TMJ Arthroscopy Compared to Arthrocentesis in Disc Displacement Without Reduction

1. BACKGROUND The temporomandibular joint (TMJ) is a bilateral joint comprised of two cartilage covered bone surfaces that articulates against each other during mouth opening and closing. A dense cartilage disc is situated between the two joint surfaces. Disc displacement (DD) is characterised by an improper position of the TMJ disc relative to the articulating surfaces and affects up to 30% of the population. DD without reduction (DDwoR) is a sub-diagnosis of DD where the inaccurate disc position and associated inflammation hampers the TMJ movement and osteoarthritis successively develops. The affected patient suffers from reduced mouth opening capacity and pain when opening the mouth and during chewing, which leads to restrains in food intake, speech and social activities. A recent Swedish publication has shown that patients with TMJ disorders had more days of work disability (2-3 times longer) compared to a non-TMD cohort followed over a ten year period. The reliance on social security benefits in the group of patients that had TMJ surgery more than once were more accentuated compared to other TMJ disorder patients. The primary treatment for DDwoR is non-surgical, i.e., physiotherapy, pharmacological treatment and/or occlusal splint therapy. If non-surgical treatment fails, arthroscopic lysis and lavage (ALL) or arthrocentesis (AC) might be considered. ALL is performed with a rod-shaped lens inserted into the TMJ cavity enabling direct visualisation, The operative procedure involves manipulation of disc position, release of adhesions (lysis), and irrigation (lavage) of the joint with either saline solution or Ringer's solution. Generally the procedure is performed under general anaesthesia, but there are reports on ALL performed in local anaesthesia or conscious sedation. Success rates of 75-88 % is reported. In AC, two needles are inserted into the joint cavity to irrigate the TMJ. The method does not allow visualisation of the joint. AC is often performed together with opening and closing manoeuvres of the mandible to enable manipulation of the discs position. AC is reported to be executed under general anaesthesia, but also under local anaesthesia or conscious sedation. AC is most often reported to have a higher success rate compared to ALL, with figures ranging from 66-100 %. Although, the compared figures are taken from cohort studies with no control group. There are a few randomised and controlled studies comparing ALL to AC. Fridrich et al. performed a randomised study without blinding and also incorporates two different diagnoses, DDwoR and disc displacement with reduction (DDwR). The patient sample was small, 20 patients. Murakami et al. conducted a study including three interventions: non-surgical , ALL and AC. The study was not blinded and not randomised, and 25 patients were included in the ALL-group and 20 patients in the AC group. The patients had the diagnosis DDwoR. In a third study, a total of 63 patients were treated, 33 had arthroscopy and 29 had arthrocentesis. The study was not blinded and 8 of the included patients was diagnosed with DDwR and the remaining with DDwoR. All three studies showed a better outcome for patients treated with arthroscopy, even though the result was not significant in two of the studies. In a meta-analysis published 2015 both MIO and patient-assessed TMJ pain had significantly better outcome in arthroscopy. Synovial fluid from the TMJ has been used since the mid-90's mostly for analysing inflammatory markers. Proteomics is a relatively new technique for analysing protein profiles and has only been applied a couple of times in the field of TMJ pathology. Since the aetiology of DD is still not known, proteomics provides a comprehensive insight of the protein profile of TMJ disorders. There are very few RCT's comparing the outcome of ALL and AC in patients with DDwoR and to this date none with a blinding. The aim of this study is to evaluate outcome of ALL and AC in treating DDwoR in a double-blind fashion. Another aim is to investigate synovial fluid protein profile in TMJ DD and to see if patient-specific differences might be connected to surgical outcome, local inflammation, local pain, or level of degenerative changes.
2. HYPOTHESIS/RESEARCH QUESTIONS ALL has a superior outcome compared to AC treating patients with DDwoR. The outcome will be longitudinally evaluated by maximum mouth opening capacity, patient reported TMJ pain and TMJ disability, and Quality of life (QoL) questionnaires at pre-determined time-points during a 24-month postoperative period. The primary research question is which surgical method that best improves the above stated variables. A secondary question is if there is an identifiable protein profile regarding DDwoR. A tertiary question is if any pre- or peri-operative variable/-s could be identified as a predictor for outcome. A quaternary is the impact from DDwoR on QoL and the surgical outcomes impact on QoL. A last question is if none of the therapies are better than the other, do the recovery time after surgery differ between ALL and AC.
3. METHODS 3.1. Study design

Randomised, double-blind, controlled study where the outcome of ALL and AC on patients with DDwoR is compared.

3.2. Data source

All patients referred to the Department of Oral- and Maxillofacial Surgery, Karolinska University Hospital, Stockholm, with the diagnosis DDwoR and with a history of not responding to non-surgical treatment will be screened for inclusion.

3.3. Exposure/Intervention

Patients will be assigned to either ALL or AC. Both treatments will be performed under general anaesthesia. The randomisation will be made in blocks of two in a 1:1 ratio. A computer software provided by KTA (Karolinska Trial Alliance) will handle the randomisation process. At time of surgery, when the patient is under general anaesthesia and draped in sterile cloth, the patient will be allocated to either ALL or AC. In this manner the intervention or placebo treatment will be blinded to the patient but not to the surgeon. The person evaluating the outcome of surgery (the assessor) will not know which treatment the patient has undertaken. The two surgical procedures are described below.

