- Email: [email protected]
CONTROLLED TRIAL OF PREDNISOLONE IN ACUTE POLYNEUROPATHY
750 The
reported objection to suture closure has been that
50-70% of patients will have further ulcer symptoms
or
complications.8 10 Sawyers and Herrington4 have performed proximal gastric vagotomy with suture closure for perforated duodenal ulcer in 21 patients, with no mortality and no recurrent symptoms over a 3-year period. This procedure may represent a compromise between operations which do not protect against recurrent ulceration and those which subject symptom-free patients to the hazards of gastric surgery. 30-47% of patients have been reported to be symptom-free after suture closure.36 11-13 In our study 48% of patients are still free of symptoms 3 months to 6 years after suture closure. Therefore, if emergency definitive surgery had been performed on these patients,, nearly half would have had an unnecessary and extended operation. Most surgeons agree that patients with a long history of symptoms should be considered for ulcer-curative surgery,36 12 14 and this has led to the classification of
peptic ulcers into "acute" and "chronic", depending
on
the duration of symptoms before perforation.8 Several workers use a period of 3 months as the dividing line. However, 28% of our patients had no symptoms before perforation.’o 11 16 Age, race, and sex do not influence prognosis significantly but the prevalence of perforation is higher in men than women. 13 17
Having determined the natural history of duodenalpatients treated with simple suture with Tutt and van Heerden, who sugclosure, agree choice of treatment, with long-term the first this as gest considerations exclude us from now follow-up.18 Ethiçal trial a clinical of emergency definitive treatundertaking ment. If primary definitive surgery is to be undertaken, then it should be reserved for patients with a long history of dyspepsia. Even this may be unjustified if H2-receptor antagonists are shown to heal chronic duodenal ulcers. 19-23 Requests for reprints should be addressed to S.N.J. ulcer
perforation
in
we
REFERENCES 1. MacLaren, R. Br. med. J. 1894, ii, 863. 2. Cooley, D. A., Jordan, G. L., Brockham, H.
L., De Bakey, M. E. Ann. Surg. 1955, 141, 840. 3. Jarrett, F., Donaldson, G. A. Am. J. Surg. 1972, 123, 406. 4. Sawyers, J. L., Herrington, J. L. Ann. Surg. 1977, 185, 656. 5. Jordan, G. L., Jr, De Bakey, M., Cooley, D. A. Am. J. Surg. 1963, 105, 396. 6. Jordan, G. L., Angel, R. T., De Bakey, M. D. Archs Surg., Chicago, 1965, 92, 449. 7. Booth, R. A. D., Williams, J. A. Br. J. Surg. 1971, 58, 42. 8. Illingworth, C. E. W., Scott, L. D. W., Jamieson, R. A. Br. med. J. 1946, i, 787. 9. Dean, A. C. B., Clark, C. G., Sinclair-Gieben, A. H. Gut, 1962, 2, 60. 10. Hofkin, G. A. Am. J. Surg. 1966, 111, 193. 11. McDonough, J. M., Foster, J. H. ibid. 1972, 123, 411. 12. Greco, R. S., Cahow, C. E. ibid. 1974, 127, 109. 13. Robbs, J. V., Moshal, M. G., Baker, L. W. S. Afr. J. Surg. 1977, 15, 39. 14. Maynard, A. de L., Froix, C. J. L., Oropeza, G. Archs Surg., Chicago, 1968, 97, 96. 15. Taylor, H., Warren, R. P. Lancet, 1956, i, 397. 16. Cassell, P. Gut, 1969,10, 572. 17. MacKay, C. Br. med. J. 1966, i, 701. 18. Tutt, G.O., van Heerden, J.A.S. Afr. J. Surg. 1976, 14, 1. 19. Blackwood, W. S., Maudgal, D. P., Pickard, R. G., Lawrence, D., Northfield, T. C. Lancet, 1976, ii, 174. 20. Gray, G. R., McKenzie, I., Smith, I. S., Crean, G. P., Gillespie, G. ibid. 1977, i, 4. 21. Gray, G. R., Smith, I. S., Mackenzie, I., Gillespie, G. Scott. med. J. 1977, 22, 293. 22. Semb, L. S., Berstad, A., Myren, J., Foss, J. C., Carlsen, E., Kruse-Jenson, A. in 2nd Int. Symp. Histamine H2-receptor Antagonist (1977) (edited by W. L. Burland and M. A. Simpkins); p. 248. Amsterdam, 1977. 23. Spence, R. W., Celestin, L. R., McCormick, D. A. Gut, 1977, 18, 420 (abstr.).
CONTROLLED TRIAL OF PREDNISOLONE IN ACUTE POLYNEUROPATHY R. A. C. HUGHES
Guy’s Hospital, London SE1
J. M. NEWSOM-DAVIS Royal Free Hospital, London NW3 and National Hospital for Nervous Diseases, London WC1
G. D. PERKIN
Charing Cross Hospital, London W6
J. M. PIERCE Department of Biomathematics, University of Oxford multicentre, randomised trial of prednisolone in acute polyneuropathy of undetermined ætiology (Guillain-Barré syndrome), 21 patients were treated with prednisolone (60 mg daily for one week, 40 mg daily for four days, and then 30 mg daily for three days) and 19 did not have steroid treatment. Patients were graded on a six-point scale by one of two neurologists who had no knowledge of the treatment schedule. Reassessment at one, three, and twelve months consistently showed greater improvement in the
Summary
In
a
control than the
prednisolone group but the only statistically significant result was in the improvement at three months among patients entered to the trial within a week of onset of illness. The 6 control patients had improved by 2·5±0·43 grades by three months from entry to the trial whereas the 10 prednisolone patients had only improved by 0·9±0·46 grades (P<0·05). There was 1 death related to the polyneuropathy in each group, and 1 suicide in a control patient during convalescence. 6 prednisolone patients were left with considerable disability compared with 1 control patient. There were 3 relapses in the prednisolone group, but none in the control group. The results indicate that steroid treatment is not beneficial and can be detrimental in acute neuropathy of undetermined ætiology.
