- Email: [email protected]
AN AUGMENTED SCHILLING TEST IN THE DIAGNOSIS OF PERNICIOUS ANÆMIA
545 in patients with congenital clotting disorders IS sugthat complete correction of the coagulation abnorMtities should be attempted at biopsy to avoid bleeding ;omplications. The rationale underlying our adoption of a single administration of replacement therapy was that the natural concentrations of clotting factors present in these patients should be sufficient to maintain haemostasis, once this has been firmly established at biopsy. ures
A relatively large dose of F.F.P. was found to be least effective in correcting abnormal coagulation tests, as was found by others.16 The concentrate containing factors, u, ix, and x was also not sufficiently effective, but the additional administration of factor-vn concentrate was followed by return to normal Of P.T. and N.T. results in most patients. However, as was found by Green et al,," this regimen failed to correct K.P.T.T., and normal results in the tests exploring both extrinsic (P.T. and B.T.) and intrinsic (K.P.T.T.) coagulation systems could only be produced by the combined administration of less F.F,P. (8 ml/kg) and of the concentrates (12 units/kg). These findings are maybe related to the infusion with plasma of factors v, xi, and XII, which are often considerably reduced in patients with liver disease (unpublished observations) and may contribute to the prolongation of K.P.T.T. Besides being more effective, such a combined therapeutic regimen allows for the volume of plasma administered to be reduced and thus decreases the risk ofhypervolaemia. Hepatitis and thromboembolism are regarded as major hazards associated with the use of prothrombincomplex concentrates. In the twelve months after biopsy, neither acute hepatitis nor HBsAg antigen developed in our patients. Even though hepatitis was not reported in two large studies18 19 involving patients treated with the concentrates used in this investigation, there is still a real risk associated with their use. This risk must be weighed against that of bleeding when a liver biopsy is contemplated in a patient with chronic liver disease who has a coagulation defect. Impaired clearance of activated clotting factors20 and reduced synthesis of naturally occurring inhibiiorsl3 facilitate intravascular coagulation in patients with liver disease. In our series, there was no change in those tests that reflect the occurrence of intravascular coagulation fplatelet-count, plasma fibrinogen, factors v and VIII, and serum-F.D.p.) immediately and up to four hours after the infusion. It was also reassuring to find that the infusion of concentrates which might contain clotting factors in activated form was not accompanied by a decrease of antithrombin in (anti Xa). The administration of F.F.P., which provides a source of inhibitors, should help to reduce the risk of thromboembolism. In our opinion, activated clotting factors contained in
prothrombin-complex concentrates are unlikely to cause thromboembolic complications unless these preparations are used in patients who are severely ill and have other underlying conditions facilitating the development of intravascular coagulation (such as major surgery, hsmorrhagic shock, and coma). Thus, concentrates together with F.F.P. may be used before biopsy, thereby allowing the procedure to be carried out safely in pa’.icnts with coagulation defects. thank Dr D. C. B. Mills for his help in the preparation of the ".s:iis:npt.
