Low-dose, short-term oral prednisone in the treatment of carpal tunnel syndrome

Carpal tunnel syndrome (CTS)

is the most common peripheral nerve entrapment and a significant source of morbidity. Conservative, nonoperative therapy is appropriate, at least initially, for most patients with mild to moderate CTS. ^[1] Established therapies such as splinting and steroid injection into the carpal canal are frequently effective, though recurrences are common. ^[1-5] There are no adequate studies of the utility of common therapies such as nonsteroidal anti-inflammatory drugs (NSAIDs), diuretics, or oral steroids. ^[1,6-8] This study evaluates the effectiveness of low-dose, short-term oral prednisone in ameliorating the pain and other symptoms of CTS.

Methods.

We conducted a prospective, randomized, double-blind, placebo-controlled study. Patients referred to our EMG laboratory for suspected CTS were eligible if they met the following criteria: (1) focal signs and symptoms in at least one hand consistent with a clinical diagnosis of CTS, including sensory symptoms restricted to the median nerve distribution (pain, numbness, tingling, nocturnal exacerbation of symptoms) and only minimal to moderate weakness of thenar muscles; (2) confirmatory electrophysiologic evidence of focal slowing of transcarpal (palm to wrist) median mixed nerve conduction compared with the distal (palm to digit 2) segment (more than 12 m/sec slowing) or, if no sensory response, a prolonged distal motor latency (more than 4.5 msec); and (3) patients were aged 18 years or older. Exclusion criteria included (1) clinical or electrophysiologic evidence of accompanying conditions that could mimic CTS or interfere with its evaluation, such as cervical radiculopathy, proximal median neuropathy, significant polyneuropathy, or marked orthopedic abnormalities, (2) moderate to severe thenar muscle weakness or atrophy, or EMG evidence of more than mild motor axon degeneration, (3) cognitive impairment that interfered with the subject's ability to follow instructions and relate symptoms, and (4) recent peptic ulcer disease or history of steroid intolerance.

At the screening visit, evaluation included standardized medical and neurologic history, detailed neurologic examination (NE), and bilateral nerve conduction studies (NCSs) and needle EMG. The NE of the hands included evaluating thenar atrophy, weakness of abductor pollicis brevis and opponens pollicis, light touch and pinprick sensation, and Tinel's and Phalen's signs. Eligible subjects who consented to the study discontinued prior treatment, if any, and returned for a baseline assessment after 1 week. The baseline assessment included bilateral quantitative thermal and vibratory threshold testing (QST) of the index finger and a standardized symptom questionnaire. The QST devices used were the Sensortek Vibratron II for vibration and the Sensortek NTE-2 for thermal testing (warm; thermode applied to fingertip), using a forced-choice algorithm. The questionnaire used a 10-point ordinal scale on which symptoms were rated from 0 (no symptoms) to 10 (severe); patients rated pain, numbness, paresthesias, weakness/clumsiness, and nocturnal awakening.

The primary endpoint variable was the mean score of the five rated symptoms; this score was termed the "global symptom score" (GSS). Following the baseline assessment, patients were randomized to either a 2-week treatment period with prednisone, 20 mg daily for the first week followed by a second week of 10 mg daily, or 2 weeks of matched placebo treatment. Follow-up assessments at 2, 4, and 8 weeks were identical, including NE, NCSs, QST, and standardized symptom questionnaire. Patients also recorded adverse side effects, use of analgesics, and interval change in symptoms. Needle EMG was not performed at the follow-up assessments. No other treatments (including splints and NSAIDs) were allowed during the study period, aside from acetaminophen or acetaminophen plus codeine.

In patients with bilateral CTS, the most symptomatic hand was selected for analysis; if both hands were comparable, the dominant one was chosen. Data analysis was performed with t tests for continuous variables and chi-square or Fisher's exact test for dichotomous variables.

Results.

Six patients were randomized to prednisone and nine to placebo. Two patients initially randomized to prednisone and one to placebo were lost to follow-up immediately after baseline evaluation and were not included. Subject demographic and clinical characteristics were comparable Table 1.

The primary endpoint variable, the mean GSS, comprised the sum of ratings for pain, numbness, paresthesias, weakness/clumsiness, and nocturnal awakening generated for each evaluation point Figure 1. Significant improvement for the prednisone-treated group was present at 2 weeks (p less than 0.05). After treatment was discontinued, the 4- and 8-week mean GSS was not significantly different from the control group but remained below the baseline value. By self-report, all patients in the prednisone-treated group, versus three of nine in the placebo group, reported overall improvement (p equals 0.02) at any of the observation points; at 8 weeks, improvement over baseline was still greater for the treated group (p equals 0.01). Self-reported improvement specifically for pain favored the treated group at 2 (p equals 0.07), 4 (p equals 0.03), and 8 (p equals 0.008) weeks. In addition, four in the treated group and none in the placebo group had complete resolution of pain and nocturnal exacerbations at some point in the course. In both groups, the contralateral hand in those with bilateral symptoms paralleled the predominantly symptomatic or dominant hand. Three patients in the prednisone-treated group had a history of splint use; two of these reported no benefit from the splint but responded to prednisone.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Mean global symptom score at baseline and 2, 4, and 8 weeks after prednisone or placebo. *p less than 0.05.