Introduction THE usefulness of steroids in the Guillain-Barre syndrome has remained in doubt despite many retrospective studies. The disease (incidence about 1-7/100 0001) is sufficiently rare to make a single-centre study of steroids difficult. Acute polyneuropathy is not always a mild disease with a good prognosis: 10-23% require tracheostomy,2-5 only about 60% recover completely,2-5 7-22% are left significantly disabled,2-4 mortality 5 rates of 2.4-5% are still reported4’ and 3-10% relapse.2-4 We carried out a randomised multicentre trial of prednisolone versus no steroid treatment to establish whether steroids improve the outcome.
Methods All
with with
neurologists in London were invited to enter patients polyneuropathy of undetermined aetiology. Patients porphyria, collagen disease, carcinoma, severe diabetes,
acute
751 clinical evidence of current infective illnesses were excluded. Patients were included even if they had a normal protein concentration or pleocytosis in the cerebrospinal fluid. Patients who had been started on steroids before being seen by the admitting neurologist and those with contraindications to steroid treatment were excluded. Since it was decided during the course of the trial that those who had begun to improve before being seen should also be excluded 1 patient had to be withdrawn from each group: Approval of the trial was independently given by the ethical committee of two hospitals. The decision to ask for the consent of the patient was left to the physician in charge of each case. At entry patients were randomised to prednisolone or control treatment according to a central register. One group received prednisolone 15 mg four times daily for one week, 10 mg four times daily for four days, and 10 mg three times daily for three days after which the dose was reduced or continued at the discretion of the physician-incharge. The dose for children was reduced according to age: the starting dose was 10 mg four times daily at age 7, 10 mg three times daily at age 3, and 5 mg four times daily at age 1. Assessments were made within 24 h of entry by one of two assessors who had no knowledge of the treatment schedule, and again after one, two, three, and four weeks, two months, three months, six months, nine months, and one year. All cases were followed-up for one year although 1 boy who emigrated to Australia was assessed on the last three occasions by another neurologist. On each occasion a full neurological assessment was made and graded as follows: 0. Healthy. 1. Minor signs or symptoms of neuropathy but capable of manual work. 2. Able to walk without support of a stick but incapable of manual work. 3. Able to walk with a stick, appliance, or support. 4. Confined to bed or chairbound. 5. Requiring assisted ventilation. 6. Dead. 2 more patients were withdrawn from the control group, 1 because the diagnosis was later considered to be incorrect and 1 because prednisolone had been given before entry to the trial. 3 other patients were retained within the groups to which they had been allocated despite the following irregularities in the conduct of the trial. 1 mild case was allocated to the prednisolone group on a Friday but the drug was not given, and since, by the Monday, the patient had improved treatment was considered unnecessary. 1 control patient had a persistent demyelinating motor polyneuropathy and a grossly raised E.S.R. After four months he was started on steroids but soon died of bronchopneumonia. A third control patient deteriorated and thirteen days after entry was given A.c.T.H. for two weeks by her physician; she committed suicide seven months after entry when she had a depressive illness unrelated to the neuropathy and was included in the analysis only until the time of her suicide.
TABLE I-COMPARISON OF TRIAL GROUPS AT ENTRY
or
half, paraesthesiae or loss of sensation in a third, and pain in a further third. Combinations of these symptoms were common. 3 patients presented with ataxic limbs and 1 with diplopia. The pattern of development of clinical signs after presentation was also similar (table II). Limb weakness was present in every case. Facial palsy, when present, was always bilateral but not necessarily symmetrical. Changes in disability grades at one, three, and twelve months were used in the analysis of results. The improvement in disability at one month after entry to the trial was less in the prednisolone group than the control group (table ill). The greatest difference was observed in those patients who entered the trial within a week of the onset of symptoms, where it was close to significance at the 5% level. This difference reached significance at this level three months after entry (table iv): those treated in
TABLE II—CLINICAL SIGNS IN TRIAL GROUPS
*
Results 22 patients had been entered to each treatment group between May, 1974 and April, 1977, when an interim
analysis showed less improvement in the prednisolonetreated group and entries to the trial ceased. 4 patients were withdrawn so that 21 prednisolone and 19 control patients were available for analysis. The groups were similar in their sex ratio, age, time from onset to entry into the trial, and severity at entry (table i). More than three-quarters of each group had raised cerebrospinal fluid protein concentrations while only 3 patients had a mild cerebrospinal fluid pleocytosis (never more than 10/p.l). The distribution of presenting symptoms was also similar in the groups, with weakness
Excluding suicide
TABLE III-IMPROVEMENT IN DISABILITY AT ONE MONTH
*Not significant (t=-1 95, and only values below -2-14 or above +2.14 are significant, at p
752 TABLE IV-IMPROVEMENT IN DISABILITY AT THREE MONTHS
Discussion Most
assessments
of the value of steroids in the Guil-
lain-Barre syndrome have been based either on literature reviews or on retrospective uncontrolled studies of patients. In one influential early literature review cases treated with steroids seemed to compare favourably with cases which had not been so treated: for instance, of the 68 treated cases 21 recovered within a month and only 2
*Significant (t=-2-35, P<0.05)
TABLE V-IMPROVEMENT "IN DISABILITY AT ONE YEAR
died,6 whereas
in 20 non-treated
recovery never 4 occurred earlier than ten weeks and died.’ Subsequent reviewers, who compared uncontrolled studies of patients treated with or without steroids, have reached conflicting conclusions. Some authors have stated that their impression is that steroids have shortened the duration of the diseaseor at least hastened the onset of improvement even if the eventual outcome remained unaffected.9 10 Others have not noted any significant effect from steroids .4 11 12 On the other hand there is a report that cases
of mild or moderate of the disease." duration the prolongs
steroid *Not significant
(t=—1-93).
