- Email: [email protected]
CONTROLLED TRIAL OF CYSTEAMINE IN TREATMENT OF ACUTE PARACETAMOL (ACETAMINOPHEN) POISONING
111 RESULTS USING ANTISERA DETECTING B-LYMPHOCYTE ANTIGENS IN NORMAL INDIVIDUALS AND PATIENTS WITH G.S.E. AND D.H. aI
I
linkage disequilibrium may be necessary to the organism in that the gene products have complementary functions. In this regard we have already presented evidence that the HLA-B8 gene is not merely a marker gene but seems to contribute a gene product (surface antigen) that participates in the disease process.’" If this Such
is the case,
suggest that the HLA gene and the B-lymphocyte antigen gene act in a complementary fashion at some critical step in the pathogenesis ofD.H. and ..-.
non
l..___L
i
y
i
we
G.S.E.
The methods used in this study can be easily applied the investigation of other diseases where associations with HLA antigens have been identified. In these cases, unique B-lymphocyte antigens may be found which will lead to an understanding of underlying disease processes. Second, the antigens detected in this report can be used to develop a method of identifying patients with gastrointestinal gluten sensitivity. Such a method will not depend on interpretation of jejunal-biopsy specimens and therefore will lead more easily to the correct diag-
to
nineteen D.H. patients, while antiserum W-1 reacted with fifteen of sixteen G.s.E. patients and fifteen of fifteen D.H. patients. Five other antisera obtained from mothers with children having G.S.E. or D.H. showed no consistent pattern of B-lymphocyte reactivity. Antisera obtained from Amish or randomly selected donors which have been shown to react with B lymphocytes obtained from normal persons carrying the HLA-B8 antigen, showed a similar pattern of reactivity with the lymphocytes in the disease population studied-i.e., they tended to react with HLA-B8-positive patients.
noses.
Requests for reprints should be addressed to D. L. M., Immunology Branch, National Cancer Institute, Building 10, Room 4B08, Bethesda, Maryland 20014, U.S.A.
Discussion The studies show that
antisera reacted with B-lymphocyte surface antigens which are very strongly associated with G.S.E. and D.H. and which are lacking in the normal population so far studied. However, before one can say that these antigens are unique to G.s.E. and D.H. many additional normal individuals must be studied to determine whether these antigens occur in the normal population in low frequency. In addition, patients with other diseases must be studied to determine whether the antigen occurs in other pathological states. The relation of the cell-surface antigen(s) reported here with the pathophysiology of G.S.E. and D.H. is unknown. Lymphocyte surface antigens in mice, termed Ia antigens, are known to influence the ability of mice to respond to certain antigens,7 and humoral enhancing and suppressing factors produced by T lymphocytes which are active on B lymphocytes have la specificity.8 9 The antigens detected in this study may be human analogues of the murine Ia antigens. If this is so the B-lymphocyte antigen associated with G.S.E. and D.H. may represent the gene product of an abnormal Ir gene in man, which in turn results in an immune response to wheat protein and immunologically mediated gut epithelial cell toxicity. A second possibility, not necessarily mutually exclusive from the first possibility, is that the B-lymphocyte antigen identified here is a receptor or binding site for wheat proteins. Such sites may exist on the surface of gut epithelial cells in patients with G.s.E. and D.H. and lead to immune reactions to wheat protein and, in the case of D.H., to the subsequent deposition of IgA immune complexes in the skin. To study these possibilities, inhibition of B-lymphocyte cytotoxicity by wheat protein is currently being investigated. The association of both HLA-B8 and the B-lymphocyte antigen identified here with G.S.E. and D.H. is of some interest. If the gene controlling the B-lymphocyte antigen proves to be part of the histocompatability gene complex, as in the case of previously studied B-lymphocyte antigens,3then the occurrence of these antigens in G.S.E./D.H. may be due to linkage disequilibrium-i.e., cross-over suppression between the two involved genes.
REFERENCES
two
Rogentine, G. N., Strober, W. J. clin. Invest. 1972, 51, 1602. 2. Katz, S. I., Falchuk, Z. M., Dahl, M., Rogentine, G. N., Strober, W. ibid. 1.
Falchuk,
A. M.,
p. 2977. 3. Mann, D. L., Abelson, L., Henkart, P., Harris, S. D., Amos, B. D. Proc. natn. Acad. Sci. U.S.A. (in the press). 4. Alexander, E., Henkart, P. A. Unpublished. 5. Mann, D. L., Abelson, L., Harris, S., Amos, B. D. Nature, (in the press). 6. Abelson, L. D., Henkart, P. A., Mann, D. L. in Manual of Clinical Immunology (edited by E. Yunis); (in the press). 7. Sachs, D. H., Dickler, H. B. Transplant. Rev. 1975, 23, 159. 8. Munro, A. J. Taussig, M. J. Nature, 1975, 256, 103. 9. Takemori, T., Tada, T. J. exp. Med. 1975, 142, 1241. 10. Nelson, D. L., Falchuk, Z. M., Kasarda, D., Strober, W. Clin. Res. 1975, 23, 254abstr.
CONTROLLED TRIAL OF CYSTEAMINE IN TREATMENT OF ACUTE PARACETAMOL (ACETAMINOPHEN) POISONING ADRIAN P. DOUGLAS ADRIAN N. HAMLYN OLIVER JAMES
Gastroenterology Group, Department of Medicine, Royal Victoria Infirmary, Newcastle on Tyne NE1 4LP A randomised controlled trial of the use of intravenous cysteamine in the treatment of severe paracetamol poisoning has been performed. Thirty-eight patients presenting 3-17 h after ingestion were admitted to the trial; of these eighteen received cysteamine. Two patients died from hepatic failure, one in each treatment group. Analysis of the series as a whole showed no advantage of cysteamine in preventing biochemical abnormalities of liver function except for aspartate aminotransferase and serum ferritin levels, which were significantly less after cysteamine therapy. Separate analysis of the patients treated within 9 h of paracetamol ingestion and of those treated 9-17 h after paracetamol ingestion similarly showed no definite advantage of cysteamine. Histological evidence of liver damage showed a possible beneficial effect of cys-
Summary
112
teamine.