Requests for reprints should be addressed to P. M. M., Haemophilia and Thrombosis Cent.-e, Angelo Bianchi Bonomi, Via Pace 15, 20122 Milan, Italy. REFERENCES 1. Roberts, R. H., Cederbaum, A. I. Gastroenterology, 1972, 63, 297. 2. Alderson, M. R., Poller, L., Thomson, J. M. J. clin. Path. 1970, 23, 281. 3. Matchett, M. O., Ingram, G. I. C. ibid. 1965, 18, 465. 4. Owren, P. A. Farmakoterapi, 1969, 1-2, 14. 5. Praga, C., Valentini, L., Maidrano, M., Cortellaro, M. in Platelet Function and Thrombosis: A Review of Methods (edited by P. M. Mannucci and S. Gorini) p. 271. New York, 1972. 6. Denson, K. W. E. in Human Blood Coagulation Haemostasis and Thrombosis (edited by R. Biggs); p. 602. Oxford, 1972. 7. Wilson, W. J., G. I. C., Hills, M. Coagulation, 1971, 4, 113. 8. Vermylen, C., De Vreker, R. A., Verstraete, M. Clinica chim. Acta, 1963,
Ingram,
8, 418. Hawiger, D., Niewiarowski, S., Gurewich, V., Thomas, D. P. J. Lab. clin. Med. 1970, 75, 93. 10. Owren, P. A. Lancet, 1959, ii, 754. 11. Yin, E. T., Wessler, S., Stell, P. J. biol. Chem. 1971, 246, 3712. 12. Fagerhol, M. K., Abildgaard, V. Scand. J. Hœmat. 1970, 7, 10. 13. Biggs, R., Denson, K. W. E., Akman, N., Bonett, R., Hadden, M. Br. J. Hœmat. 1970, 19, 283. 14. Ling, C. M., Overby, L. R. J. Immun. 1972, 109, 834. 15. Biggs, R. in Treatment of Haemophilia and other Coagulation Disorders (edited by R. Biggs and R. G. MacFarlane); p. 166. Oxford, 1966. 16. Gazzard, B. G., Henderson, S. M., Williams, R. Gut, 1975, 16, 621. 17. Green, G., Dymock, I. W., Poller, L., Thomson, J. M. Lancet, 1975, i, 1311. 9.
18. Fischer, M. Medsche Welt, 1973, 24, 1899. 19. Vinazzer, H., Bergmann, H. Trans Xth Congr. Wld Fedn Hæmophilia (in the press). 20. Wessler, S., Yin, E. T., Gaston, L. W., Nicol, I., Thromb. Diath. Hœmorr.
1967, 18, 12.
AN AUGMENTED SCHILLING TEST IN THE DIAGNOSIS OF PERNICIOUS ANÆMIA
JAMES A. STRAUCHEN* Department of Medicine, University of California Medical Center, San Francisco, California 94122, U.S.A. A patient with otherwise typical addisonian pernicious anæmia and serum anti-intrinsic-factor antibody failed to respond to oral intrinsic factor on repeated testing during three years of therapy with parenteral vitamin B12. There was no evidence of generalised malabsorption. To test the hypothesis that binding of intrinsic factor to gut secretory anti-
Summary
body
was
responsible,
an
"augmented" Schilling
test
devised using eight times the usual dose of intrinsic factor. This increased dose of intrinsic factor resulted in normal absorption of the test dose of vitamin B12, confirming the diagnosis. It is suggested that the augmented Schilling test may be useful in the diagnosis of the occasional patient with features of pernicious anæmia who fails to respond to conventional doses of intrinsic factor in the Schilling test. was
Introduction
Schilling test has been a major tool in the diagmegaloblastic anaemia since its description by Schilling in 1953.’ The pattern of low excretion of a test dose of labelled vitamin B12 with an increase to normal after administration of intrinsic factor is diagnostic of addisonian pernicious anaemia, and serves to distinguish this condition from other causes of vitamin B12 malabsorption. A variety of autoantibodies has been described in patients with pernicious anaemia including anti-inTHE
nosis of
*Present address: Laboratory of Pathology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20014.
546 trinsic-factor antibodies of both binding and blocking type.23 These antibodies might have a role in the pathoof the disorder by interfering with the secretion and function of intrinsic factor and may also interfere with the Schilling test.4 The present report describes a patient with otherwise classic addisonian pernicious anaemia who consistently failed to respond to oral intrinsic factor on repeated testing over a three-year period in association with serum anti-intrinsic-factor antibody. This patient did, however, show normal excretion of a test dose of labelled vitamin B12 in an "augmented" Schilling test using eight times the usual dose of oral intrinsic factor. This may be a useful approach in the diagnosis of pernicious anaemia in the presence of antiintrinsic-factor antibody.