treatment
cases
significantly
I
While
our trial was in progress a controlled trial of 100 units daily for ten days was reported from Kentucky.14 Severe cases requiring assisted ventilation were excluded, an important difference from our protocol. 9 patients received A.C.T.H. and 7 placebo: the mean duration of illness was 4’4±2-3 months (range 1-8) in 8 of the 9 A.C.T.H.-treated patients compared with 9’0±3’22 months (range 2-18) in the 7 placebo-treated patients. This difference was considered significant by the authors but 1 patient allocated to A.C.T.H., who discharged himself from hospital and died at home, was excluded from the analysis, which might be considered to bias the result. Furthermore, the duration of hospital stay was somewhat longer in the A.c.T.H. than the placebo group. This report did not seem to us to show a convincing effect of A.C.T.H.
A.C.T.H.
with prednisolone within a week of onset had improved on average 1-6 "grades" of disability less than those in the control group. By twelve months from entry the difference between the groups was smaller (table v) but the same trend existed, with less improvement in the prednisolone-treated patients who had entered the trial within one week from onset. This difference was close to significance at the 5% level. Although other analyses of the groups did not reveal significant differences the prednisolone patients generally fared worse than controls (table vi). Many patients in both groups were left with minor symptoms such as paraathesix or numbness or residual signs such as mild distal weakness or areflexia. 6 patients in the prednisolone group had considerable disability (grade 2 or more) after one year compared with 1 in the control group (excluding the patient who committed suicide). 1 patient in the prednisolone group died of acute respiratory failure with bulbar involvement and chest infection on the day after entry and a control patient died of bronchopneumonia late in the course of the disease (as mentioned above). 3 patients relapsed during the year of observation in the prednisolone group, but none in the control group. 1 patient in the prednisolone group had hypertension and papilloedema.
TABLE VI——OTHER COMPARISONS OF TRIAL GROUPS
The most interesting finding in our trial is that patients on prednisolone who were entered within a week of onset showed significantly less improvement by three months compared with those not on prednisolone. This suggests that prednisolone delays recovery. Disability at 12 months was greater in the prednisolone group. This finding is not significant with 2-tailed tests which have been used throughout. It does, however, imply that there are no detectable differences in effectiveness, and consequently that the drug has little to offer the clinician who is solely interested in making his patients significantly better. The present results should thus influence clinicians against the use of steroids in patients with acute
inflammatory polyneuropathy. Acute
*Excluding the one suicide during convalescence.
-
polyneuropathy may present in many ways and
embraces a number of different conditions of which a small minority might be steroid-sensitive. If this is so and the conclusions from the present trial are correct, this minority is more than counterbalanced by a larger number who are made worse by steroids. It is widely considered that relapsing cases are steroid-reponsivell and such cases might be expected to be steroidresponsive in their first attack. While this may be true, it is notable that the three patients who relapsed during the years of observation were those who had been treated with steroids initially. Furthermore, among 15
probably
753 series treated with cyclophospharelapses. Both figures are higher than the 3-10% incidence of relapse usually reported.2-4 Immunosuppression during the acute attack may actually encourage relapse. It has been said that A.c.T.H. is superior to prednisolone in the treatment of Guillain-Barre syndrome" but there appears to be no objective evidence to support this view and some hold an opposing view. Furthermore we do not know of other situations where A.C.T.H. has been clearly shown to be superior to prednisolone. In Bell’s palsy the reverse has been found: prednisolone 80 mg daily was more effective than A.c.T.H. 60 units daily in reducing the degree of denervation. 18 ’It seems unlikely that A.C.T.H. would succeed in Guillain-Barre syndrome where prednisolone has failed. According to the classical hypothesis acute inflammatory polyneuropathy is caused by a cell-mediated immune response to peripheral nerve antigens analogous to the mechanism demonstrated in experimental allergic encephalomyelitis and thought to operate in experimental allergic neuritis.t5 Steroids and immunosuppressive agents are effective in the treatment of experimental allergic encephalomyelitis’9 and they might be expected to be beneficial in experimental allergic neuritis although this has not been satisfactorily demonstrated. Our finding that prednisolone delays rather than enhances improvement is unexpected. Several factors could contribute to this outcome. Firstly, by the time clinical symptoms appear the immune response may already have reached a stage when it cannot be suppressed. Secondly, any reduction of oedema which occurs with steroids may be relatively unimportant in the peripheral nerve compared with the central neuraxis and steroids may actually slow repair. Thirdly, steroids could inhibit the suppressor T and B cell mechanisms which normally operate to limit the disease. Finally, if the disease is due to an immune response to a persistent viral infection, steroids may prolong the infection. Instances of acute inflammatory neuropathy occurring in patients with reduced cell-mediated immunity associated with Hodgkin’s disease,20 or with immunosuppression after renal transplantation 21 have been reported; cases associated with Cushing’s syndrome have also been
patients
in
a recent
mide16 there
were
4
recorded. 22
SYSTEMIC LUPUS ERYTHEMATOSUS AND LYMPHOMA A. A. DAWSON WILLIAM WALKER
J. A. GREEN
Departments of Medicine and Therapeutics, University of Aberdeen In 4 women lymphomas developed 2 months to 12 years after the onset of systemic lupus erythematosus. An association between the two diseases had previously been reported in 14 cases, in 6 of which the lymphoma either preceded or was diagnosed at the same time as the autoimmune disease. In systemic lupus erythematosus early biopsy of suspect lymph-nodes is recommended.