Cysteamine therapy did pancreatic damage.
not
preventrenal
or
-
Introduction THE North East Regional Survey, centred at Newcastle upon Tyne, has been concerned with aspects of paracetamol (acetaminophen) poisoning, and an average of 15 such patients are admitted each month. Over 1000 patients with paracetamol overdose are admitted to hospitals in England and Wales every year, and about 3% die.1 Many different measures have been advocated for their management, including corticosteroids,2 haemodialysis,3 forced diuresis,2 orally administered cholestyramine or activated charcoal, and charcoal-column hæmoperfusion.5 None of these measures have been proved to prevent liver damage in severe paracetamol poisoning. In 1974 Prescott et al. reported6 the results of intravenous dosage with cysteamine in patients after paracetamol poisoning with apparent favourable results. The work of Mitchell and his colleagues,7-13 has suggested that paracetamol is toxic because of its oxidised metabolites. In therapeutic doses paracetamol is safely conjugated with sulphate and glucuronic acid in the liver but, as the dose is increased, the metabolites are neutralised by glutathione, which contains a sulphydryl (SH) group, until all the glutathione is used up. Then other SH groups in the cell (such as are present in many intracellular membranes and enzymes) are consumed, and damage ensues. This hypothesis provides the rationale for cysteamine treatment. Cysteamine (&bgr;-mercaptoethylamine) is a nucleophilic SH-containing compound which reduces the hepatotoxicity of paracetamol in laboratory animals,12 13 perhaps by supplying SH groups or possibly by slowing the intracellular metabolism of paracetamol.6 The clinical course of paracetamol poisoning varies6 14 15 so assessment of the effect of treatment is difficult. Previous workers have pointed out that "many forms of therapy... have been advocated without control data."6 We have carried out a controlled trial of cysteamine in the management of paracetamol poisoning.
Patients and Methods Patients AU patients who were admitted to our hospital within 17 h of paracetamol ingestion, who were HbsAg negative with no history of pre-existing liver disease, and whose plasma-paracetamol was above the line indicated in fig. 1 were included in the trial. The line shown. in fig. 1 is very similar to that of Prescott et al.6 and is constructed from our own data on more than 250 patients over the past 2 years.’6 Patients whose presenting plasma-paracetamol fell above the line had a substantial risk of death or serious liver damage. 38 patients were admitted to the trial. 18 were assigned to receive cysteamine and 20 patients did not receive cysteamine. The patients had ingested between 7 and 85 g of paracetamol; no correlation was observed between the stated dose and plasma-paracetamol concentration.
Design of Trial The trial was planned to answer four questions. Does cysteamine therapy affect mortality after paracetamol overdose? Does cysteamine prevent severe, non-fatal, liver damage or other injury caused by paracetamol overdose? In view of the theoretical considerations governing the use of cysteamine, is it important to give cysteamine early after the overdose? And, is age an important factor in determining mortality or mor-
bidity after paracetamol overdose as it is in other forms of acute hepatic failure17? Patients were assigned on admission to one of four groups: (I) young (less than 30 years) and early (less than 9 h); (n) young and late; (III) old and early; and (tv) old and late. Patients were randomly allocated to receive cysteamine or not using a table of random numbers. Using a table of random numbers, it was possible that one or other group could become weighted with cases receiving only one of the treatment regimens. Adjustment was made therefore, so that every six patients in each group included three who had received cysteamine.
Methods Liver function
was assessed daily by measurement, on fast. ing blood-samples, of bilirubin, aspartate aminotransferase (A.S.T., normal range 4-20 I.U.!1), alkaline phosphatase, albumin and total protein, prothrombin ratio (Quick one-stage method), total bile acidsl6 (normal < 14 µmol/l), and ferritin" (normal > 240 mg/1). Liver biopsy was performed with a Menghini needle on the fourth day after admission, or as soon thereafter as the prothrombin-time and patient’s condition permitted. Histopathological changes were graded severe, moderate, or mild. 16 Renal function was assessed daily by measurement of urine volume, plasma urea, electrolytes, and creatinine, and, where appropriate, creatinine clearance and urinary urea and
sodium. Pancreatic
damage was assessed by daily estimation of serum-amylase (normal<300 units/l). Plasma-paracetamol was measured spectrophotometrically. 19 Treatment All patients received a standard regimen.16 Briefly this consisted of intravenous 5% dextrose, 2-3 litres daily, with added vitamins and potassium as necessary. This was maintained for 72 h or until liver function was satisfactory, and for the fist 48 h nothing was given by mouth. If liver function became moderately or severely impaired (A.S.T. > 100 units/1, proof control) the infusion was continued thrombin-time <50% and lactulose, 10 ml 6-hourly was given. If hepatic encephalopathy developed, this was treated conventionally. Two patients who reached grade iv hepatic coma were treated with charcoal-column hoemoperfusion; both died. Cysteamine was given as described by Prescott et al." except that it was dissolved in 5% dextrose and injected, or added to the 5% dextrose infusion, using a Millipore filter attached to a syringe. 2-hourly estimations of capillary-blood sugar were made by a fingerprick method (Ames ’Dextrostix’), and if the bloodsugar dropped below 2.5 mmol/1 (4) mg/dl), 100 ml of 5% dextrose was
injected intravenously.
Statistical Analysis Where appropriate the data rank-sum test.
were
analysed using Wilcoxon’s
Results
Only two of our possible groups--young and early (six cysteamine, seven no cysteamine) and young and late (eight cysteamine, ten no cysteamine) contained sufficient numbers to be analysed separately from the entire trial sample. This is a reflection of the age distribution of patients presenting with paracetamol self-poisoning. In addition we analysed all fourteen early patients (young and old) separately from the nineteen late patients. Paracetamol "Index" Using the data in fig. 1 a paracetamol "index" was derived for each patient. This was the concentration, in
113 Minimal Prothrombin
Maximal Bilirubin
Received Cysteamine No Cysteamine A Severe histological change
0
t Died
Cysteamine No Cysteamine Time in hours after ingestion of Paracetamol
Fig. 1-Plasma-paracetamol levels with paracetamol poisoning
on
admission to
hospital in patients
mg/1, by
which the patient exceeded the theoretical "safe" upper limit, indicated by the line, at the time when the plasma-paracetamol was measured. The mean value for the 18 cysteamine patients was 72 mg/1 and the mean value for the 20 non-cysteamine patients was 98 mg/l. The difference between these values is significant (P<0.02) and indicates that the non-cysteamine group were potentially at greater risk from paracetamol poisoning. This difference arose by chance because of the design of the trial. The mean index for the cysteamine-treated young and early patients was 43 mg/1 and the mean index for the young and early non-cysteamine patients was 138 mg/1 (P<001). There was no significant difference in these values in the young and late patients (75 mg/1 cystea-
mine, 67 mg/l non-cysteamine). Deaths Two
patients died of fulminant hepatic failure. One,
Fig. 2—Distribution of maximum bilirubin and minimum prothrombin in patients with paracetamol poisoning. t Denotes patients who died.
