|Home | About | Journals | Submit | Contact Us | Français|
Temporomandibular joint (TMJ) involvement is common in Juvenile Idiopathic Arthritis (JIA). Dexamethasone iontophoresis (DIP) uses low-grade electric currents for transdermal dexamethasone delivery into deeper anatomic structures. The purpose of this study was to assess the safety and effectiveness of DIP for the treatment of TMJ involvement in JIA, and to delineate variables that are associated with improvement after DIP.
Medical records of all JIA patients who underwent DIP for TMJ involvement at a larger tertiary pediatric rheumatology center from 1997 to 2011 were reviewed. DIP was performed using a standard protocol. The effectiveness of DIP was assessed by comparing the maximal inter-incisor opening (MIOTMJ) and the maximal lateral excursion (MLETMJ) before and after treatment.
Twenty-eight patients (ages 2– 21 years) who received an average of eight DIP treatment sessions per involved TMJ were included in the analysis. Statistically significant improvement in the median MIOTMJ (p< 0.0001) was observed in 68%. The median MLETMJ (p= 0.03) improved in 69%, and resolution of TMJ pain occurred in 73% of the patients who had TMJ pain at baseline. Side effects of DIP were transient site erythema (86%), skin blister (4%), and metallic taste (4%). Improvement in TMJ range of motion from DIP is associated with lower MIOTMJ, lower MLETMJ, and absence of TMJ crepitus at baseline.
In this pilot study DIP appeared to be an effective and safe initial treatment of TMJ involvement in JIA, especially among patients with decreased TMJ measurements. Prospective controlled studies are needed.
In about 30–80% of children and adolescents with Juvenile Idiopathic Arthritis (JIA) the temporomandibular joint (TMJ) is affected1–3. Besides pain, swelling, and limitation in the range of motion, TMJ arthritis can present with headaches, neck pain, and pain with mastication. If untreated and persistent, TMJ arthritis can result in micro-retrognathia, facial asymmetry and, therefore, decreased quality of life2,3.
Despite the effectiveness of currently available treatments for JIA in general, the best and safest treatment for a child with TMJ arthritis in isolation or when present with peripheral arthritis has not been well studied4–8. Intra-articular steroid injections for TMJ arthritis in JIA have been shown to improve TMJ range of motion and improvement of TMJ inflammation on magnetic resonance imaging (MRI)9–11.
However, intra-articular steroid injections to the TMJ are not without drawbacks. Generally, procedural sedation is required, possibly lipoatrophy at the injection site can occur, and TMJ avascular necrosis has been reported3,9,11–13. Likewise, intra-articular steroid injection to the TMJ is usually performed by interventional radiologists as only a few pediatric rheumatologists have acquired the procedural expertise of intra-articular steroid injection to the TMJ as part of their training.
Conversely, dexamethasone iontophoresis (DIP) is a non-invasive physiotherapy modality which allows for transdermal delivery of dexamethasone14. Low-grade electric currents lead to the dissociation of hydrophilic medications into ions which move to penetrate deeper anatomic structures15. DIP therapy has been utilized for more than 30 years to treat various musculoskeletal conditions including tendinitis, epicondylitis, enthesitis, and inflammatory peripheral arthritis16–19. Given the relatively superficial position of the TMJ, we hypothesized that DIP is beneficial for the treatment of the TMJ of children with JIA.
This pilot study aimed (1) to assess the effectiveness and safety of DIP when used for the treatment of TMJ involvement in JIA, and (2) to delineate variables that are associated with improvement after DIP.
With approval of the institutional review board, patients with TMJ involvement and JIA who underwent the procedure ’dexamethasone iontophoresis’ from 1997 to 2011 were identified from the electronic medical record (EMR) and billing databases of the Division of Occupational and Physical Therapy at Cincinnati Children’s Hospital Medical Center (CCHMC). Data extraction was performed by RM, PM and SP.
For each patient the EMR was reviewed for gender, JIA subtype, activity of JIA as measured by the number of joints with active arthritis, medication prescribed for the treatment of JIA, duration of JIA, and markers of systemic inflammation [erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP)]. We also recorded the patient’s age at the time of DIP, duration of TMJ involvement prior to DIP, and DIP treatment period (i.e. number of days from start to final treatment sessions), and the number of DIP sessions. Adherence to the prescribed DIP treatment schedule was also noted.
Two anatomical measurements were obtained routinely by the physical or occupational therapist. These were: (1) the maximum inter-incisal opening (MIOTMJ), defined as the distance (in mm) between the upper incisor and the lower incisor on full mouth opening with neutral head position, and (2) the maximum lateral excursion (MLETMJ), defined as the maximum horizontal distance (in mm) that is measured between the upper and lower central incisors (or between the lip frena) with excursion of the mandible to the left or the right side, respectively. The MIOTMJ and MLETMJ measurements were performed using the TheraBite range of motion scale (Atos Medical, Milwaukee, WI).