Summary
Introduction THE association between autoimmune disease and
malignancy is well known. Up to 20% of patients with dermatomyositis have an underlying carcinoma, and other non-organ-specific autoimmune diseases-Sjogren’s syndrome, polyarteritis nodosa, and rheumatoid arthritis-have been associated with lymphoid malignancies.2-4 14 cases of systemic lupus erythematosus (S.L.E.) and lymphoma are documented, the diagnoses being made simultaneously in 44-6 and lymphoma preceding the s.L.E. in 2.46 This paper describes 4 more patients with established S.L.E. in whom lymphoma subsequently developed. The diagnosis of S.L.E. was made on the basis of clinical and laboratory findings. The temporal course of the disease was defined by the terms s.L.E. onset for the time of the first significant symptom and zero time (to) for the time at which treatment for S.L.E. was decided.7 A definite histological diagnosis of Hodgkin’s disease, non-
DR HUGHES AND OTHERS: REFERENCES
2.
3.
Our results provide no grounds for the use of steroids in the management of acute inflammatory neuropathy since the prognosis is not improved, the rate of recovery is slowed, and the chance of relapse may be increased. Careful nursing, physiotherapy, and attention to the risk of ventilatory failure and circulatory complications remain the mainstay of treatment.
We thank Prof P. Armitage for his advice, Sister I. Dillon and staff of the Batten Unit, National Hospital for Nervous Diseases, for their assistance in holding the central register, the Medical Research Council for financial support and the physicians who entered patients to the trial (Dr J. N. Blau, Dr J. H. J. Durston, Dr A. G. Freeman, Dr W. W. Gooddy, Dr M. Green, Dr B. P. Harold, Dr P. K. P. Harvey, Dr A. P. Hopkins, Dr J. M. B. Hughes, Dr N. Jones, Dr R. S. Kocen, Dr N. J. Legg, Dr 1. C. K. Mackenzie, Prof. J. Marshall, Dr J. N. Milnes, Dr B. G. R. Nevilie, Dr B. G. Parsons Smith, Prof. R. J. Robinson, Dr R. W. Ross Russell, Dr P. Rudge, Dr G. Hillas Smith).
References at foot of next column
Kurland, L. T., Mulder, D. W. Neurology, Minneap., 1973, 23, 1269. Wiederholt, W. C., Mulder, D. W., Lambert, E. H. Proc. Mayo Clin. 1964, 39, 427. McLeod, J. G., Walsh, J. C., Prineas, J. W., Pollard, J. D. J. neurol. Sci.,
1. Lesser, R. P., Hauser, W. A.,
4. 5.
1976, 27, 145. Löffel, N. B., Rossi, L. N., Mumenthaler, M., Lütschg, J., Ludin, H. P. ibid. 1977, 33, 71. Dowling, P. C., Menonna, J. P., Cook, S. D. J. Am. med. Ass. 1977, 238, 317.
6. 7.
Jackson, R. H., Miller, H., Schapira, K. Br. med. J. 1957, i, 480. Gilpin, S. F., Moersch, F. P., Kernohan, J. W. Archs Neurol. Psychiat., Chicago, 1936, 35, 937. 8. Heller, G. L., De Jong, R. N. Archs Neurol., Chicago, 1963, 8, 179. 9. Bammer, H., Schaltenbrand, G. Münch. med. Wschr. 1965, 107, 1629. 10. Prineas, J. Acta neurol. scand. 1970, 46, (suppl. 44),1 11. Marshall, J. Brain, 1963, 86, 55. 12. Frick, E., Angstwurm, H. Munch. med. Wschr. 1968, 110, 1265. 13. Goodall, J. A. D., Kosmidis, J. C., Geddes, A. M. Lancet, 1974, i, 524. 14. Swick, H. M., McQuillen, M. P. Neurology, Minneap 1976, 26, 205. 15. Arnason, B. G. W. in Peripheral Neuropathy Vol II, (edited by P. J. Dyck, P. K. Thomas and E. H. Lambert); p. 1110. Philadelphia, 1975. 16. Rosen, A. D., Vastola, E. F. J. neurol. Sci. 1976, 30, 179. 17. Walton, J. N. in Brain’s Diseases of the Nervous System; p. 952. Oxford, 1977. 18. Taverner, D., Cohen, S. B., Hutchinson, B. C. Br. med. J. 1971, iv, 20. 19. Hughes, R. A. C., Leibowitz, S. L. in Immunology in Medicine (edited by E. J. Holborow and W. G. Reeves); p. 878, New York: 1977. 20. Lisak, R. P., Mitchell, M., Zweiman, B, Orrechio, E., Asbury, A. K. Ann Neurol. 1977, 1, 72. 21. Drachman, D. A., Paterson, P. Y., Berlin, B., Roguska, J. Neurology, Minneap. 1970, 20, 390. 22. Bresler, R., Johnson, C. T. Ann. intern. Med. 1959, 50, 1298.
