STEROID INJECTION FOR CARPAL TUNNEL SYNDROME L. R, IRWIN, R. BECKETT and R. K. SUMAN
From the Hand Unit, District GeneralHospital, Grimsby, UK We studied the medium- to long-term results of steroid injection into the carpal tunnel of women with the carpal tunnel syndrome (CTS). Of 45 hands, only II had lasting relief of symptoms and 22 had no relief whatsoever. There was no correlation of the typical signs and symptoms of CTS with outcome. Other series have offered various predictive factors for the 6utcome of injection but we found little or no correlation between these factors and outcome.
Journal of Hand Surgery (British and European Volume, 1996) 2lB." 3:355-357 The use of steroid injection into the carpal tunnel for relief of carpal tunnel syndrome (CTS) is contentious. Such injections are widely practised, but their efficacy has been the subject of much debate (Foster, 1960; Phalen, 1966; Myles et al, 1973; Wood, 1980; Girlanda et al, 1993). Even in those series with a good initial response to injection there appears to be a marked recurrence in the longer term (Wood, 1980; Phalen, 1972; Green, 1984; Girlanda et al, 1993). Studies have attempted to determine factors predictive of response to injection. Gelberman et al (1980) suggested that the factors predictive of a poor response were: symptoms present for over 12 months; thenar muscle atrophy; weakness; distal motor latencies > 6m/s.; absent sensory responses. Kaplan et al (1990) suggested the following factors: age over 50 years; symptoms present for more than 10 months; constant paraesthesiae; stenosing flexor tenosynovitis; a Phalen test positive in under 30 seconds. They went on to assess the likely success rates of injection when one or more of these factors was present. If none were present, the likely success rate was said to be 67%. This fell to 40.4% with one factor present, 16.7% with two present, 6.8% with three present and 0% with four or more present. We report a prospective study of the results of steroid injection for carpal tunnel syndrome and we have also studied the predictive factors of Gelberman et al (1980) and Kaplan et al (1990).
100 mg hydrocortisone with 4 ml lignocaine directly into the carpal tunnel. Splintage was not used. All patients were reviewed 1 month later, then again 3 months and 6 months later. Patients were seen for this study a minimum of 1 year after injection. Response to the injection was determined by direct questioning of each patient and was based on subjective answers. Information was collected and assessed according to the protocols of Gelberman et al (1980) and Kaplan et al (1990; see above) and a judgement made of the reliability and relevance of these factors in predicting eventual outcome.
RESULTS
Forty-two patients were included in the study but records of one were lost, leaving 41 patients with 45 affected wrists. Ages of the patients ranged from 24 years to 88 years (average 47.7 years). The duration of symptoms ranged from 2 weeks to 20 years (average 18 months). No patient complained of constant pain,, nor did anyone have symptoms or signs compatible with tenosynovitis. Phalen's test was positive in 32 hands and negative in 13. Thenar atrophy was present in 4 hands and absent in 41. Nerve conduction studies were carried out on 11 hands. Of the tests carried out, seven were diagnostic of nerve entrapment and four were negative. The assessment of Gelberman's and Kaplan's predictive factors revealed that there was virtually no correlation between these factors and outcome in our patients (Table 1). Statistical analysis of these data, deriving correlation coefficients, showed that only "Gelberman status" had any meaningful positive correlation at about 0.2, and this is indicative of only a very weak effect (Table 2). Eleven hands were injected and went on to enjoy symptomatic relief in excess of 1 year. These hands are regarded as having had a lasting response to the injection. Twelve hands had initial relief of symptoms, but relapsed within the t year study period. Twenty-two hands had no response to the injection.
METHOD
All female patients referred to one consultant (RKS) and diagnosed as havi~ag CTS were studied prospectively over a period of 9 months. No case was excluded and no attempt was made to blind data collection. Diagnosis of carpal tunnel syndrome was made from clinical signs and symptoms. Any female patient presenting with the classical signs and symptoms of CTS was entered in the study. When diagnostic doubt existed, nerve conduction studies (NCS) were carried out. Rheumatoid patients were not actively excluded from the study group, but the group only contained two patients with the diagnosis of rheumatoid arthritis. First line treatment for all cases was injection of 355
THE JOURNAL OF HAND SURGERY VOL. 21B No. 3 JUNE 1996
356
Table 3 - - S u c c e s s rate of injection in other studies
Table 1--Recorded factors and eventual outcome
Parameters
Age in years Symptom duration (months) N u m b e r of cases Positive Phalen tests Negative Phalen tests Thenar atrophy present Positive NCS Negative NCS "Gelberman" positive "Gelberman" negative 0 " K a p l a n " factors 1 " K a p l a n " factors 2 " K a p l a n " factors 3 " K a p l a n " factors
No response to injection
Initial response only
Lasting response
48.8 (24 88) 22.8
41.7 (24-53) 17.4
53 (31 79) 43.1
22 14 8
12 10 2
11 8 3
2
1
1
4 2 11
2 1 6
1 1 3
11
6
8
3 8 8 3
0 6 5 1
1 5 3 2
Table2--Statistical correlation between various factors with the outcome of injection
Factor Age < 5 0 Duration of symptoms < 1 year Duration of symptoms < 18 m o n t h s Thenar atrophy Phalen test Nerve conduction studies "Gelberman status" No. of " K a p l a n factors"
Correlation coefficient --0.11747 0.039078 --0.06333 -0.14192 - 0.12267 0.109135 0.196561 0.00828