ALL: The surgical procedure is performed according to the original description of the technique. Local During the operation the joint space is irrigated with a minimum of 300 mL saline solution. A pressure infuser (InfuseITTM) set at 175 mmHg will be used to standardise the irrigation pressure. At the start of joint irrigation, the first 50 mL outflow irrigation fluid will be collected in a sterile container for later proteomic analysis. At the end of surgery, the joint is flushed with 10 mL of Marcaine-Adrenaline (2.5 mg/mL). No other medication will be injected or otherwise introduced into the joint compartment. The incision is sutured with a single non-resorbable suture, and a wound dressing is put over the incision. The described intervention is standardized and takes no more than 30 minutes.

AC: Arthrocentesis of the TMJ was first described in 1991 and the same technique is used here with the exception that no corticosteroids are injected in the joint postoperatively. During the operation the joint space is irrigated with a minimum of 300 mL saline solution. A pressure infuser (InfuseITTM) set at 175 mmHg will be used to standardise the irrigation pressure. At the start of joint irrigation, the first 50 mL outflow irrigation fluid will be collected in a sterile container for later proteomic analysis. After finishing the mandibular manipulation and the joint irrigation, the joint is flushed with 10 mL of Marcaine-Adrenaline (2.5 mg/mL). No other medication will be injected or otherwise introduced into the joint compartment. Finally, a vertical preauricular skin incision (approximately 5 mm long) is made to mimic an arthroscopic procedure. The incision is sutured with a single non -resorbable suture and a wound dressing is put over the incision.

To keep the blinding intact for the patient and the assessor, the medical chart will be written as if all the patients had AC performed.

Both the intervention groups will have postoperative physiotherapy training for a period of one month from the day after surgery in accordance with a specified home exercise program (see attachment). If needed, the home exercise program can be prolonged or individually adjusted after the first month. If the patient uses an occlusal splint prior to surgery the splint may have to be adjusted and that has to be performed by the assessor.

3.4. Outcome

Please see "Outcome measure" below.

3.5. Sample size

50 patients in each intervention group, i.e., 100 patients in total, in accordance with the power analysis, see "Statistical analysis".

3.6. Inclusion and exclusion criteria

Please, see "Eligibility" below.

3.7. Data collection

At registration for surgery and after obtained informed consent, baseline data is collected by the surgeon designating the patient to surgery. Predetermined follow-up visits are at 1 week when the surgeon will remove the stich and do a reinstruction for the home-exercise program, and at 1, 3, 6, 12 and 24 months postoperative when the assessor examine the patient according to the specific exam protocol (see attachment) The assessor/-s will be calibrated alongside the surgeon/-s before the start of the study in accordance with the DC/TMD criteria. At 6-month intervals the surgeons and assessors will have a meeting for re-calibrating purposes.

Objective data only registered preoperative: name, personal identification number, age, sex, affected joint (left, right).

Objective data registered both pre- and post-operatively: MIO with and without pain (mm), lateral excursive movement (mm), protrusion (mm), pain upon palpating masticatory muscles, pain upon lateral joint palpation.

Subjective data registered both pre- and post-operatively: patient-reported TMJ disability (NRS 0-10), patient-reported TMJ pain (NRS 0-10), patient-reported psycho-social impairment because of TMJ disorder (NRS 0-10), patient-reported global pain (NRS 0-10), and quality of life questionnaire EQ-5D-5L, JFSL-8, and OHIP 14-S.

Perioperative data: antibiotic prophylaxis (yes/no), performed surgery (AC/ALL), maximum assisted mouth opening (mm), duration of surgery (min), duration of general anaesthesia (min), lavage volume (mL), preauricular swelling due to lavage (yes/no),Wilkes grading (3-5), negative event during and due to surgery. Grading of intraarticular conditions will be made according to the scale proposed by Gynther et al.: synovitis (0-3), degenerative changes (0-3), and fibrosis/adhesions (0-2), in case of ALL.

3.8. Revelation of blinding At the 6-month visit the patients will be asked if they believe that they had ALL or AC. The assessor will also give their opinion in that matter. After the 6-month follow-up, the blinding will be revealed to the patient if they want to. Patients with a non-successful outcome might be offered another therapy; non-surgical or surgical. Controls will be made at 12- and 24-months post-operative regarding all patients.

3.9. Analysis of synovial fluid Proteomic analysis and protein quantification in synovial fluid will be performed at PainOmics laboratory, University hospital in Linköping. In this project, we will follow a discover and verification/validation schema. In the first step peptides and protein contents of the synovial fluid will be analysed using 2-DE in combination with mass spectrometry. With this technique it is possible to separate and quantify more than 1000 proteins and it is the only technique that allows separating and studying different protein isoforms. Quantification of the separated proteins is achieved by spot integration using specialized 2-DE software that makes it possible to compare protein patterns from different subjects/groups. Identification of the protein is performed by in-gel digestion of each protein, recovery of the fragments from the gel, and analysis by mass spectrometry (MS) using MALDI-T

1. STATISTICAL ANALYSIS The primary objective is to evaluate the superiority of ALL compared to AC in patients diagnosed with DDwoR and not responding to non-surgical treatment. The outcome variable is dichotomous (successful/not successful) where a successful outcome in the ALL group is assessed to be 80 % and in the AC group 50 %. With 80 % power and a 5 % significance level the groups must contain 45 patients each. Dropout rate of 10 % is accounted for which means that 50 patients/group must be included to verify superiority.

Statistical analyses will be performed using the Chi-Square test for bivariate data and the Students t-test (unpaired and paired) for mean values, as well as repeated measures ANOVA. Longitudinal data will be analysed with linear regression analysis and predictors for surgical outcome will be evaluated through logistic regression. The intention to treat analysis model will be used.

Potential known confounders may be adjusted for in the analysis and the risk of residual unknown confounders should be reduced by the randomisation process.