genesis
mg/dl
and serum-vitamin B12 was 480 pg/ml. ErythrocB sedimentation-rate was 6 mm/h." Serum-calcium was 9f
mg/dl, phosphorus 3-8mg/dl, glucose 84 mg/dl, blood-urea-ru. trogen 11 mg/dl, and creatinine 0.8mg/dl. Total protein was 7.7g/dl, albumin was 4.66 g/dl, serum-cholesterol 160mi. serum-carotene 72 mg/dl, serum-thyroxine 4.0 g/dt, and tn. iodothyronine resin 29%. Urinalysis was normal. Stool fat was less than 5 g/24 h. Stool examination for ova and parasites was negative. Serum was positive for both antiparietat-teti and anti-intrinsic-factor antibodies, but negative for antinuclear antibody. Stimulated gastric analysis with ametazole hydro. chloride (’Histalog’) (1-7mg/kg subcutaneously) demonstrated histamine-fast achlorhydria with gastric fluid pH greater than 7-0in all specimens. A standard Schilling test with 1 N.F, unit intrinsic factor showed less than 1% excretion of the test dose in 48 h. A repeat Schilling test with 8 N.F. units intrinsic factor (eight times the standard dose) showed 13% excretion in 48 h
(see accompanying table). Methods and Materials RESULTS OF SCHILLING TESTS
Schilling tests were performed in a standard manner with a test dose of 0.7 p.Ci of 57Co-cyanocobalamin, followed by an intramuscular flushing dose of 1 mg unlabelled cyanocobalamin. Urine was collected for 24 h before the test and for the following 48 h. Hog intrinsic-factor of standardised potency (E.R. Squibb Pharmaceutical Co. 1 National Formulary (N.F.) unit equals approximately 50 mg) was administered orally in a dose of 1 N.F. unit in the ordinary Schilling test and in a dose of 8 N.F. units in the augmented test. The anti-intrinsic-factor antibody test was done by Dr Hugh Fudenberg.
Case-history The patient is a 37-year-old White female admitted for evaluation ofvitamin-B12 deficiency. The patient has a longstanding history of seizure disorder for which she was maintained on phenytoin, (’Epanutin’ [’Dilantin’]) phenobarbitone, and primidone (’Mysoline’). Ansemia was first noted following her second pregnancy in 1959 and was said to have responded to iron therapy. In 1970 she presented with generalised weakness and severe anaemia with Hb 4.3 g/dl. No glossitis or neurological signs were noted. Evaluation revealed a megaloblastic bone-marrow. Reticulocyte count, Coombs’ test, serum-iron/ total iron binding capacity, Hb A2, F, and electrophoresis, and serum-haptoglobin were normal. Increased serum-folate of 35 ng/ml and markedly decreased serum-vitamin-B12 of 70 pg/ml was noted. Schilling test showed less than 1% excretion without intrinsic factor, and 2% excretion with intrinsic factor (normal greater than 7%). Pernicious ansemia was suspected and the patient began monthly injections of vitamin B12 resulting in a brisk reticulocyte response, and progressive increase in Hb and hsematocrit to normal levels. In 1972 haematocrit was 37%, serum-iron/total iron binding capacity was 35/360 mg/dl, and oral iron supplements were added to the regimen. A repeat Schilling test at this time again showed less than 1% excretion with no increase with intrinsic factor. Because of the persistently abnormal Schilling test with intrinsic factor suggesting malabsorption rather than pernicious anaemia the patient was admitted to the University of California Hospital for further evaluation. There was no history of weight-loss, diarrhoea, steatorrhsra, parasitic infection, or abdominal complaints. Physical examination revealed no significant findings. There was normal papillation of the tongue and normal neurological examination. X-ray examinations of the chest and abdomen were within normal limits. Upper gastrointestinal and small bowel series showed no evidence of malabsorption or other significant finding. Laboratory studies revealed Hb 12.1 g/dl, haematocrit 36%, red-blood-cell count 5 210 000/mm3, reticulocytes 0-6%, platelets 245 000/mm3, white-blood-cell count 4300/mm3 with 47% neutrophils, 42% lymphocytes, 2% monocytes, and 1% eosino-
phils.