DISCUSSION We found no correlation between any of the factors studied and eventual outcome except the duration of symptoms. Injection, perhaps surprisingly, appeared to be more beneficial for patients whose CTS symptoms had been present for longer. The results also show that 22 of 45 (49%) patients failed to respond to injection, 12 (27%) responded initially but then relapsed and 11 (24%) had reasonably long-term relief of symptoms. Three of the four bilateral cases had different outcomes in their two hands. Many series have been published addressing this subject and a selection is tabulated below Table 3. Almost all showed a high response rate in the early stages (less than a month) and all showed a much poorer long-term (more than a year) symptom-free response. Some authors have enthusiastically recommended injection as an effective, permanent treatment (Kendall, 1962), yet other studies reveal a recurrence rate of 92% in the longer term (Girlanda et al, 1993). Others have found that injection does not influence the natural history (Myles et al, 1973) which suggests in fact that many
Author
Number of hands
% initially successful
% symptom free long-term
Gelberman et al
50
76
22
Ozdo~an & Yazma (1984) Foster (1960) Kendall (1962) Wood (1980) Phalen (1966) Myles et al (1973) Girlanda et al (1993) Crow (1960) Phalen (1972) Green (1984)
37
-
22
25 170 61 270 65 91 81 497 222
92 80 90 62 44 90 25 66 81
4 40 6 24 33 8 15 13 11
()980)
patients will improve spontaneously. Most authors feel that injection should have a better defined role as a temporary relief for self-limiting CTS (e.g. that occurring in pregnancy; Crow, 1960) or as a trial to limit the number of patients requiring surgery (Phalen, 1966; Phalen, 1972; McConnell and Bush, 1990). Others would accept the technique as being suitable to "tide-over" a patient who is awaiting surgery (Crow, 1960; McConnell and Bush, 1990). Identifying the correct role for steroid injection is therefore not easy when confronted with such contradictions. One group suggests that those responding poorly to injection will do the same when subjected to surgery (Frederick et al, 1992). A logical step therefore, is to suggest that there is one group that will do badly whatever the type of treatment and another that will always do well. Our experience of injection in males has not been favourable. We are aware that many studies include both males and females. Direct comparison of our figures with those studies is therefore not entirely reliable. Other studies Carried out in females (Ozdo~an and Yazici, 1984) or almost entirely on females (Crow, 1960), however, reveal results broadly similar to ours. Our approach to the use of injection in this condition is to reserve it exclusively for women. We accept its low long-term efficacy and observe the patients closely following injection, believing that if symptoms are allowed to persist the results of surgery will be compromised. We are aware of the potential risks of injecting the median nerve (McConnell and Bush, 1990; Frederick et al, 1992; Linskey and Segal, 1990), which is another reason for exercising caution with the technique. Other than in cases of minor trauma and pregnancy, we advise our patients that most will benefit only for a short period and that it serves to relieve their symptoms temporarily while awaiting surgery which will be expedited if the injection fails completely. We had intended to make use of the factors of Gelberman et al (1980) and Kaplan et al (1990) in future cases to predict the likely outcome of injection.
357
STEROID INJECTION FOR CTS
We have now found that they do not correlate with outcome in our patients and we have abandoned this plan. We also feel that we will probably inject fewer wrists in future. General practitioners who are now carrying out more carpal tunnel injections should be advised that injection is more likely than not to fail. They should not delay referral as prolonged symptoms may compromise the results of later surgery. References CROW R S (1960). Treatment of the carpal-tunnel syndrome. British Medical Journal, 1: 1611-1615. FOSTER J B (1960). Hydrocortisone and the carpal-tunnel syndrome. Lancet, 1: 454-456. FREDERICK H A, CARTER P R and LITTLER J W (1992). Injection injuries to the median and ulnar nerves at the wrist. Journal of Hand Surgery, 17A: 645 647. GELBERMAN R H, ARONSON D and WEISMAN M H (1980). Carpaltunnel syndrome: Results of a prospective trial of steroid injection and splinting. Journal of Bone and Joint Surgery, 62A: 1181-1184. GIRLANDA P, DATTOLA R, VENUTO C et al. (1993). Local steroid treatment in idiopathic carpal tunnel syndrome: short- and long-term efficacy. Journal of Neurology, 240: 187-190. GREEN D P (1984). Diagnostic and therapeutic value of carpal tunnel injection. Journal of Hand Surgery, 9A: 850-854.
KAPLAN S J, GL1CKEL S Z and EATON R G (1990). Predictive factors in the non-surgical treatment of carpal tunnel syndrome. Journal of Hand Surgery, 15B: 106-108. KENDALL P H (1962). Carpal-tunnel syndrome. British Medical Journal, l: 115. LINSKEY M E and SEGAL R (1990). Median nerve injury from local steroid injection in carpal tunnel syndrome. Neurosurgery, 26: 512-515. McCONNELL J R and BUSH D C (1990). Intraneural steroid injection as a complication in the management of carpal tunnel syndrome. Clinical Orthopaedics and Related Research, 250: 181-184. MYLES A B, CASEMORE V A, COULTHARD M, G1LLIAM G and SCHILLER L F G (1973). Management of the carpal tunnel syndrome with local corticosteroid injections. Rheumatology and Rehabilitation, 12: 205-208. OZDO(~AN H and YAZICI H (1984). The efficacy of local steroid injections in idiopathic carpal tunnel syndrome: A double blind study. British Journal of Rheumatology, 23: 272-275. PHALEN G S (1966). The carpal-tunnel syndrome. Journal of Bone and Joint Surgery, 48A: 211-228. PHALEN G S (1972). The carpal-tunnel syndrome: clinical evaluation of 598 hands. Clinical Orthopaedics and Related Research, 83: 29-40. WOOD M R (1980). Hydrocortisone injections for carpal tunnel syndrome. The Hand, 12: 62-64.
Accepted: 16 October 1995 Mr L. R. Irwin, Orthopaedic Department, St James's University Hospital, BeckeU Street, Leeds LS9 7TF, UK. © 1996The British Society tbr Smgery of the Hand