About 60% of the measurements were done by a one physical therapist (PM). Based on previous studies, the MIOTMJ and MLETMJ have excellent inter-rater and intra-rater reliability. The intraclass correlation coefficient (ICC)inter-rater for MIOTMJ and MLETMJ are 0.9 and 0.77, respectively, while the ICCintra-rater are 0.87 and 0.85, respectively20,21. Earlier research suggested that decreased TMJ range of motion is a good surrogate measure (sign) of TMJ arthritis in JIA22,23.
We recorded whether TMJ findings were unilateral or bilateral. We also documented the presence (versus absence) of symptoms and signs associated with TMJ involvement. They are TMJ pain when chewing, TMJ pain at rest, clicking and crepitus on TMJ examination. The results of MRI and x-ray studies of the TMJ done within three months of the initial and final DIP sessions were also reviewed.
DIP was administered by a trained physical or occupational therapist. The standard protocol consisted of eight to 10 DIP sessions. Time intervals between DIP sessions were one to three days for the initial four to six sessions; thereafter, the frequency of DIP was decreased to once weekly. Fewer sessions were done if the treatment goal of inducing maximal improvement of the TMJ range of motion (MIOTMJ and MLETMJ), and/or resolution of TMJ symptoms and signs was met.
The iontophoresis equipment (Dupel, Empi, St Paul, MN) that was used for DIP featured bipolar electrodes (see Figure 1), the drug delivery and the dispersive electrode (IOGEL®, IOMED Inc; Salt Lake, UT). The delivery electrode was prepared by adding 1.5 mL of dexamethasone sodium phosphate (total dose 6 mg of dexamethasone per TMJ per session). The skin over the TMJ was cleaned with alcohol prior to the placement of the delivery electrode which was connected to the negative pole of a direct current (DC) generator. The dispersive electrode was placed over the trapezius or biceps muscle on the same side of the body as where the delivery electrode was positioned, and connected to the generator’s positive pole. Typically, the electrical current was initiated at 0.5 milliamperes (mA) for the initial 30 seconds of treatment session. After a slow increase as per patient’s tolerance, the electric current was maintained at the highest tolerable level for patient comfort, usually at 4 mA. Settings of the iontophoresis equipment ensured that the current flow continued for a total constant current dose of 40 mA · min which allowed for the entire dexamethasone dose to be administered. Drug delivery was typically achieved over 15 to 30 minutes. After the procedure the electrodes were removed and the skin inspected for signs of irritation.
Descriptive statistics was done using medians and inter-quartile ranges (IQR) for numeric variables, and frequencies (percentages) for categorical data. The primary outcome variable to assess the effectiveness of DIP was the change (in mm) of the MIOTMJ, and the other primary outcome variable was the change in MLETMJ. For the purpose of the analysis we averaged the measurement of the MLETMJ for the left and right excursion when bilateral involvement was present. The absolute MIOTMJ and MLETMJ measurements at baseline were compared to the corresponding final MIOTMJ and MLETMJ measurements after the last DIP session using Wilcoxon signed-rank test. Resolution of TMJ chewing or resting pain was also evaluated.
Furthermore, the response to DIP therapy (yes/no) was defined as achievement of a normal age-adjusted MIOTMJ or MLETMJ24,25 upon completion of the DIP treatment course. Exploratory analyses focused on the presence versus absence of any improvement in the MIOTMJ or MLETMJ measurements upon completion of the DIP therapy.
To identify variables that are associated with improvement to DIP, spearman correlation was performed using the change in the MIOTMJ or the change in the MLETMJ as dependent variable. Variables considered univariately were the number of DIP sessions, bilateral TMJ signs/symptoms (yes/no), number of joints with active arthritis, duration of TMJ involved, JIA disease duration, baseline TMJ resting or chewing pain (yes/no), baseline TMJ click (yes/no), baseline TMJ crepitus (yes/no), JIA subtype, patient’s age at time of DIP, and presence/absence of concomitant medications. Statistical analysis was done using SAS 9.2 software (Cary, NC) and R software (www.r-project.org). P-values ≤ 0.05 are considered statistically significant.
Among 32 JIA patients who underwent DIP for TMJ involvement, four patients were excluded from the subsequent analysis because there was only a single MIOTMJ recorded for each of these patients despite multiple DIP therapy sessions. Twenty-eight patients had serial MIOTMJ measurements, and for 16 of the 28 patients MLETMJ measurements were available. Details on the patient population are provided in Table 1. All subtypes of JIA except for the undifferentiated subtype were represented. Five patients tested positive for antinuclear antibody, and two patients had abnormal ESR and CRP at baseline. Medication regimens remain stable during the DIP treatment period. None of the patients had undergone TMJ intra-articular steroid injections prior to receiving DIP. No other therapeutic modality for the TMJ was prescribed to the patients during the treatment period.