The
serum-iron/level
iron
binding capacity
was
63/280
On the basis of histamine-fast achlorhydria, positive serum antiparietal-cell, and anti-intrinsic-factor antibodies, and response to an augmented dose of intrinsic factor in the Schilling test a diagnosis of pernicious anaemia was confirmed and the patient discharged to continue monthly injections of vitamin B1.
Discussion This patient was evaluated because a repeatedly abnormal Schilling test with intrinsic factor raised the question of generalised malabsorption as the cause of her vitamin-B12 deficiency rather than pernicious anaemia. The response to intrinsic factor had remained abnormal despite continuous prolonged therapy with parental vitamin B12 ruling out secondary intestinal atrophy as the cause of her persistently abnormal Schilling test. Screening tests for malabsorption including small bowel X-ray series, faecal fat, stool ova, and parasite examination, and determinations of serum carotene, calcium, and cholesterol were all within normal limits. The diagnosis of pernicious anaemia, however, was strongly supported by the findings of histamine-fast achlorhydria and positive serum tests for antiparietalcell and anti-intrinsic-factor antibodies. The finding of serum anti-intrinsic-factor antibody suggested that the failure to respond to oral intrinsic factor in the Schilling test might be due to the presence of secretory anti-intrinsic-factor antibody in the gut. It has been shown that patients with pernicious anaemia may have secretorya well as serum anti-intrinsic-factor antibodies, both of the binding and blocking types, the latter of which blocks the binding of vitamin B12 to intrinsic factor preventing ileal absorption.4An augmented Schilling test was therefore devised, using a larger dose of intrinsi factor in order to saturate any secretory antibody present. Administration of eight times the standard dose c: intrinsic factor resulted in excretion of 13% of t labelled dose of vitamin B12’ whereas administration : the standard dose had resulted in excretion of less thar 1% (table). This response to an augmented Schillings
547
confirmed the diagnosis of pernicious anaemia in this patient. Unfortunately, the. patient refused a second nasogastric aspiration for direct assay of secretory antibody in the gastric fluid. I suggest that the
augmented Schilling test may be a useful diagnostic tool in the occasional patient with features of pernicious anaemia who fails to respond to conventional doses of intrinsic factor. I am grateful to Dr Stephen B. Shohet for kindly reviewing the
,
manuscript. REFERENCES
1. Schilling, lab. clin. Med. 1953, 42, 860 2. Ardeman, S., Chanarin, I. Lancet, 1963, ii, 1350. 3. Schade, S. G., Abels, J., Schilling, R. F. J. clin. Invest. 1967, 46, 615. 4. Fisher, J. M., Rees, C., Taylor, K. B. Lancet, 1966, ii, 88. 5. Bardhan, K. D., Hall, J. R., Spray, G. H., Callender, S. T. E. ibid. 1968, R. F. J.
ii, 62.
Hypothesis INTERACTION OF HLA MOLECULES WITH NON-IMMUNOLOGICAL LIGANDS AS AN EXPLANATION OF HLA AND DISEASE ASSOCIATIONS A.
L. P. RYDER SVEJGAARD Tissue Typing Laboratory, Blood Grouping Department, State University Hospital (Rigshospitalet), Copenhagen, Denmark
It is postulated that alloantigens (HLA in particular) may interfere with ligand/ receptor interactions not directly involved in immune reactions. This interference may be regarded as a sideeffect of the creation of new mutant HLA alleles which are maintained in the population for other reasons. As an example, it is suggested that some HLA molecules may have structures resembling those on receptors for certain hormones which could cause competition between HLA and receptor molecules for the hormone. Facilitation of ligand/receptor interaction may also be envisaged. Such interactions could under certain conditions lead to disease and might explain some associations between HLA and non-immunological diseases (e.g., manic-depressive disease and hæmachromatosis).
