Local vs systemic corticosteroids in the treatment of carpal tunnel syndrome

Methods.

This was a prospective, randomized, double-blind, parallel treatment study in which all patients received steroid treatment either by local injection or by mouth as well as placebo treatment by the alternate route; ethics approval had been obtained from the institutional committee, and patients gave written informed consent. We recruited patients with newly diagnosed CTS of longer than 3 months’ duration from our orthopaedic and medical clinics. Co-interventions such as splints and other oral medications were withheld for the duration of the study. To confirm the diagnosis, motor and sensory nerve conduction studies were performed in median and ulnar nerves by using standard techniques of supramaximal stimulation and surface electrodes, according to the American Academy of Neurology summary statement on the desired nerve conduction and electromyographic studies in patients with suspected CTS.8 Compound muscle action potentials (CMAP), distal motor latencies (DML), nerve conduction velocities (NCV) and sensory nerve action potentials (SNAP) were measured, and needle electromyographic recordings (Viking III P, Nicolet W) of the abductor pollicis brevis (APB) were performed. Patients were enrolled into the study if they fulfilled the criteria below:

1. 1. Sensory symptoms over median nerve distribution

2. 2. Confirmatory electrophysiologic results: prolonged median nerve DML >4 ms or median-ulnar palmer sensory latency difference >0.5 msec

3. 3. Failed splinting for 2 months or more

The following exclusion criteria applied:

1. 1. Patients with evidence of severe CTS: fibrillation potentials or reinnervation on needle examination of the abductor pollicis brevis muscle or clinical examination showing wastage of the thenar muscles. (These patients were referred for surgical decompression on presentation).

2. 2. Clinical or electrophysiologic evidence of accompanying conditions that could mimic CTS or interfere with its evaluation, such as proximal median neuropathy, cervical radiculopathy, or significant polyneuropathy.

3. 3. Contradiction to steroid use.

4. 4. History of underlying disorders associated with CTS such as diabetes mellitus, rheumatoid arthritis, pregnancy, acromegaly, or hypothyroidism.

5. 5. Patients who have received previous steroid injection or oral steroid therapy for CTS.

Patients were randomly assigned to one of two groups: the first group received oral placebo daily for 10 days and a single 15-mg methylprednisolone acetate injection3 locally into the carpal tunnel, and the second group received prednisolone 25 mg daily for 10 days and the same volume of saline injection into the carpal tunnel. In those patients with bilateral CTS, both hands were injected. The randomization was performed by a computer-generated list of numbers. A single operator (S.M.W.) performed all the injections according to the method described by Girlanda,3 and a research nurse according to the randomization protocol drew up the injections with micropore tapping over the syringes, thus ensuring blinding. The placebos were identical in appearance to the prednisolone tablets and distributed by the same research nurse according to the randomization protocol. Neither the doctor (S.M.W.) nor the patients knew what treatment was given, and they remained masked to treatment during the assessments.

The primary end point variable was the Global Symptom Score (GSS) as described by Herskovitz et al.7 The GSS rated symptoms from 0 (no symptoms) to 10 (severe) in each of five categories: pain, numbness, paresthesia, weakness/clumsiness, and nocturnal awakening. The GSS was the summation of these five scores and ranged from 0 (absence of symptoms) to 50 (most severe symptoms). Subjects’ GSS was recorded at baseline, 2, and 8 weeks, and any side effects were recorded by the same physician (C.F.H.), who was masked to the treatment allocation, by telephone interview. All subjects returned for assessment by the author (C.F.H.) at 12 weeks, and the GSS was also recorded.

The sample size of 60 was planned to enable achievement of 80% power with an α = 0.05 for detecting a 40% difference in GSS between the treatment groups, assuming the response rate in the injection group to be 70% and that of the oral corticosteroids group 30%.

Two-sample t-test was used for comparison between the oral steroid group and the injection group during each time phase, considering the null hypothesis because there was no mean difference in GSS between the two groups. In addition, the 95% confidence intervals for means at each interval were calculated and display graphically. All t-tests were two-sided, with a significance level of 0.05, and their use was justified by the normality test. SPSS 9.0 was used for performing the statistical analysis.

Results.

Sixty-two patients consented to this study, and two were excluded because of coexisting diabetes mellitus. Thirty patients were randomized to steroid injection and oral placebo and 30 patients to oral steroid and placebo injection (figure 1). Subject demographic and clinical characteristics were comparable (table 1). The mean GSS was not significant between groups at baseline and at 2 weeks. Mean GSS at baseline was 25.0 (SD 6.4) for the steroid injection group and 25.7 (SD 8.3) for the oral steroid group, p = 0.70. At 2 weeks, the mean GSS was 13.6 (SD 7.5) for the injection group and 17.8 (SD 10.0) for the oral group, p = 0.07.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Patient flow.

Significant improvement for the injection group was present at both 8 weeks (p = 0.002) and 12 weeks (p = 0.004). Mean GSS was 13.7 (SD 8.3) for the injection group and 20.8 (SD 8.7) for the oral group at 8 weeks and a mean of 14.3 (SD 8.4) for the injection group and 21.4 (SD 9.6) for the oral group at 12 weeks (table 2,figure 2A). Side effects were minimal in both groups and consisted of local injection pain, gastrointestinal upset, insomnia, and increased appetite. Comparison of changes from baseline between groups was also performed, showing a significant decrease in mean GSS at all the time points (2, 8, and 12 weeks) in the injection group (p < 0.001). In the oral group, a significant difference was only found at 2 weeks (p = 0.001) and 8 weeks (p = 0.03) (figure 2B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. (A) Error plot of mean scores between the oral corticosteroids and steroid injection in relation to time. ▪ = Oral group; ▴ = injection group. (B) Error plot of Global Symptom Score mean difference between baseline and specific time phase in the oral corticosteroids and steroid injection. ▪ = Oral group; ▴ = injection group; dashed line = reference line at 0.

Discussion.

To our knowledge, this is the first randomized clinical trial comparing the efficacy of steroid injection and oral corticosteroids in the treatment of CTS. Our study shows that local steroid injection was superior to oral corticosteroids over a 3-month period in patients with CTS. The age and sex distribution of the studied population corresponds with the data reported in the literature, and there was no difference in comparison between the two groups on the demographic parameters (not shown). Despite the widespread use of corticosteroids injection into the carpal tunnel in CTS, there are actually few randomized controlled trials in confirming its usefulness.2,3,5⇓⇓ The only controlled trial comparing the effect of local and systemic steroid route was in favor of the local route,2 but the dose of systemic steroids used by intramuscular injection was much smaller than in our study. In some way our study confirms the efficacy of steroid injections in the treatment of CTS. Moreover, the duration of efficacy (at least 12 weeks) was also impressive and is in keeping with a recent report.5 This cannot be accounted for by the half-life of depomedrone, whereas the oral prednisolone group was only effective for up to 8 weeks from intragroup analysis (figure 2b). Both treatment groups were associated with minimal steroid side effects, and no withdrawals were noted. Because CTS is primarily a symptomatic disorder, we used self-assessment by masked patients through the GSS as the primary outcome measure in this study, although the patient-oriented outcomes do not always progress as neurophysiologic measures do.7,9,10⇓⇓ This study showed a clear benefit from steroid injection versus oral steroid in the treatment of CTS.