reported objection to suture closure has been that
50-70% of patients will have further ulcer symptoms
or
complications.8 10 Sawyers and Herrington4 have performed proximal gastric vagotomy with suture closure for perforated duodenal ulcer in 21 patients, with no mortality and no recurrent symptoms over a 3-year period. This procedure may represent a compromise between operations which do not protect against recurrent ulceration and those which subject symptom-free patients to the hazards of gastric surgery. 30-47% of patients have been reported to be symptom-free after suture closure.36 11-13 In our study 48% of patients are still free of symptoms 3 months to 6 years after suture closure. Therefore, if emergency definitive surgery had been performed on these patients,, nearly half would have had an unnecessary and extended operation. Most surgeons agree that patients with a long history of symptoms should be considered for ulcer-curative surgery,36 12 14 and this has led to the classification of
peptic ulcers into "acute" and "chronic", depending
on
the duration of symptoms before perforation.8 Several workers use a period of 3 months as the dividing line. However, 28% of our patients had no symptoms before perforation.’o 11 16 Age, race, and sex do not influence prognosis significantly but the prevalence of perforation is higher in men than women. 13 17
Having determined the natural history of duodenalpatients treated with simple suture with Tutt and van Heerden, who sugclosure, agree choice of treatment, with long-term the first this as gest considerations exclude us from now follow-up.18 Ethiçal trial a clinical of emergency definitive treatundertaking ment. If primary definitive surgery is to be undertaken, then it should be reserved for patients with a long history of dyspepsia. Even this may be unjustified if H2-receptor antagonists are shown to heal chronic duodenal ulcers. 19-23 Requests for reprints should be addressed to S.N.J. ulcer
perforation
in
we
REFERENCES 1. MacLaren, R. Br. med. J. 1894, ii, 863. 2. Cooley, D. A., Jordan, G. L., Brockham, H.
L., De Bakey, M. E. Ann. Surg. 1955, 141, 840. 3. Jarrett, F., Donaldson, G. A. Am. J. Surg. 1972, 123, 406. 4. Sawyers, J. L., Herrington, J. L. Ann. Surg. 1977, 185, 656. 5. Jordan, G. L., Jr, De Bakey, M., Cooley, D. A. Am. J. Surg. 1963, 105, 396. 6. Jordan, G. L., Angel, R. T., De Bakey, M. D. Archs Surg., Chicago, 1965, 92, 449. 7. Booth, R. A. D., Williams, J. A. Br. J. Surg. 1971, 58, 42. 8. Illingworth, C. E. W., Scott, L. D. W., Jamieson, R. A. Br. med. J. 1946, i, 787. 9. Dean, A. C. B., Clark, C. G., Sinclair-Gieben, A. H. Gut, 1962, 2, 60. 10. Hofkin, G. A. Am. J. Surg. 1966, 111, 193. 11. McDonough, J. M., Foster, J. H. ibid. 1972, 123, 411. 12. Greco, R. S., Cahow, C. E. ibid. 1974, 127, 109. 13. Robbs, J. V., Moshal, M. G., Baker, L. W. S. Afr. J. Surg. 1977, 15, 39. 14. Maynard, A. de L., Froix, C. J. L., Oropeza, G. Archs Surg., Chicago, 1968, 97, 96. 15. Taylor, H., Warren, R. P. Lancet, 1956, i, 397. 16. Cassell, P. Gut, 1969,10, 572. 17. MacKay, C. Br. med. J. 1966, i, 701. 18. Tutt, G.O., van Heerden, J.A.S. Afr. J. Surg. 1976, 14, 1. 19. Blackwood, W. S., Maudgal, D. P., Pickard, R. G., Lawrence, D., Northfield, T. C. Lancet, 1976, ii, 174. 20. Gray, G. R., McKenzie, I., Smith, I. S., Crean, G. P., Gillespie, G. ibid. 1977, i, 4. 21. Gray, G. R., Smith, I. S., Mackenzie, I., Gillespie, G. Scott. med. J. 1977, 22, 293. 22. Semb, L. S., Berstad, A., Myren, J., Foss, J. C., Carlsen, E., Kruse-Jenson, A. in 2nd Int. Symp. Histamine H2-receptor Antagonist (1977) (edited by W. L. Burland and M. A. Simpkins); p. 248. Amsterdam, 1977. 23. Spence, R. W., Celestin, L. R., McCormick, D. A. Gut, 1977, 18, 420 (abstr.).
CONTROLLED TRIAL OF PREDNISOLONE IN ACUTE POLYNEUROPATHY R. A. C. HUGHES
Guy’s Hospital, London SE1
J. M. NEWSOM-DAVIS Royal Free Hospital, London NW3 and National Hospital for Nervous Diseases, London WC1
G. D. PERKIN
Charing Cross Hospital, London W6
J. M. PIERCE Department of Biomathematics, University of Oxford multicentre, randomised trial of prednisolone in acute polyneuropathy of undetermined ætiology (Guillain-Barré syndrome), 21 patients were treated with prednisolone (60 mg daily for one week, 40 mg daily for four days, and then 30 mg daily for three days) and 19 did not have steroid treatment. Patients were graded on a six-point scale by one of two neurologists who had no knowledge of the treatment schedule. Reassessment at one, three, and twelve months consistently showed greater improvement in the
Summary
In
a
control than the
prednisolone group but the only statistically significant result was in the improvement at three months among patients entered to the trial within a week of onset of illness. The 6 control patients had improved by 2·5±0·43 grades by three months from entry to the trial whereas the 10 prednisolone patients had only improved by 0·9±0·46 grades (P<0·05). There was 1 death related to the polyneuropathy in each group, and 1 suicide in a control patient during convalescence. 6 prednisolone patients were left with considerable disability compared with 1 control patient. There were 3 relapses in the prednisolone group, but none in the control group. The results indicate that steroid treatment is not beneficial and can be detrimental in acute neuropathy of undetermined ætiology.