Summary
INTRODUCTION
MANY diseases are more common in individuals carrying certain antigens of the HLA systeml 2-i.e., the major histocompatibility system in man. Genetic systems resembling the HLA system have been found in all vertebrates investigated so far. Such systems control (i) strong alloantigens, (ii) many immune responses, and iti) some components of the complement cascade.3-6 We have called them ARMS (antigen/receptor/mediator system) systems6 because immune-response (Ir) determinants may be regarded as receptors for antigens. So far, the associations between HLA and disease have mainly been demonstrated for some HLA-B and HLA-D (mixed lymphocyte culture [M.L.C.]) determinants, but there is no proof that these antigens are causally responsible for disease susceptibilities. The as-
sociations found may be secondary to primary associations with other HLA factors for which we cannot type at the moment-e.g., Ir determinants, and which are thought to be in strong linkage disequilibrium with factors which have been studied so far-HLA-A, B, and D determinants. Several theories have been put forward to explain the associations. 1 7-9 They could be due to (i) the adverse effects of specific Ir genes’ or "lacunar immunodeficiencies";10 (ii) molecular mimicry7 (i.e., resemblance between some microbial antigens and some HLA antigens); (iii) a receptor function of HLA factors for certain viruses; (iv) an abnormal function of one or more of the complement components coded by the HLA system; and/or (v) inadequately operating differentiation antigens8 (some HLA antigens [e.g., HLA-D which may be identical to B-lymphocyte antigens] are differentiation antigens in the sense that they are present only on some cell types). The first of these theories is likely to be the explanation for many of the associations observed so far. Molecular mimicry has not yet been shown to operate, while the receptor theory is receiving increasing attention due to the observation of Doherty and Zinkernagel11 that virus-induced alterations of selfantigens are important for the immunological elimination of virus-infected cells. Complement deficiencies may cause associations, and our knowledge of differentiation antigens is yet too limited to allow more than speculation. We suggested that ARMS, apart from dealing with foreign matter, may interfere with the interaction between hormones and receptors in the body.69 Increasing evidence supports this assumption which gives rise to the following hypothesis which could be tested. HYPOTHESISS
Within the term HLA
we
include all factors coded for
genes in linkage disequilibrium with the HLA system-i.e., both the classic HLA-A, B, and C antigens; the HLA-D antigens; some complement factors; the B-cell antigens; and other as yet unknown factors (e.g., homologues of the Tia, Ly, and other H-2 linked antigens in mice). Some HLA determinants-and perhaps other surface and fluid phase molecules---may interfere with the interaction between ligands (e.g., hormones) and their receptors on cell surfaces. One reason to believe this hypothesis is that HLA is itself a ligand/receptor system,
by
and with the considerable number (probably over 100) different HLA factors in the population, it is quite possible that some of these would be non-neutral in relation to other ligand/receptor interactions. Interference with the interactions may arise by various mechanisms, and one of the more simple ones is outlined in the accompanying figure. If an HLA antigen has some incidental resemblance to the binding site of a cell-surface receptor molecule for a given hormone, there could be competition between the receptor and the HLA antigen molecule. Naturally, the affinity between the ligand and its receptor must be much stronger than that between ligand and the HLA antigen, but since the receptor is present only on the specific target cells, whereas many HLA antigens are present on all cells and even in body fluids, the affinity of HLA for ligand need perhaps not