aged 24, had not received cysteamine (admitted to hospital 6.5h after a stated ingestion of 75 g of paracetamol) and the other, aged 28, had received cysteamine (admitted to hospital 9.5h after ingesting 15 g of paracetamol). Liver Function
These data are summarised in the table. The distribution of maximum bilirubin and minimum prothrombin (both achieved on the third or fourth day) in our patients is shown in fig. 2. There is no difference between the two groups. Similarly, there is no difference when the young and late and early and late groups are analysed separately. However, the difference for prothrombin in the young early group just reaches statistical significance (P=0.05). Fig. 3 shows the results for three tests of hepatocellular damage. No significant differences are seen between cysteamine and non-cysteamine treated patients
LIVER-FUNCTION TESTS IN PATIENTS WITH ACUTE PARACETAMOL POISONING I
Including one patient who died. Differences between cysteamine and
I
.
no
Cysteamine No Cysteamine
cysteamine not significant except for ferritin and A.S.T. (P<0.01
f<0.05).
(MEAN VALUES)
114 there was a significantly lower paracetamol index in those receiving cysteamine, all six patients on cysteamine had moderate (grade n) changes, whereas in those who did not receive cysteamine and who had a greatly higher paracetamol index, five out of seven showed
grade-m changes. Renal Function Four of
our
patients developed histologically
proven
tubular necrosis (one of whom died in hepatic failure). Two of these had received cysteamine and the other two had not. acute
.
Cysteamine No Cysteamine Cysteamine No Cysteamine
Cysteamine No Cysteamine
3-Distribution of maximum serum-A.S.T., maximal serum-bileacids and maximum serum-ferritin in patients with paracetamol poi-
Fig.
soning.
with respect to peak serum concentrations of bile acids. However, there are striking and significant differences (P<0.01) between the two sets of patients with regard to peak serum-A.s.T. In the entire series, and also in both early and late patients, the serum-A.S.T. is lower in the cysteamine-treated than in the non-cysteamine patients. A similar difference was seen with regard to the serumferritin in the entire series (P<0.01), and in the early not but in the late groups (P<0.05 ) groups. Liver Histology
Fig. 4 illustrates the distribution of the three grades of histological abnormality. More patients had severe (grade m) abnormalities in the non-cysteamine group than amongst those who received cysteamine. Fig. 1 shows that most grade-in changes were in patients with high plasma-paracetamol concentrations and that when patients receiving or not receiving cysteamine had comparable paracetamol levels the severity of histological change was similar. In the young early group, in which
Serum-amylase Abnormally high serum-amylases were recorded (often within 48 h of paracetamol ingestion) in six of the cysteamine-treated patients (430-780 units/1) and in three of the non-cysteamine-treated patients (370-440
units/1). Toxic Effects of Cysteamine Our experience was similar to that of Prescott et al.6 All our patients rapidly became nauseated and repeatedly vomited and retched for the duration of cysteamine therapy and up to 36 h afterwards. In addition they lay in bed looking utterly dejected and miserable and their breath stank of mercaptans. Although the patients in the non-cysteamine group, with severe liver damage, commonly experienced some nausea, they did not have the general appearance of the cysteamine group. We have not established any long-term toxic effects of cysteamine. Liver function and histology of patients who received cysteamine were identical at 3 months’ follow-up with those of survivors who did not receive cys-
teamine.2o Discussion The importance of preventing death from paracetamol-induced hepatic necrosis lies not least in the fact that most of these patients are young and quickly regret their action.However, because only 3-5% of such overdosages prove fatal, a very large trial is needed to demon strate that any new treatment saves lives; and our trial is on too small a scale to permit any conclusions about the effect of cysteamine on mortality-rates. One patient on cysteamine and one not on cysteamine died. Prescott et al. recorded no deaths in eight treated patients and two deaths in eleven previous, and not sequential, noncysteamine-treated cases. This difference is not statistic-
ally significant. Cysteamine did
prevent raised transaminase and ferritin levels, but is this truly indicative of an effect in preventing hepatic centrilobular necrosis? In the rat, peak transaminase values seem to correlate well with the extent of paracetamol-induced liver necrosis21 but in man this correlation is less exact,12 18 22 and an increase in serum-bilirubin and prothrombin ratio may be a better prognostic index.’ In our patients, serum-bilirubin and prothrombin ratio were much the same in the cysteamine and non-cysteamine patients. Nor did differences emerge when another sensitive indicator of No
Cysteamine
Cysteamine
Fig. 4-Distribution of grades histological abnormality paracetamol poisoning.
in
patients
with
hepatocellular dysfunction, serum-bile-acids, was considered. However, changes in serum-ferritin, a newly recognised indicator of acute hepatocellular
115
damage, 18 23
do accord with the alterations of Further evidence on this point is provided serum-A.S.T. data. Cysteamine therapy seemed to by our histological severe (grade in) changes. fewer with be associated this difference However, may just as well be due to the fact that, by chance, the cysteamine-treated had lower plasma-paracetamol concentrations. Our observations certainly accord with the hypothesis generated by the work of Mitchell and his colleagues,H3 but administration of cysteamine, a toxic substance, cannot yet be said to be justifiable routinely for all patients with paracetamol poisoning. Our data suggest that if cysteamine is used at all, it must be given within 9 h of paracetamol ingestion. Further work with other SH donors certainly seems worthwhile, however. The incorporation of methionine into tablets containing paracetamol might be helpful,24 but methionine given after a paracetamol overdose may be harmful since it has a comagenic effect in liver disease.25 The extrahepatic effects of paracetamol are often forgotten in the face of the potential hepatoxicity. Acute renal failure is a well-recognised feature of paracetamol poisoning,26-29 and we have no evidence that cysteamine prevents this. Acute pancreatitis (not complicating renal failure) also occurred in our patients and was as often associated with cysteamine as with no cysteamine. This series is too small to permit conclusions about the treatment of paracetamol poisoning. It is important to bear in mind that we did not carry out a controlled trial of cysteamine versus nothing, but rather studied patients receiving dextrose, some of whom also received cysteamine. In the past 15 months we have had six deaths in hepatic failure due to paracetamol poisoning: two in this trial and four who had been transferred to our hospital more than 72 h after drug ingestion during which time they had received no specific therapy. We recommend that such patients should be treated with intravenous 5% dextrose and no food by mouth for the first 48 h at least. This may or may not be beneficial, but it seems to do no harm. We thank our junior medical colleagues for their tireless help; the physicians of the region for referring cases; and Prof. D. N. S. Kerr
for his encouragement; Dr P. A. Smith, Dr A. J. Fleetwood, Dr M. J. Watson, and others who estimated the plasma-paracetamol levels; Mrs L. Pulman for the serum-bile-acid estimations; Dr E. J. Eastham and Mr I. Bell for the ferritin estimations; the Departments of Clinical Biochemistry and Hxmatology for their help; and Dr M. Lesna and Dr A. J. Watson for the liver histology.
