injections and obtaining consent will be the physicians or physician assistants (listed in attached) of the St. Luke's Orthopaedic Specialists group who, in the scope of their current practice, evaluate and treat patients with knee osteoarthritis. All listed personnel will maintain appropriate CITI and HIPPA training.

Inclusion Criteria:

• Knee osteoarthritis as demonstrated on x-ray as Kellgren-Lawrence stage II or higher (Kellgren & Lawrence, 1957)
• Clinical symptoms and physical exam consistent with osteoarthritis

Exclusion Criteria:

• Prior treatment with corticosteroid injections within the last 1 year
• Failure of past treatment with corticosteroid injections
• Medical treatment including oral medication, physical therapy, or other analgesia use within the last 20 days prior to injection
• Pregnant
• Under 18
• Serious systemic disease
• Gout arthropathy
• Inflammatory joint disease including rheumatoid or psoriatic arthritis
• Metabolic bone disease
• Anserine bursitis or pain referred from other structures such as ipsilateral hip or lumbar spine
• Patients with a significant effusion that requires aspiration

Will plan to submit to orthopaedic journals for publication