The hierarchy of different treatments for arthrogenous temporomandibular disorders: A network meta-analysis of randomized clinical trials

Different treatment options for patients with arthrogenous Temporomandibular Disorders (TMDs) have been reported. However, evidence regarding the most effective intervention using network meta-analysis (NMA) has not been performed. Thus, we conducted a NMA of randomized clinical trials (RCTs) to identify the most effective treatment of arthrogenous TMDs with respect to pain reduction and improved mouth opening, and to generate a ranking according to their effectiveness. An electronic search on three major databases was undertaken to identify RCTs published before August 2019, comparing up to fourteen different treatments against control/placebo patients for arthrogenous TMDs with respect to pain reduction and improved mouth opening. The treatment variables were controls/placebo, conservative treatment (muscle exercises and occlusal splint therapy), occlusal splint therapy alone, intraarticular injection (IAI) of hyaluronic acid (HA) or corticosteroid (CS), arthrocentesis with or without HA, CS and platelet-rich plasma (PRP), arthroscopy with or without HA and PRP, open joint surgery, and physiotherapy. Frequentist NMA was performed using STATA software. Studies meeting the inclusion criteria were divided according to the length of follow-up (short-term (≤5 months) and intermediate-term (≥6 months to 4 years) and type of TMJ arthrogenous disorders; internal derangement (ID) and TMJ osteoarthritis (OA). The standardized mean differences (SMD) in post-treatment pain reduction and maximum mouth opening (MMO) were analysed. Thirty-six RCTs were identified that performed comparative outcome assessments for pain and 33 RCTs for MMO. At the short term (≤5 months), IAI-HA (SMD = −2.8, CI: −3.7 to −1.8) and IAI-CS (SMD = −2.11, CI: −2.9 to −1.2) (all very low quality evidence) achieved a substantially greater pain reduction than control/placebo. At intermediate term (≥6 months), a statistically significant decrease in posttreatment pain intensity was observed following Arthroscopy-PRP (SMD = −3.5, CI: −6.2 to −0.82), Arthrocentesis-PRP (SMD = −3.08, CI: −5.44 to −0.71), Arthroscopy-HA (SMD = −3.01, CI: −5.8 to −0.12), TMJ surgery (SMD = −3, CI: −5.7 to −0.28), IAI-HA (SMD = −2.9, CI: −4.9 to −1.09) (all very low quality evidence), Arthroscopy-alone (SMD = −2.6, CI: −5.1 to −0.07, low quality evidence) and Arthrocentesis-HA (SMD = −2.3, CI: −4.5 to −018, moderate-quality evidence) when compared to the control/placebo groups. Relative to MMO, the most effective treatments for short- and intermediate-term improvement were the arthroscopy procedures (PRP > HA > alone, all very low-quality evidence) followed by Arthrocentesis-PRP (very low-quality evidence) and Arthrocentesis-HA (moderate-quality evidence). The non-invasive procedures of occlusal splint therapy, physical therapy, conservative therapy, placebo/control provided significantly lower quality outcomes relative to pain and MMO. The results of the present meta-analysis support a paradigm shift in arthrogenous TMJ disorder treatment. There is a new evidence (though on a very low to moderate quality level) that minimally invasive procedures, particularly in combination with IAI of adjuvant pharmacological agents (PRP, HA or CS), are significantly more effective than conservative treatments for both pain reduction and improvement of MMO in both short (≤5 months) and intermediate term (6 months–4 years) periods. In contrast to traditional concepts mandating exhaustion of conservative treatment options, minimally invasive procedures, therefore, deserve to be implemented as efficient first-line treatments (e.g. IAIs and/or arthrocentesis) or should be considered rather early, i.e. as soon as patients do not show a clear benefit from an initial conservative treatment.