Requests for reprints should be addressed to
0.
J.
REFERENCES 1.
Clark, R., Thompson, R. P. M., Borkrakchanyurat, V., Widdop, B., Davidson, A. R., R., Williams, R. Lancet, 1973, i, 66. 2. MacLean, D., Peters, T. J., Brown, R. A. C., McCathie, M., Baines, G. E., Robertson, P. G. C. ibid. 1968, ii, 849. 3. Fand, N. R., Glynn, J. P., Kerr, D. N. S. ibid. 1972, iii, 396. 4. Dordom, B., Wilson, R. S., Thompson, R. P. H., Williams, R. Br. med. j. 1973, iii, 86. 5. Wilson, R. S., Winch, J., Thompson, R. P. M., Williams, R. Lancet, 1973, i, 77. 6.
Prescott, L. F., Newton, R. W., Swainson, C. P., Wright, N., Forrest, A. R. W., Matthew, H. ibid. 1974, i, 588. 7. Mitchell, J. R., Jollow, D. J., Porter, W. Z., Davis, D. C., Gillette, J. R., Brodie, B. B. J. Pharmac. exp. Ther. 1973, 187, 185. 8. Jollow, D. J., Mitchell, J. R., Potter, W. Z., Davis, D. C., Gillette, J. R., Brodie, B. B. ibid. 1973, 187, 195. 9. Potter, W. Z., Davis, D. C., Mitchell, J. R., Jollow, D. J., Gillette, J. R., Brodie, B. B. ibid. 1973, 187, 203. 10 Mitchell, J. R., Jollow, D. J., Potter, W. Z., Gillette, J. R., Brodie, B. B. ibid. 1973, 187, 211.
ATOPIC DISORDERS AND ADULT CŒLIAC DISEASE R.
H. J. F. HODGSON* A. E. St. Thomas’
J. DAVIES
GENT†
Hospital, London SE1
M. E. HODSON
Brompton Hospital,
London SW3
A history of asthma, hay fever, and flexural eczema was significantly more common in patients with adult cœliac disease (A.C.D.) than in normal controls. Autoantibodies were also more common in A.C.D. First-degree relatives of A.C.D. patients were more likely than controls to have atopic disorders. A deficiency of local mucosal immunity due to abnormal IgA responses may underly this association between
Summary
A.C.D.
and atopy. Introduction
ATOPY and immunoglobulin deficiencies are associated. Patients with selective IgA deficiency often have atopic disorders, 12 and Kaufman and Hobbs found
serum-immunoglobulin deficiencies, particularly IgA deficiency, in 6.7% of a group of atopic patients.3 In adult coeliac disease (A.C.D.) the quantity and distribution of IgA may be abnormal. Frank IgA deficiency, present in about 1 in 500 of the general population, is probably twelve times as common in coeliac disease.4 Even when serum-IgA concentrations are normal, and patients have responded well to a gluten-free diet, the ability to mount a normal IgA response to ingested antigens may be impaired.5 Also, the normal preponderance of IgA-producing plasma-cells in the lamina propria of the gut is reduced.6 We found that atopic disorders were more common in patients with treated A.C.D. than in a control hospital population. *Present address: †Present address:
Department of Medicine, Royal Free Hospital, London NW3. Salisbury General Hospital, Wiltshire.
Mitchell, J. R., Thorgeirssan, S. S., Potter, W. Z., Jollow, D. J., Keiser, H. Clin., Pharmac. Ther. 1974, 16, 676. 12. Mitchell, J. R., Jollow, D. J., Gillette, J. R., Brodie, B. B. Drug Metab. Dispos. 1973, 1, 418. 13. Jollow, D. J., Mitchell, J. R., Zampaglione, N., Potter, W. Z. Pharmacology, 1974, 11, 151. 14. Prescott, L. F., Wright, N. Br. J. Pharmac. 1973, 49, 602. 15. Goulding, R. Lancet, 1975, i, 1162. 16. James, O., Lesna, M., Roberts, S. H., Pulman, L., Douglas, A. P., Smith, P. A., Watson, A. J. ibid. 1975, ii, 579. 17. Peters, R. L., Edmondson, H. A. Gastroenterology, 1970, 58, 289. 18. Eastham, E. J., Bell, I. D., Douglas, A. P. Br. med. J. (in the press). 19. Routh, J. R., Shane, N. A., Arrendondo, E. G., Paul, W. D. Clin. Chem. 1968, 14, 882. 20. Hamlyn, A. N., James, O. F. W., Lesna, M., Roberts, S. H., Watson, A. J., Douglas, A. P. Gut, 1975, 16, 838. 21. Dixon, M. F., Fulker, M. J., Waler, B. E., Kelleher, J., Losowsky, M. A. ibid. 1975, 16, 800. 22. Prescott, L. F., Roscoe, P., Wright, N., Brown, S. S. Lancet, 1971, i, 519. 23. Prieto, J., Barry, M., Sherlock, S. Gastroenterology, 1975, 62, 525. 24. McLean, A. E. M. Lancet, 1973, ii, 1490. 25. Phear, E. A., Ruebner, B., Sherlock, S., Summerskill, W. H. J. Clin. Sci. 1955, 15, 93. 26. Edwards, O. M., Edwards, P., Huskisson, E. C., Taylor, R. T. Br. med. J. 1971, ii, 87. 27. Boyer, J. D., Rouff, S. L. J. Am. med. Ass. 1971, 218, 440. 28. Snow, M., Wilkinson, R., James, O. Clin. Sci. Mol. Med. (in the press).