Subjective improvement in the treated group began within 1 to 4 days in five of six patients and at 3 weeks in the sixth. Symptoms began to return gradually within 1 to 17 days of the end of the 2-week treatment period. One patient in each group used analgesics. Three patients in each group reported adverse reactions. In the prednisone-treated group these included nausea/abdominal discomfort, constipation, and dysgeusia; one, a diabetic, had mild hyperglycemia. The placebo group reported nausea/abdominal discomfort, constipation, insomnia, headache, dysuria, and burning nostrils. None of these symptoms prompted discontinuation of therapy.

At screening, mean NCS data were similar in the prednisone and placebo groups, including median distal sensory latency (3.4 versus 3.4 msec), sensory conduction velocity (41 versus 43 m/sec), transcarpal mixed nerve conduction velocity (35 versus 37 m/sec), and distal motor latency (5.5 versus 4.2 msec). Thenar motor amplitudes were normal (7.2 versus 10.8 mV), and three patients in each group had mild (1 plus) active denervation in the abductor pollicis brevis muscle. Mean thermal and vibration index finger QST was normal and comparable in the two groups. There was no significant change in the NE, NCSs, or QST on follow-up testing.

Discussion.

In this small, controlled trial, a 2-week course of oral prednisone, 20 mg daily for the first week followed by 10 mg daily for a second week, resulted in significant improvement in pain by self-report as well as in GSSs of pain, numbness, paresthesias, weakness/clumsiness, and nocturnal exacerbation at the 2-week point. Subjective improvement began within 1 to 4 days of treatment in five of six patients. Symptoms gradually worsened after the treatment period, but self-reported benefits were maintained at the 4- and 8-week evaluations. Though symptom scores did not return to baseline, group differences were no longer statistically significant. There were no significant changes in the NE, NCSs, or QST. Side effects of treatment were minor and comparable in the two groups. We deliberately chose a low-dose, short-term regimen to minimize the likelihood of adverse side effects that would make this approach unappealing.

This study had some limitations. First, all patients had mild to moderate CTS; accordingly, results cannot be generalized to those with more severe CTS. Second, the sample size is small, although statistically significant differences were observed. Although there are no other data on low-dose oral steroids, several lines of evidence lend credence to our results. Arguments suggesting that oral steroids might be effective include the following: (1) nonspecific tenosynovitis is thought to be the most common cause of CTS, (2) injection of local steroids into the carpal tunnel is effective, (3) high-dose oral steroids are reported to ameliorate CTS in a few cases associated with connective tissue diseases, ^[9] and (4) anecdotal experience, including our own, suggests that steroids in CTS are helpful. ^[1,6-8] A questionnaire survey revealed that 10% of hand surgeons included oral steroids as a nonsurgical treatment modality. ^[6]

Additional, effective, conservative/nonoperative therapies would be welcome in the management of CTS. Although splinting is simple, safe, and often effective, its usefulness is frequently short-lived. ^[1,2] As many as 90% of patients treated with steroid injection in the carpal tunnel obtain initial relief, but most relapse after weeks to months. ^[1-5] Because this procedure can be complicated by mechanical or chemical nerve injury, the number of injections should be limited. ^[1,2,10] Low-dose oral steroids, either alone or perhaps in combination with splinting, may be a helpful conservative treatment option for rapid symptomatic improvement. This strategy seems rational given that approximately 25% of patients obtain no relief with splints and that in another one-quarter of cases, the onset of relief is delayed for 1 to 2 weeks. ^[1] Whether oral steroids will have any effect on the subsequent long-term course of CTS is unknown.

There are a number of questions that this study raises and that future studies might address. (1) Would larger doses or longer treatment periods result in better or more sustained benefit? (2) Would lower doses or even alternate-day therapy be equally effective with fewer side effects? (3) Would repeated short courses every few months be an effective approach in managing some patients? (4) What is the relative efficacy of steroids, NSAIDs, and pyridoxine in the short- and long-term management of CTS?

Acknowledgments

Our thanks to Lisa Mirsky for help with coordinating the study and to Denis Giblin, MD, who inspired its creation.