The most common indications for DIP included decreased TMJ range of motion and TMJ pain. At baseline, ten patients (10/28=36%) and nine patients (9/16=56%) had normal age-adjusted MIOTMJ and MLETMJ, respectively. Among the 28 patients, DIP was performed bilaterally in 54% and unilaterally in the others. The median number of DIP sessions was 8 ± 1 (range: 2–14) and the median treatment period was 33 ± 17.5 days (range: 7–74 days).
Adherence to the DIP therapy was seen in 27 (96%). One patient received only two DIP sessions and discontinued the treatment for unknown reasons.
For seven patients a contrast MRI study of the TMJ was done prior to DIP therapy. Findings included condylar flattening (n=5), condylar erosions (n=3), synovial effusion (n=3), synovial hypertrophy (n=4), and synovial enhancement (n=6). Only one patient had a panoramic x-ray done prior to receiving DIP therapy, which showed condylar flattening.
Upon completion of DIP therapy, the median increase in the MIOTMJ and in the MLETMJ was 4.5 mm (p-value<0.0001) and 2.25 mm (p-value=0.01), respectively (see Table 2).
Nineteen of 28 patients (68%) experienced some increase in their MIOTMJ. There was a median increase of 5 mm with DIP in the MIOTMJ of 18 patients with abnormal MLETMJ at baseline (p-value<0.0001), while the median increase in the MIOTMJ was only 0.5 mm in the ten patients who had MIOTMJ that was within the age-adjusted range of normal at baseline (see Table 3).
In 11 of 16 patients (69%) the MLETMJ improved with DIP. Also in Table 3, there was a median increase of 3 mm with DIP in the MLETMJ of seven patients with abnormal MLETMJ at baseline (p-value= 0.03), and a small gain in the MLETMJ by 1 mm in the nine patients who had a MLETMJ that was within the age–adjusted range of normal at baseline.
Fifteen children included in the study reported pain of the TMJ with chewing and/or at rest, which resolved in 11 of them (73%) with DIP therapy. In addition, pre-treatment TMJ click was reported in seven but resolved in only one patient (14%). Similarly, TMJ crepitus resolved in one of five patients (20%).
About a third of the patients did not experience an improvement in their MIOTMJ and MLETMJ with DIP. For three patients the post-treatment MIOTMJ or MLETMJ were smaller than the respective measurements at baseline. A decrease of the MIOTMJ by 1 mm was observed in a 12-year old girl with recently diagnosed very active enthesitis-related JIA but without systemic therapy at the time of DIP. Likewise, despite the resolution of a TMJ click or crepitus, two patients with long-standing oligoarticular JIA who were treated with NSAIDs had a decrease in the MIOTMJ or MLETMJ from baseline; there was a decrease in the final MIOTMJ in one (i.e. from 50 mm to 46 mm) and in the final MLETMJ in the other (i.e. from 13.5 mm to 10 mm). Of note, the post-treatment TMJ range of motion measurements remained within the normal range in the latter two patients.
Transient non-painful site erythema which lasted for 15 minutes was observed in 24 children (24/28=86%) after DIP sessions. One child (1/28=4%) reported a metallic taste during DIP, and another patient experienced a small skin blister (1/28=4%). The latter occurred in an 18 year old patient after a rapid increase in the intensity of the current flow during her final treatment session (8th session).
Improvement in the MIOTMJ after DIP therapy was associated with lower MIOTMJ (p<0.0001), absence of TMJ crepitus (p=0.003) and absence of TMJ click (p=0.02) at baseline. Compared to non-responders, patients who achieved a MIOTMJ within the age-adjusted range of normal were younger [median ± IQR (range): 8 ± 3 years (4–15) vs. 15 ± 5 years (8–21), p-value=0.01], and fewer had baseline TMJ pain (38% vs. 89%, p-value=0.05). Similarly, lower MLETMJ (p=0.02) and absence of TMJ crepitus at baseline (p=0.05) were associated with improvement of the MLETMJ after DIP.
To the best of our knowledge this study is the first to evaluate the use of DIP for the treatment of TMJ involvement with JIA. About two-thirds of the patients experienced an improvement or normalization of the TMJ range of motion which was generally accompanied by resolution of TMJ pain.
Despite the lack of its use for JIA-associated TMJ involvement, DIP has been employed for the treatment of arthritis in the past. An earlier pilot study, adult patients with rheumatoid arthritis experienced a reduction in knee pain with DIP to the knee joints16. Likewise, Ozgocmen et al showed that triamcinolone iontophoresis resulted in a reduction in synovial tissue vascularity on power Doppler sonography, a surrogate for reduced inflammation18.