Introduction THE usefulness of steroids in the Guillain-Barre syndrome has remained in doubt despite many retrospective studies. The disease (incidence about 1-7/100 0001) is sufficiently rare to make a single-centre study of steroids difficult. Acute polyneuropathy is not always a mild disease with a good prognosis: 10-23% require tracheostomy,2-5 only about 60% recover completely,2-5 7-22% are left significantly disabled,2-4 mortality 5 rates of 2.4-5% are still reported4’ and 3-10% relapse.2-4 We carried out a randomised multicentre trial of prednisolone versus no steroid treatment to establish whether steroids improve the outcome.
Methods All
with with
neurologists in London were invited to enter patients polyneuropathy of undetermined aetiology. Patients porphyria, collagen disease, carcinoma, severe diabetes,
acute
751 clinical evidence of current infective illnesses were excluded. Patients were included even if they had a normal protein concentration or pleocytosis in the cerebrospinal fluid. Patients who had been started on steroids before being seen by the admitting neurologist and those with contraindications to steroid treatment were excluded. Since it was decided during the course of the trial that those who had begun to improve before being seen should also be excluded 1 patient had to be withdrawn from each group: Approval of the trial was independently given by the ethical committee of two hospitals. The decision to ask for the consent of the patient was left to the physician in charge of each case. At entry patients were randomised to prednisolone or control treatment according to a central register. One group received prednisolone 15 mg four times daily for one week, 10 mg four times daily for four days, and 10 mg three times daily for three days after which the dose was reduced or continued at the discretion of the physician-incharge. The dose for children was reduced according to age: the starting dose was 10 mg four times daily at age 7, 10 mg three times daily at age 3, and 5 mg four times daily at age 1. Assessments were made within 24 h of entry by one of two assessors who had no knowledge of the treatment schedule, and again after one, two, three, and four weeks, two months, three months, six months, nine months, and one year. All cases were followed-up for one year although 1 boy who emigrated to Australia was assessed on the last three occasions by another neurologist. On each occasion a full neurological assessment was made and graded as follows: 0. Healthy. 1. Minor signs or symptoms of neuropathy but capable of manual work. 2. Able to walk without support of a stick but incapable of manual work. 3. Able to walk with a stick, appliance, or support. 4. Confined to bed or chairbound. 5. Requiring assisted ventilation. 6. Dead. 2 more patients were withdrawn from the control group, 1 because the diagnosis was later considered to be incorrect and 1 because prednisolone had been given before entry to the trial. 3 other patients were retained within the groups to which they had been allocated despite the following irregularities in the conduct of the trial. 1 mild case was allocated to the prednisolone group on a Friday but the drug was not given, and since, by the Monday, the patient had improved treatment was considered unnecessary. 1 control patient had a persistent demyelinating motor polyneuropathy and a grossly raised E.S.R. After four months he was started on steroids but soon died of bronchopneumonia. A third control patient deteriorated and thirteen days after entry was given A.c.T.H. for two weeks by her physician; she committed suicide seven months after entry when she had a depressive illness unrelated to the neuropathy and was included in the analysis only until the time of her suicide.
TABLE I-COMPARISON OF TRIAL GROUPS AT ENTRY
or
half, paraesthesiae or loss of sensation in a third, and pain in a further third. Combinations of these symptoms were common. 3 patients presented with ataxic limbs and 1 with diplopia. The pattern of development of clinical signs after presentation was also similar (table II). Limb weakness was present in every case. Facial palsy, when present, was always bilateral but not necessarily symmetrical. Changes in disability grades at one, three, and twelve months were used in the analysis of results. The improvement in disability at one month after entry to the trial was less in the prednisolone group than the control group (table ill). The greatest difference was observed in those patients who entered the trial within a week of the onset of symptoms, where it was close to significance at the 5% level. This difference reached significance at this level three months after entry (table iv): those treated in
TABLE II—CLINICAL SIGNS IN TRIAL GROUPS
*
Results 22 patients had been entered to each treatment group between May, 1974 and April, 1977, when an interim
analysis showed less improvement in the prednisolonetreated group and entries to the trial ceased. 4 patients were withdrawn so that 21 prednisolone and 19 control patients were available for analysis. The groups were similar in their sex ratio, age, time from onset to entry into the trial, and severity at entry (table i). More than three-quarters of each group had raised cerebrospinal fluid protein concentrations while only 3 patients had a mild cerebrospinal fluid pleocytosis (never more than 10/p.l). The distribution of presenting symptoms was also similar in the groups, with weakness
Excluding suicide
TABLE III-IMPROVEMENT IN DISABILITY AT ONE MONTH
*Not significant (t=-1 95, and only values below -2-14 or above +2.14 are significant, at p
752 TABLE IV-IMPROVEMENT IN DISABILITY AT THREE MONTHS
Discussion Most
assessments
of the value of steroids in the Guil-
lain-Barre syndrome have been based either on literature reviews or on retrospective uncontrolled studies of patients. In one influential early literature review cases treated with steroids seemed to compare favourably with cases which had not been so treated: for instance, of the 68 treated cases 21 recovered within a month and only 2
*Significant (t=-2-35, P<0.05)
TABLE V-IMPROVEMENT "IN DISABILITY AT ONE YEAR
died,6 whereas
in 20 non-treated
recovery never 4 occurred earlier than ten weeks and died.’ Subsequent reviewers, who compared uncontrolled studies of patients treated with or without steroids, have reached conflicting conclusions. Some authors have stated that their impression is that steroids have shortened the duration of the diseaseor at least hastened the onset of improvement even if the eventual outcome remained unaffected.9 10 Others have not noted any significant effect from steroids .4 11 12 On the other hand there is a report that cases
of mild or moderate of the disease." duration the prolongs
steroid *Not significant
(t=—1-93).