11.
I
linkage disequilibrium may be necessary to the organism in that the gene products have complementary functions. In this regard we have already presented evidence that the HLA-B8 gene is not merely a marker gene but seems to contribute a gene product (surface antigen) that participates in the disease process.’" If this Such
is the case,
suggest that the HLA gene and the B-lymphocyte antigen gene act in a complementary fashion at some critical step in the pathogenesis ofD.H. and ..-.
non
l..___L
i
y
i
we
G.S.E.
The methods used in this study can be easily applied the investigation of other diseases where associations with HLA antigens have been identified. In these cases, unique B-lymphocyte antigens may be found which will lead to an understanding of underlying disease processes. Second, the antigens detected in this report can be used to develop a method of identifying patients with gastrointestinal gluten sensitivity. Such a method will not depend on interpretation of jejunal-biopsy specimens and therefore will lead more easily to the correct diag-
to
nineteen D.H. patients, while antiserum W-1 reacted with fifteen of sixteen G.s.E. patients and fifteen of fifteen D.H. patients. Five other antisera obtained from mothers with children having G.S.E. or D.H. showed no consistent pattern of B-lymphocyte reactivity. Antisera obtained from Amish or randomly selected donors which have been shown to react with B lymphocytes obtained from normal persons carrying the HLA-B8 antigen, showed a similar pattern of reactivity with the lymphocytes in the disease population studied-i.e., they tended to react with HLA-B8-positive patients.
noses.
Requests for reprints should be addressed to D. L. M., Immunology Branch, National Cancer Institute, Building 10, Room 4B08, Bethesda, Maryland 20014, U.S.A.
Discussion The studies show that
antisera reacted with B-lymphocyte surface antigens which are very strongly associated with G.S.E. and D.H. and which are lacking in the normal population so far studied. However, before one can say that these antigens are unique to G.s.E. and D.H. many additional normal individuals must be studied to determine whether these antigens occur in the normal population in low frequency. In addition, patients with other diseases must be studied to determine whether the antigen occurs in other pathological states. The relation of the cell-surface antigen(s) reported here with the pathophysiology of G.S.E. and D.H. is unknown. Lymphocyte surface antigens in mice, termed Ia antigens, are known to influence the ability of mice to respond to certain antigens,7 and humoral enhancing and suppressing factors produced by T lymphocytes which are active on B lymphocytes have la specificity.8 9 The antigens detected in this study may be human analogues of the murine Ia antigens. If this is so the B-lymphocyte antigen associated with G.S.E. and D.H. may represent the gene product of an abnormal Ir gene in man, which in turn results in an immune response to wheat protein and immunologically mediated gut epithelial cell toxicity. A second possibility, not necessarily mutually exclusive from the first possibility, is that the B-lymphocyte antigen identified here is a receptor or binding site for wheat proteins. Such sites may exist on the surface of gut epithelial cells in patients with G.s.E. and D.H. and lead to immune reactions to wheat protein and, in the case of D.H., to the subsequent deposition of IgA immune complexes in the skin. To study these possibilities, inhibition of B-lymphocyte cytotoxicity by wheat protein is currently being investigated. The association of both HLA-B8 and the B-lymphocyte antigen identified here with G.S.E. and D.H. is of some interest. If the gene controlling the B-lymphocyte antigen proves to be part of the histocompatability gene complex, as in the case of previously studied B-lymphocyte antigens,3then the occurrence of these antigens in G.S.E./D.H. may be due to linkage disequilibrium-i.e., cross-over suppression between the two involved genes.
REFERENCES
two
Rogentine, G. N., Strober, W. J. clin. Invest. 1972, 51, 1602. 2. Katz, S. I., Falchuk, Z. M., Dahl, M., Rogentine, G. N., Strober, W. ibid. 1.
Falchuk,
A. M.,
p. 2977. 3. Mann, D. L., Abelson, L., Henkart, P., Harris, S. D., Amos, B. D. Proc. natn. Acad. Sci. U.S.A. (in the press). 4. Alexander, E., Henkart, P. A. Unpublished. 5. Mann, D. L., Abelson, L., Harris, S., Amos, B. D. Nature, (in the press). 6. Abelson, L. D., Henkart, P. A., Mann, D. L. in Manual of Clinical Immunology (edited by E. Yunis); (in the press). 7. Sachs, D. H., Dickler, H. B. Transplant. Rev. 1975, 23, 159. 8. Munro, A. J. Taussig, M. J. Nature, 1975, 256, 103. 9. Takemori, T., Tada, T. J. exp. Med. 1975, 142, 1241. 10. Nelson, D. L., Falchuk, Z. M., Kasarda, D., Strober, W. Clin. Res. 1975, 23, 254abstr.
CONTROLLED TRIAL OF CYSTEAMINE IN TREATMENT OF ACUTE PARACETAMOL (ACETAMINOPHEN) POISONING ADRIAN P. DOUGLAS ADRIAN N. HAMLYN OLIVER JAMES
Gastroenterology Group, Department of Medicine, Royal Victoria Infirmary, Newcastle on Tyne NE1 4LP A randomised controlled trial of the use of intravenous cysteamine in the treatment of severe paracetamol poisoning has been performed. Thirty-eight patients presenting 3-17 h after ingestion were admitted to the trial; of these eighteen received cysteamine. Two patients died from hepatic failure, one in each treatment group. Analysis of the series as a whole showed no advantage of cysteamine in preventing biochemical abnormalities of liver function except for aspartate aminotransferase and serum ferritin levels, which were significantly less after cysteamine therapy. Separate analysis of the patients treated within 9 h of paracetamol ingestion and of those treated 9-17 h after paracetamol ingestion similarly showed no definite advantage of cysteamine. Histological evidence of liver damage showed a possible beneficial effect of cys-
Summary
112
teamine.