The biologic rationale for this modality is based on the principle that like charges repel. In this case, dexamethasone sodium phosphate is a negatively charged ion and is applied to the negative pole (anode) of the bipolar iontophoresis set-up; electrical charge applied to the anode forces the drug ions to be pushed through the skin into the deeper tissues. In a study evaluating tissue penetration of iontophoretically administered dexamethasone in rhesus monkeys, dexamethasone was demonstrated in all tissues underlying the electrode down to, and including, tendinous structures and cartilaginous tissues26.
DIP is associated with a tingling sensation which intensifies with higher electric current flow. The use of a lower current flow to achieve the total constant current dose lengthens the treatment time but minimizes this tingling sensation. DIP is appealing for use in JIA since it is painless, non-invasive, and can be performed without sedation. None of the patients developed skin or soft tissue atrophy during the follow-up period.
Intra-articular steroid injections are also effective to treat TMJ involvement with JIA. Previous studies that evaluated intra-articular steroid injection of the TMJ in JIA patients reported increase in the MIOTMJ that ranged between 1.8 mm and 6.9 mm9–11,27, measurements comparable to the overall median increase in the MIOTMJ of 4.5 mm observed in our study.
Intra-articular steroid injections, however, carry the risk of avascular necrosis, soft tissue atrophy, and infections2,9,11–13. Conversely, iatrogenic infections are virtually impossible with DIP, given its non-invasive nature.
Another advantage of DIP over CT-guided intra-articular steroid injection of the TMJ is that the total direct cost of DIP is likely lower. Based on the review of local billing databases the cost of DIP are about 40% of that of CT-guided intra-articular steroid injection of the TMJ. It remains to be determined how the overall costs, direct and indirect, differ between the two treatment modalities.
Although DIP resulted in statistically significant improvement of the MIOTMJ and MLETMJ, the clinical relevance of such quite small absolute gains in range of motions in children remains to be determined. Because the minimal clinically significant change in TMJ measurements in JIA patients is unknown, we used age-adjusted normal ranges24,25 in our secondary analysis to help with the interpretation of the response to DIP therapy.
Not all patients who underwent DIP experienced a therapeutic effect. We hypothesize that non-responders to DIP already had experienced significant TMJ internal derangement or damage which would not be expected to improve in response to non-surgical interventions. This notion is supported by two randomized controlled trials28,29 of DIP therapy for TMJ internal derangement, capsulitis, and osteoarthritis; one reported an increase in TMJ range-of motion but no difference in pain while the other study suggested stable TMJ range-of motion and improved pain.
Our pilot study has several limitations, including the retrospective nature of the study and the lack of controls. However, data were prospectively recorded using either standardized clinic forms or the EMR, resulting in few missing data for the primary outcome variables of this study. Furthermore, to enhance the quality of data collection30, there were three abstractors who were all health providers familiar with the measures used in the study.
The selection of patients who underwent the DIP over other treatment modalities may have introduced some bias as well. DIP for TMJ involvement in our institution is usually done based on patient/parental preference. It is also conceivable that the improvement we observed in the patients was partly because of the patients’ concomitant systemic medications, which were not controlled in this study. Nevertheless, almost all patients were on stable doses of medications for at least six months before the treatment period, and were actually diagnosed to have TMJ involvement while on systemic medications. Persistent TMJ arthritis despite adequate control of peripheral arthritis has been reported in the past11. In our study patients, the presence of concomitant methotrexate or biologic therapy was also not significantly associated with improvement in our correlation analysis.
Ideally, baseline MRI should be done to help assess the degree of inflammation and damage prior to the initiation of DIP therapy, and additional imaging would be desirable to confirm the resolution of inflammation after completion of DIP. Routine serial MRI was not performed in our clinical setting due the need for sedation in the young patients and the substantial cost of MRI. In this study, we considered the increase of TMJ range of motion (MIOTMJ and MLETMJ) as surrogate of TMJ inflammatory changes in JIA22,23.
In conclusion, we found DIP to be an effective and safe treatment modality for JIA patients who have TMJ involvement, especially among those with abnormally low TMJ range of motion measurements and without TMJ crepitus at baseline. Further research is required to determine the optimal number of DIP sessions based on sensitive imaging approaches, durability of treatment response, and performance of DIP in direct comparison to intra-articular steroid injection for the TMJ.
The authors would like to acknowledge Joshua Pendl and Irene Calderon for their contribution to the paper.
Dr. Brunner is supported by the NIH-grants: 5U01-AR51868, P60-AR047884 and 2UL1RR026314. Dr. Mina is supported by a NIAMS Training award T32100291.