A relatively large dose of F.F.P. was found to be least effective in correcting abnormal coagulation tests, as was found by others.16 The concentrate containing factors, u, ix, and x was also not sufficiently effective, but the additional administration of factor-vn concentrate was followed by return to normal Of P.T. and N.T. results in most patients. However, as was found by Green et al,," this regimen failed to correct K.P.T.T., and normal results in the tests exploring both extrinsic (P.T. and B.T.) and intrinsic (K.P.T.T.) coagulation systems could only be produced by the combined administration of less F.F,P. (8 ml/kg) and of the concentrates (12 units/kg). These findings are maybe related to the infusion with plasma of factors v, xi, and XII, which are often considerably reduced in patients with liver disease (unpublished observations) and may contribute to the prolongation of K.P.T.T. Besides being more effective, such a combined therapeutic regimen allows for the volume of plasma administered to be reduced and thus decreases the risk ofhypervolaemia. Hepatitis and thromboembolism are regarded as major hazards associated with the use of prothrombincomplex concentrates. In the twelve months after biopsy, neither acute hepatitis nor HBsAg antigen developed in our patients. Even though hepatitis was not reported in two large studies18 19 involving patients treated with the concentrates used in this investigation, there is still a real risk associated with their use. This risk must be weighed against that of bleeding when a liver biopsy is contemplated in a patient with chronic liver disease who has a coagulation defect. Impaired clearance of activated clotting factors20 and reduced synthesis of naturally occurring inhibiiorsl3 facilitate intravascular coagulation in patients with liver disease. In our series, there was no change in those tests that reflect the occurrence of intravascular coagulation fplatelet-count, plasma fibrinogen, factors v and VIII, and serum-F.D.p.) immediately and up to four hours after the infusion. It was also reassuring to find that the infusion of concentrates which might contain clotting factors in activated form was not accompanied by a decrease of antithrombin in (anti Xa). The administration of F.F.P., which provides a source of inhibitors, should help to reduce the risk of thromboembolism. In our opinion, activated clotting factors contained in
prothrombin-complex concentrates are unlikely to cause thromboembolic complications unless these preparations are used in patients who are severely ill and have other underlying conditions facilitating the development of intravascular coagulation (such as major surgery, hsmorrhagic shock, and coma). Thus, concentrates together with F.F.P. may be used before biopsy, thereby allowing the procedure to be carried out safely in pa’.icnts with coagulation defects. thank Dr D. C. B. Mills for his help in the preparation of the ".s:iis:npt.
Requests for reprints should be addressed to P. M. M., Haemophilia and Thrombosis Cent.-e, Angelo Bianchi Bonomi, Via Pace 15, 20122 Milan, Italy. REFERENCES 1. Roberts, R. H., Cederbaum, A. I. Gastroenterology, 1972, 63, 297. 2. Alderson, M. R., Poller, L., Thomson, J. M. J. clin. Path. 1970, 23, 281. 3. Matchett, M. O., Ingram, G. I. C. ibid. 1965, 18, 465. 4. Owren, P. A. Farmakoterapi, 1969, 1-2, 14. 5. Praga, C., Valentini, L., Maidrano, M., Cortellaro, M. in Platelet Function and Thrombosis: A Review of Methods (edited by P. M. Mannucci and S. Gorini) p. 271. New York, 1972. 6. Denson, K. W. E. in Human Blood Coagulation Haemostasis and Thrombosis (edited by R. Biggs); p. 602. Oxford, 1972. 7. Wilson, W. J., G. I. C., Hills, M. Coagulation, 1971, 4, 113. 8. Vermylen, C., De Vreker, R. A., Verstraete, M. Clinica chim. Acta, 1963,
Ingram,
8, 418. Hawiger, D., Niewiarowski, S., Gurewich, V., Thomas, D. P. J. Lab. clin. Med. 1970, 75, 93. 10. Owren, P. A. Lancet, 1959, ii, 754. 11. Yin, E. T., Wessler, S., Stell, P. J. biol. Chem. 1971, 246, 3712. 12. Fagerhol, M. K., Abildgaard, V. Scand. J. Hœmat. 1970, 7, 10. 13. Biggs, R., Denson, K. W. E., Akman, N., Bonett, R., Hadden, M. Br. J. Hœmat. 1970, 19, 283. 14. Ling, C. M., Overby, L. R. J. Immun. 1972, 109, 834. 15. Biggs, R. in Treatment of Haemophilia and other Coagulation Disorders (edited by R. Biggs and R. G. MacFarlane); p. 166. Oxford, 1966. 16. Gazzard, B. G., Henderson, S. M., Williams, R. Gut, 1975, 16, 621. 17. Green, G., Dymock, I. W., Poller, L., Thomson, J. M. Lancet, 1975, i, 1311. 9.
18. Fischer, M. Medsche Welt, 1973, 24, 1899. 19. Vinazzer, H., Bergmann, H. Trans Xth Congr. Wld Fedn Hæmophilia (in the press). 20. Wessler, S., Yin, E. T., Gaston, L. W., Nicol, I., Thromb. Diath. Hœmorr.
