Research Programs

Manipulation Under Anesthesia (MUA) to Treat Postoperative Stiffness after Total Knee Arthroplasty:  A Randomized Clinical Trial

Purpose:  To determine the efficacy of manipulation under anesthesia (MUA), with and without perioperative NSAIDS and IV dexamethasone, in terms of postoperative range of motion and Knee Society scores (KSS) in patients experiencing stiffness following primary total knee arthroplasty (TKA).

Background and Scientific Review:  Postoperative stiffness is a relatively common complication following TKA, which is responsible for substantial patient dissatisfaction.  It is among one of the most common reasons for early revision TKA and reoperations.  Manipulation under anesthesia is commonly used as the primary treatment modality in patients who have not achieved 90º of flexion within the first 4 – 8 weeks postoperatively.  Preliminary clinical and translational research has indicated that the inflammatory cascade plays a central role in the formation, and recurrence, of arthrofibrosis, particularly at the time of insult (i.e. time of index arthroplasty and/or MUA).  This is essential as some insurance companies have stopped covering this procedure.

Study Design:  Prospective, multicenter, randomized controlled trial (max of 10 centers)
Inclusion Criteria:  Any patient with flexion ≤ 90° between 4 and 12 weeks following a primary TKA

Exclusion Criteria: Allergy or contraindication to dexamethasone or celecoxib; diagnosis other than OA for the index TKA; age < 18 or > 90 years

Treatment Groups:

1. MUA + PT – Patient will receive MUA, followed by supervised PT for an additional 6 weeks
2. MUA + PT + 8mg IV dexamethasone at MUA + 2 weeks PO celecoxib 200mg QD

Sample Size:  The mean and standard deviation for sample size calculation were derived from a study by Yercan et al. In that study, the authors found that the mean ROM following MUA was 114° ± 16°.  We hypothesized that a 10° difference in total range of motion would constitute a clinically significant difference. With a type I error rate of 5%, 108 total patients (54 per arm) are required to have 90% power to detect this difference. Accounting for a drop our rate of 20%, 65 patients per arm (130 total patients) will be recruited.

Demographics/Patient Specifics:  Age at Index Arthroplasty, Sex, BMI, ASA score, Charlson Index

Randomization: Patients will be randomized using a computer generated (SAS PROC) block randomization schedule to ensure equivalent numbers of patients in each group over the course of the study

Outcome Measurements: 

1. Arc of motion and degrees of terminal knee flexion.
2. Knee Society Score (KSS).
3. SF-12, KOOS and PROMIS.