Cysteamine therapy did pancreatic damage.
not
preventrenal
or
-
Introduction THE North East Regional Survey, centred at Newcastle upon Tyne, has been concerned with aspects of paracetamol (acetaminophen) poisoning, and an average of 15 such patients are admitted each month. Over 1000 patients with paracetamol overdose are admitted to hospitals in England and Wales every year, and about 3% die.1 Many different measures have been advocated for their management, including corticosteroids,2 haemodialysis,3 forced diuresis,2 orally administered cholestyramine or activated charcoal, and charcoal-column hæmoperfusion.5 None of these measures have been proved to prevent liver damage in severe paracetamol poisoning. In 1974 Prescott et al. reported6 the results of intravenous dosage with cysteamine in patients after paracetamol poisoning with apparent favourable results. The work of Mitchell and his colleagues,7-13 has suggested that paracetamol is toxic because of its oxidised metabolites. In therapeutic doses paracetamol is safely conjugated with sulphate and glucuronic acid in the liver but, as the dose is increased, the metabolites are neutralised by glutathione, which contains a sulphydryl (SH) group, until all the glutathione is used up. Then other SH groups in the cell (such as are present in many intracellular membranes and enzymes) are consumed, and damage ensues. This hypothesis provides the rationale for cysteamine treatment. Cysteamine (&bgr;-mercaptoethylamine) is a nucleophilic SH-containing compound which reduces the hepatotoxicity of paracetamol in laboratory animals,12 13 perhaps by supplying SH groups or possibly by slowing the intracellular metabolism of paracetamol.6 The clinical course of paracetamol poisoning varies6 14 15 so assessment of the effect of treatment is difficult. Previous workers have pointed out that "many forms of therapy... have been advocated without control data."6 We have carried out a controlled trial of cysteamine in the management of paracetamol poisoning.
Patients and Methods Patients AU patients who were admitted to our hospital within 17 h of paracetamol ingestion, who were HbsAg negative with no history of pre-existing liver disease, and whose plasma-paracetamol was above the line indicated in fig. 1 were included in the trial. The line shown. in fig. 1 is very similar to that of Prescott et al.6 and is constructed from our own data on more than 250 patients over the past 2 years.’6 Patients whose presenting plasma-paracetamol fell above the line had a substantial risk of death or serious liver damage. 38 patients were admitted to the trial. 18 were assigned to receive cysteamine and 20 patients did not receive cysteamine. The patients had ingested between 7 and 85 g of paracetamol; no correlation was observed between the stated dose and plasma-paracetamol concentration.
Design of Trial The trial was planned to answer four questions. Does cysteamine therapy affect mortality after paracetamol overdose? Does cysteamine prevent severe, non-fatal, liver damage or other injury caused by paracetamol overdose? In view of the theoretical considerations governing the use of cysteamine, is it important to give cysteamine early after the overdose? And, is age an important factor in determining mortality or mor-
bidity after paracetamol overdose as it is in other forms of acute hepatic failure17? Patients were assigned on admission to one of four groups: (I) young (less than 30 years) and early (less than 9 h); (n) young and late; (III) old and early; and (tv) old and late. Patients were randomly allocated to receive cysteamine or not using a table of random numbers. Using a table of random numbers, it was possible that one or other group could become weighted with cases receiving only one of the treatment regimens. Adjustment was made therefore, so that every six patients in each group included three who had received cysteamine.
Methods Liver function
was assessed daily by measurement, on fast. ing blood-samples, of bilirubin, aspartate aminotransferase (A.S.T., normal range 4-20 I.U.!1), alkaline phosphatase, albumin and total protein, prothrombin ratio (Quick one-stage method), total bile acidsl6 (normal < 14 µmol/l), and ferritin" (normal > 240 mg/1). Liver biopsy was performed with a Menghini needle on the fourth day after admission, or as soon thereafter as the prothrombin-time and patient’s condition permitted. Histopathological changes were graded severe, moderate, or mild. 16 Renal function was assessed daily by measurement of urine volume, plasma urea, electrolytes, and creatinine, and, where appropriate, creatinine clearance and urinary urea and
sodium. Pancreatic
damage was assessed by daily estimation of serum-amylase (normal<300 units/l). Plasma-paracetamol was measured spectrophotometrically. 19 Treatment All patients received a standard regimen.16 Briefly this consisted of intravenous 5% dextrose, 2-3 litres daily, with added vitamins and potassium as necessary. This was maintained for 72 h or until liver function was satisfactory, and for the fist 48 h nothing was given by mouth. If liver function became moderately or severely impaired (A.S.T. > 100 units/1, proof control) the infusion was continued thrombin-time <50% and lactulose, 10 ml 6-hourly was given. If hepatic encephalopathy developed, this was treated conventionally. Two patients who reached grade iv hepatic coma were treated with charcoal-column hoemoperfusion; both died. Cysteamine was given as described by Prescott et al." except that it was dissolved in 5% dextrose and injected, or added to the 5% dextrose infusion, using a Millipore filter attached to a syringe. 2-hourly estimations of capillary-blood sugar were made by a fingerprick method (Ames ’Dextrostix’), and if the bloodsugar dropped below 2.5 mmol/1 (4) mg/dl), 100 ml of 5% dextrose was
injected intravenously.
Statistical Analysis Where appropriate the data rank-sum test.
were
analysed using Wilcoxon’s
Results
Only two of our possible groups--young and early (six cysteamine, seven no cysteamine) and young and late (eight cysteamine, ten no cysteamine) contained sufficient numbers to be analysed separately from the entire trial sample. This is a reflection of the age distribution of patients presenting with paracetamol self-poisoning. In addition we analysed all fourteen early patients (young and old) separately from the nineteen late patients. Paracetamol "Index" Using the data in fig. 1 a paracetamol "index" was derived for each patient. This was the concentration, in
113 Minimal Prothrombin
Maximal Bilirubin
Received Cysteamine No Cysteamine A Severe histological change
0
t Died
Cysteamine No Cysteamine Time in hours after ingestion of Paracetamol
Fig. 1-Plasma-paracetamol levels with paracetamol poisoning
on
admission to
hospital in patients
mg/1, by
which the patient exceeded the theoretical "safe" upper limit, indicated by the line, at the time when the plasma-paracetamol was measured. The mean value for the 18 cysteamine patients was 72 mg/1 and the mean value for the 20 non-cysteamine patients was 98 mg/l. The difference between these values is significant (P<0.02) and indicates that the non-cysteamine group were potentially at greater risk from paracetamol poisoning. This difference arose by chance because of the design of the trial. The mean index for the cysteamine-treated young and early patients was 43 mg/1 and the mean index for the young and early non-cysteamine patients was 138 mg/1 (P<001). There was no significant difference in these values in the young and late patients (75 mg/1 cystea-
mine, 67 mg/l non-cysteamine). Deaths Two
patients died of fulminant hepatic failure. One,
Fig. 2—Distribution of maximum bilirubin and minimum prothrombin in patients with paracetamol poisoning. t Denotes patients who died.