treatment
cases
significantly
I
While
our trial was in progress a controlled trial of 100 units daily for ten days was reported from Kentucky.14 Severe cases requiring assisted ventilation were excluded, an important difference from our protocol. 9 patients received A.C.T.H. and 7 placebo: the mean duration of illness was 4’4±2-3 months (range 1-8) in 8 of the 9 A.C.T.H.-treated patients compared with 9’0±3’22 months (range 2-18) in the 7 placebo-treated patients. This difference was considered significant by the authors but 1 patient allocated to A.C.T.H., who discharged himself from hospital and died at home, was excluded from the analysis, which might be considered to bias the result. Furthermore, the duration of hospital stay was somewhat longer in the A.c.T.H. than the placebo group. This report did not seem to us to show a convincing effect of A.C.T.H.
A.C.T.H.
with prednisolone within a week of onset had improved on average 1-6 "grades" of disability less than those in the control group. By twelve months from entry the difference between the groups was smaller (table v) but the same trend existed, with less improvement in the prednisolone-treated patients who had entered the trial within one week from onset. This difference was close to significance at the 5% level. Although other analyses of the groups did not reveal significant differences the prednisolone patients generally fared worse than controls (table vi). Many patients in both groups were left with minor symptoms such as paraathesix or numbness or residual signs such as mild distal weakness or areflexia. 6 patients in the prednisolone group had considerable disability (grade 2 or more) after one year compared with 1 in the control group (excluding the patient who committed suicide). 1 patient in the prednisolone group died of acute respiratory failure with bulbar involvement and chest infection on the day after entry and a control patient died of bronchopneumonia late in the course of the disease (as mentioned above). 3 patients relapsed during the year of observation in the prednisolone group, but none in the control group. 1 patient in the prednisolone group had hypertension and papilloedema.
TABLE VI——OTHER COMPARISONS OF TRIAL GROUPS
The most interesting finding in our trial is that patients on prednisolone who were entered within a week of onset showed significantly less improvement by three months compared with those not on prednisolone. This suggests that prednisolone delays recovery. Disability at 12 months was greater in the prednisolone group. This finding is not significant with 2-tailed tests which have been used throughout. It does, however, imply that there are no detectable differences in effectiveness, and consequently that the drug has little to offer the clinician who is solely interested in making his patients significantly better. The present results should thus influence clinicians against the use of steroids in patients with acute
inflammatory polyneuropathy. Acute
*Excluding the one suicide during convalescence.
-
polyneuropathy may present in many ways and
embraces a number of different conditions of which a small minority might be steroid-sensitive. If this is so and the conclusions from the present trial are correct, this minority is more than counterbalanced by a larger number who are made worse by steroids. It is widely considered that relapsing cases are steroid-reponsivell and such cases might be expected to be steroidresponsive in their first attack. While this may be true, it is notable that the three patients who relapsed during the years of observation were those who had been treated with steroids initially. Furthermore, among 15
probably
753 series treated with cyclophospharelapses. Both figures are higher than the 3-10% incidence of relapse usually reported.2-4 Immunosuppression during the acute attack may actually encourage relapse. It has been said that A.c.T.H. is superior to prednisolone in the treatment of Guillain-Barre syndrome" but there appears to be no objective evidence to support this view and some hold an opposing view. Furthermore we do not know of other situations where A.C.T.H. has been clearly shown to be superior to prednisolone. In Bell’s palsy the reverse has been found: prednisolone 80 mg daily was more effective than A.c.T.H. 60 units daily in reducing the degree of denervation. 18 ’It seems unlikely that A.C.T.H. would succeed in Guillain-Barre syndrome where prednisolone has failed. According to the classical hypothesis acute inflammatory polyneuropathy is caused by a cell-mediated immune response to peripheral nerve antigens analogous to the mechanism demonstrated in experimental allergic encephalomyelitis and thought to operate in experimental allergic neuritis.t5 Steroids and immunosuppressive agents are effective in the treatment of experimental allergic encephalomyelitis’9 and they might be expected to be beneficial in experimental allergic neuritis although this has not been satisfactorily demonstrated. Our finding that prednisolone delays rather than enhances improvement is unexpected. Several factors could contribute to this outcome. Firstly, by the time clinical symptoms appear the immune response may already have reached a stage when it cannot be suppressed. Secondly, any reduction of oedema which occurs with steroids may be relatively unimportant in the peripheral nerve compared with the central neuraxis and steroids may actually slow repair. Thirdly, steroids could inhibit the suppressor T and B cell mechanisms which normally operate to limit the disease. Finally, if the disease is due to an immune response to a persistent viral infection, steroids may prolong the infection. Instances of acute inflammatory neuropathy occurring in patients with reduced cell-mediated immunity associated with Hodgkin’s disease,20 or with immunosuppression after renal transplantation 21 have been reported; cases associated with Cushing’s syndrome have also been
patients
in
a recent
mide16 there
were
4
recorded. 22
SYSTEMIC LUPUS ERYTHEMATOSUS AND LYMPHOMA A. A. DAWSON WILLIAM WALKER
J. A. GREEN
Departments of Medicine and Therapeutics, University of Aberdeen In 4 women lymphomas developed 2 months to 12 years after the onset of systemic lupus erythematosus. An association between the two diseases had previously been reported in 14 cases, in 6 of which the lymphoma either preceded or was diagnosed at the same time as the autoimmune disease. In systemic lupus erythematosus early biopsy of suspect lymph-nodes is recommended.