aged 24, had not received cysteamine (admitted to hospital 6.5h after a stated ingestion of 75 g of paracetamol) and the other, aged 28, had received cysteamine (admitted to hospital 9.5h after ingesting 15 g of paracetamol). Liver Function
These data are summarised in the table. The distribution of maximum bilirubin and minimum prothrombin (both achieved on the third or fourth day) in our patients is shown in fig. 2. There is no difference between the two groups. Similarly, there is no difference when the young and late and early and late groups are analysed separately. However, the difference for prothrombin in the young early group just reaches statistical significance (P=0.05). Fig. 3 shows the results for three tests of hepatocellular damage. No significant differences are seen between cysteamine and non-cysteamine treated patients
LIVER-FUNCTION TESTS IN PATIENTS WITH ACUTE PARACETAMOL POISONING I
Including one patient who died. Differences between cysteamine and
I
.
no
Cysteamine No Cysteamine
cysteamine not significant except for ferritin and A.S.T. (P<0.01
f<0.05).
(MEAN VALUES)
114 there was a significantly lower paracetamol index in those receiving cysteamine, all six patients on cysteamine had moderate (grade n) changes, whereas in those who did not receive cysteamine and who had a greatly higher paracetamol index, five out of seven showed
grade-m changes. Renal Function Four of
our
patients developed histologically
proven
tubular necrosis (one of whom died in hepatic failure). Two of these had received cysteamine and the other two had not. acute
.
Cysteamine No Cysteamine Cysteamine No Cysteamine
Cysteamine No Cysteamine
3-Distribution of maximum serum-A.S.T., maximal serum-bileacids and maximum serum-ferritin in patients with paracetamol poi-
Fig.
soning.
with respect to peak serum concentrations of bile acids. However, there are striking and significant differences (P<0.01) between the two sets of patients with regard to peak serum-A.s.T. In the entire series, and also in both early and late patients, the serum-A.S.T. is lower in the cysteamine-treated than in the non-cysteamine patients. A similar difference was seen with regard to the serumferritin in the entire series (P<0.01), and in the early not but in the late groups (P<0.05 ) groups. Liver Histology
Fig. 4 illustrates the distribution of the three grades of histological abnormality. More patients had severe (grade m) abnormalities in the non-cysteamine group than amongst those who received cysteamine. Fig. 1 shows that most grade-in changes were in patients with high plasma-paracetamol concentrations and that when patients receiving or not receiving cysteamine had comparable paracetamol levels the severity of histological change was similar. In the young early group, in which
Serum-amylase Abnormally high serum-amylases were recorded (often within 48 h of paracetamol ingestion) in six of the cysteamine-treated patients (430-780 units/1) and in three of the non-cysteamine-treated patients (370-440
units/1). Toxic Effects of Cysteamine Our experience was similar to that of Prescott et al.6 All our patients rapidly became nauseated and repeatedly vomited and retched for the duration of cysteamine therapy and up to 36 h afterwards. In addition they lay in bed looking utterly dejected and miserable and their breath stank of mercaptans. Although the patients in the non-cysteamine group, with severe liver damage, commonly experienced some nausea, they did not have the general appearance of the cysteamine group. We have not established any long-term toxic effects of cysteamine. Liver function and histology of patients who received cysteamine were identical at 3 months’ follow-up with those of survivors who did not receive cys-
teamine.2o Discussion The importance of preventing death from paracetamol-induced hepatic necrosis lies not least in the fact that most of these patients are young and quickly regret their action.However, because only 3-5% of such overdosages prove fatal, a very large trial is needed to demon strate that any new treatment saves lives; and our trial is on too small a scale to permit any conclusions about the effect of cysteamine on mortality-rates. One patient on cysteamine and one not on cysteamine died. Prescott et al. recorded no deaths in eight treated patients and two deaths in eleven previous, and not sequential, noncysteamine-treated cases. This difference is not statistic-
ally significant. Cysteamine did
prevent raised transaminase and ferritin levels, but is this truly indicative of an effect in preventing hepatic centrilobular necrosis? In the rat, peak transaminase values seem to correlate well with the extent of paracetamol-induced liver necrosis21 but in man this correlation is less exact,12 18 22 and an increase in serum-bilirubin and prothrombin ratio may be a better prognostic index.’ In our patients, serum-bilirubin and prothrombin ratio were much the same in the cysteamine and non-cysteamine patients. Nor did differences emerge when another sensitive indicator of No
Cysteamine
Cysteamine
Fig. 4-Distribution of grades histological abnormality paracetamol poisoning.
in
patients
with
hepatocellular dysfunction, serum-bile-acids, was considered. However, changes in serum-ferritin, a newly recognised indicator of acute hepatocellular
115
damage, 18 23
do accord with the alterations of Further evidence on this point is provided serum-A.S.T. data. Cysteamine therapy seemed to by our histological severe (grade in) changes. fewer with be associated this difference However, may just as well be due to the fact that, by chance, the cysteamine-treated had lower plasma-paracetamol concentrations. Our observations certainly accord with the hypothesis generated by the work of Mitchell and his colleagues,H3 but administration of cysteamine, a toxic substance, cannot yet be said to be justifiable routinely for all patients with paracetamol poisoning. Our data suggest that if cysteamine is used at all, it must be given within 9 h of paracetamol ingestion. Further work with other SH donors certainly seems worthwhile, however. The incorporation of methionine into tablets containing paracetamol might be helpful,24 but methionine given after a paracetamol overdose may be harmful since it has a comagenic effect in liver disease.25 The extrahepatic effects of paracetamol are often forgotten in the face of the potential hepatoxicity. Acute renal failure is a well-recognised feature of paracetamol poisoning,26-29 and we have no evidence that cysteamine prevents this. Acute pancreatitis (not complicating renal failure) also occurred in our patients and was as often associated with cysteamine as with no cysteamine. This series is too small to permit conclusions about the treatment of paracetamol poisoning. It is important to bear in mind that we did not carry out a controlled trial of cysteamine versus nothing, but rather studied patients receiving dextrose, some of whom also received cysteamine. In the past 15 months we have had six deaths in hepatic failure due to paracetamol poisoning: two in this trial and four who had been transferred to our hospital more than 72 h after drug ingestion during which time they had received no specific therapy. We recommend that such patients should be treated with intravenous 5% dextrose and no food by mouth for the first 48 h at least. This may or may not be beneficial, but it seems to do no harm. We thank our junior medical colleagues for their tireless help; the physicians of the region for referring cases; and Prof. D. N. S. Kerr
for his encouragement; Dr P. A. Smith, Dr A. J. Fleetwood, Dr M. J. Watson, and others who estimated the plasma-paracetamol levels; Mrs L. Pulman for the serum-bile-acid estimations; Dr E. J. Eastham and Mr I. Bell for the ferritin estimations; the Departments of Clinical Biochemistry and Hxmatology for their help; and Dr M. Lesna and Dr A. J. Watson for the liver histology.
Requests for reprints should be addressed to
0.
J.
REFERENCES 1.
Clark, R., Thompson, R. P. M., Borkrakchanyurat, V., Widdop, B., Davidson, A. R., R., Williams, R. Lancet, 1973, i, 66. 2. MacLean, D., Peters, T. J., Brown, R. A. C., McCathie, M., Baines, G. E., Robertson, P. G. C. ibid. 1968, ii, 849. 3. Fand, N. R., Glynn, J. P., Kerr, D. N. S. ibid. 1972, iii, 396. 4. Dordom, B., Wilson, R. S., Thompson, R. P. H., Williams, R. Br. med. j. 1973, iii, 86. 5. Wilson, R. S., Winch, J., Thompson, R. P. M., Williams, R. Lancet, 1973, i, 77. 6.
Prescott, L. F., Newton, R. W., Swainson, C. P., Wright, N., Forrest, A. R. W., Matthew, H. ibid. 1974, i, 588. 7. Mitchell, J. R., Jollow, D. J., Porter, W. Z., Davis, D. C., Gillette, J. R., Brodie, B. B. J. Pharmac. exp. Ther. 1973, 187, 185. 8. Jollow, D. J., Mitchell, J. R., Potter, W. Z., Davis, D. C., Gillette, J. R., Brodie, B. B. ibid. 1973, 187, 195. 9. Potter, W. Z., Davis, D. C., Mitchell, J. R., Jollow, D. J., Gillette, J. R., Brodie, B. B. ibid. 1973, 187, 203. 10 Mitchell, J. R., Jollow, D. J., Potter, W. Z., Gillette, J. R., Brodie, B. B. ibid. 1973, 187, 211.
ATOPIC DISORDERS AND ADULT CŒLIAC DISEASE R.
H. J. F. HODGSON* A. E. St. Thomas’
J. DAVIES
GENT†
Hospital, London SE1
M. E. HODSON
Brompton Hospital,
London SW3
A history of asthma, hay fever, and flexural eczema was significantly more common in patients with adult cœliac disease (A.C.D.) than in normal controls. Autoantibodies were also more common in A.C.D. First-degree relatives of A.C.D. patients were more likely than controls to have atopic disorders. A deficiency of local mucosal immunity due to abnormal IgA responses may underly this association between
Summary
A.C.D.
and atopy. Introduction
ATOPY and immunoglobulin deficiencies are associated. Patients with selective IgA deficiency often have atopic disorders, 12 and Kaufman and Hobbs found
serum-immunoglobulin deficiencies, particularly IgA deficiency, in 6.7% of a group of atopic patients.3 In adult coeliac disease (A.C.D.) the quantity and distribution of IgA may be abnormal. Frank IgA deficiency, present in about 1 in 500 of the general population, is probably twelve times as common in coeliac disease.4 Even when serum-IgA concentrations are normal, and patients have responded well to a gluten-free diet, the ability to mount a normal IgA response to ingested antigens may be impaired.5 Also, the normal preponderance of IgA-producing plasma-cells in the lamina propria of the gut is reduced.6 We found that atopic disorders were more common in patients with treated A.C.D. than in a control hospital population. *Present address: †Present address:
Department of Medicine, Royal Free Hospital, London NW3. Salisbury General Hospital, Wiltshire.
Mitchell, J. R., Thorgeirssan, S. S., Potter, W. Z., Jollow, D. J., Keiser, H. Clin., Pharmac. Ther. 1974, 16, 676. 12. Mitchell, J. R., Jollow, D. J., Gillette, J. R., Brodie, B. B. Drug Metab. Dispos. 1973, 1, 418. 13. Jollow, D. J., Mitchell, J. R., Zampaglione, N., Potter, W. Z. Pharmacology, 1974, 11, 151. 14. Prescott, L. F., Wright, N. Br. J. Pharmac. 1973, 49, 602. 15. Goulding, R. Lancet, 1975, i, 1162. 16. James, O., Lesna, M., Roberts, S. H., Pulman, L., Douglas, A. P., Smith, P. A., Watson, A. J. ibid. 1975, ii, 579. 17. Peters, R. L., Edmondson, H. A. Gastroenterology, 1970, 58, 289. 18. Eastham, E. J., Bell, I. D., Douglas, A. P. Br. med. J. (in the press). 19. Routh, J. R., Shane, N. A., Arrendondo, E. G., Paul, W. D. Clin. Chem. 1968, 14, 882. 20. Hamlyn, A. N., James, O. F. W., Lesna, M., Roberts, S. H., Watson, A. J., Douglas, A. P. Gut, 1975, 16, 838. 21. Dixon, M. F., Fulker, M. J., Waler, B. E., Kelleher, J., Losowsky, M. A. ibid. 1975, 16, 800. 22. Prescott, L. F., Roscoe, P., Wright, N., Brown, S. S. Lancet, 1971, i, 519. 23. Prieto, J., Barry, M., Sherlock, S. Gastroenterology, 1975, 62, 525. 24. McLean, A. E. M. Lancet, 1973, ii, 1490. 25. Phear, E. A., Ruebner, B., Sherlock, S., Summerskill, W. H. J. Clin. Sci. 1955, 15, 93. 26. Edwards, O. M., Edwards, P., Huskisson, E. C., Taylor, R. T. Br. med. J. 1971, ii, 87. 27. Boyer, J. D., Rouff, S. L. J. Am. med. Ass. 1971, 218, 440. 28. Snow, M., Wilkinson, R., James, O. Clin. Sci. Mol. Med. (in the press).
11.