Summary
Introduction THE association between autoimmune disease and
malignancy is well known. Up to 20% of patients with dermatomyositis have an underlying carcinoma, and other non-organ-specific autoimmune diseases-Sjogren’s syndrome, polyarteritis nodosa, and rheumatoid arthritis-have been associated with lymphoid malignancies.2-4 14 cases of systemic lupus erythematosus (S.L.E.) and lymphoma are documented, the diagnoses being made simultaneously in 44-6 and lymphoma preceding the s.L.E. in 2.46 This paper describes 4 more patients with established S.L.E. in whom lymphoma subsequently developed. The diagnosis of S.L.E. was made on the basis of clinical and laboratory findings. The temporal course of the disease was defined by the terms s.L.E. onset for the time of the first significant symptom and zero time (to) for the time at which treatment for S.L.E. was decided.7 A definite histological diagnosis of Hodgkin’s disease, non-
DR HUGHES AND OTHERS: REFERENCES
2.
3.
Our results provide no grounds for the use of steroids in the management of acute inflammatory neuropathy since the prognosis is not improved, the rate of recovery is slowed, and the chance of relapse may be increased. Careful nursing, physiotherapy, and attention to the risk of ventilatory failure and circulatory complications remain the mainstay of treatment.
We thank Prof P. Armitage for his advice, Sister I. Dillon and staff of the Batten Unit, National Hospital for Nervous Diseases, for their assistance in holding the central register, the Medical Research Council for financial support and the physicians who entered patients to the trial (Dr J. N. Blau, Dr J. H. J. Durston, Dr A. G. Freeman, Dr W. W. Gooddy, Dr M. Green, Dr B. P. Harold, Dr P. K. P. Harvey, Dr A. P. Hopkins, Dr J. M. B. Hughes, Dr N. Jones, Dr R. S. Kocen, Dr N. J. Legg, Dr 1. C. K. Mackenzie, Prof. J. Marshall, Dr J. N. Milnes, Dr B. G. R. Nevilie, Dr B. G. Parsons Smith, Prof. R. J. Robinson, Dr R. W. Ross Russell, Dr P. Rudge, Dr G. Hillas Smith).
References at foot of next column
Kurland, L. T., Mulder, D. W. Neurology, Minneap., 1973, 23, 1269. Wiederholt, W. C., Mulder, D. W., Lambert, E. H. Proc. Mayo Clin. 1964, 39, 427. McLeod, J. G., Walsh, J. C., Prineas, J. W., Pollard, J. D. J. neurol. Sci.,
1. Lesser, R. P., Hauser, W. A.,
4. 5.
1976, 27, 145. Löffel, N. B., Rossi, L. N., Mumenthaler, M., Lütschg, J., Ludin, H. P. ibid. 1977, 33, 71. Dowling, P. C., Menonna, J. P., Cook, S. D. J. Am. med. Ass. 1977, 238, 317.
6. 7.
Jackson, R. H., Miller, H., Schapira, K. Br. med. J. 1957, i, 480. Gilpin, S. F., Moersch, F. P., Kernohan, J. W. Archs Neurol. Psychiat., Chicago, 1936, 35, 937. 8. Heller, G. L., De Jong, R. N. Archs Neurol., Chicago, 1963, 8, 179. 9. Bammer, H., Schaltenbrand, G. Münch. med. Wschr. 1965, 107, 1629. 10. Prineas, J. Acta neurol. scand. 1970, 46, (suppl. 44),1 11. Marshall, J. Brain, 1963, 86, 55. 12. Frick, E., Angstwurm, H. Munch. med. Wschr. 1968, 110, 1265. 13. Goodall, J. A. D., Kosmidis, J. C., Geddes, A. M. Lancet, 1974, i, 524. 14. Swick, H. M., McQuillen, M. P. Neurology, Minneap 1976, 26, 205. 15. Arnason, B. G. W. in Peripheral Neuropathy Vol II, (edited by P. J. Dyck, P. K. Thomas and E. H. Lambert); p. 1110. Philadelphia, 1975. 16. Rosen, A. D., Vastola, E. F. J. neurol. Sci. 1976, 30, 179. 17. Walton, J. N. in Brain’s Diseases of the Nervous System; p. 952. Oxford, 1977. 18. Taverner, D., Cohen, S. B., Hutchinson, B. C. Br. med. J. 1971, iv, 20. 19. Hughes, R. A. C., Leibowitz, S. L. in Immunology in Medicine (edited by E. J. Holborow and W. G. Reeves); p. 878, New York: 1977. 20. Lisak, R. P., Mitchell, M., Zweiman, B, Orrechio, E., Asbury, A. K. Ann Neurol. 1977, 1, 72. 21. Drachman, D. A., Paterson, P. Y., Berlin, B., Roguska, J. Neurology, Minneap. 1970, 20, 390. 22. Bresler, R., Johnson, C. T. Ann. intern. Med. 1959, 50, 1298.