1967, 18, 12.
AN AUGMENTED SCHILLING TEST IN THE DIAGNOSIS OF PERNICIOUS ANÆMIA
JAMES A. STRAUCHEN* Department of Medicine, University of California Medical Center, San Francisco, California 94122, U.S.A. A patient with otherwise typical addisonian pernicious anæmia and serum anti-intrinsic-factor antibody failed to respond to oral intrinsic factor on repeated testing during three years of therapy with parenteral vitamin B12. There was no evidence of generalised malabsorption. To test the hypothesis that binding of intrinsic factor to gut secretory anti-
Summary
body
was
responsible,
an
"augmented" Schilling
test
devised using eight times the usual dose of intrinsic factor. This increased dose of intrinsic factor resulted in normal absorption of the test dose of vitamin B12, confirming the diagnosis. It is suggested that the augmented Schilling test may be useful in the diagnosis of the occasional patient with features of pernicious anæmia who fails to respond to conventional doses of intrinsic factor in the Schilling test. was
Introduction
Schilling test has been a major tool in the diagmegaloblastic anaemia since its description by Schilling in 1953.’ The pattern of low excretion of a test dose of labelled vitamin B12 with an increase to normal after administration of intrinsic factor is diagnostic of addisonian pernicious anaemia, and serves to distinguish this condition from other causes of vitamin B12 malabsorption. A variety of autoantibodies has been described in patients with pernicious anaemia including anti-inTHE
nosis of
*Present address: Laboratory of Pathology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20014.
546 trinsic-factor antibodies of both binding and blocking type.23 These antibodies might have a role in the pathoof the disorder by interfering with the secretion and function of intrinsic factor and may also interfere with the Schilling test.4 The present report describes a patient with otherwise classic addisonian pernicious anaemia who consistently failed to respond to oral intrinsic factor on repeated testing over a three-year period in association with serum anti-intrinsic-factor antibody. This patient did, however, show normal excretion of a test dose of labelled vitamin B12 in an "augmented" Schilling test using eight times the usual dose of oral intrinsic factor. This may be a useful approach in the diagnosis of pernicious anaemia in the presence of antiintrinsic-factor antibody.
genesis
mg/dl
and serum-vitamin B12 was 480 pg/ml. ErythrocB sedimentation-rate was 6 mm/h." Serum-calcium was 9f
mg/dl, phosphorus 3-8mg/dl, glucose 84 mg/dl, blood-urea-ru. trogen 11 mg/dl, and creatinine 0.8mg/dl. Total protein was 7.7g/dl, albumin was 4.66 g/dl, serum-cholesterol 160mi. serum-carotene 72 mg/dl, serum-thyroxine 4.0 g/dt, and tn. iodothyronine resin 29%. Urinalysis was normal. Stool fat was less than 5 g/24 h. Stool examination for ova and parasites was negative. Serum was positive for both antiparietat-teti and anti-intrinsic-factor antibodies, but negative for antinuclear antibody. Stimulated gastric analysis with ametazole hydro. chloride (’Histalog’) (1-7mg/kg subcutaneously) demonstrated histamine-fast achlorhydria with gastric fluid pH greater than 7-0in all specimens. A standard Schilling test with 1 N.F, unit intrinsic factor showed less than 1% excretion of the test dose in 48 h. A repeat Schilling test with 8 N.F. units intrinsic factor (eight times the standard dose) showed 13% excretion in 48 h
(see accompanying table). Methods and Materials RESULTS OF SCHILLING TESTS
Schilling tests were performed in a standard manner with a test dose of 0.7 p.Ci of 57Co-cyanocobalamin, followed by an intramuscular flushing dose of 1 mg unlabelled cyanocobalamin. Urine was collected for 24 h before the test and for the following 48 h. Hog intrinsic-factor of standardised potency (E.R. Squibb Pharmaceutical Co. 1 National Formulary (N.F.) unit equals approximately 50 mg) was administered orally in a dose of 1 N.F. unit in the ordinary Schilling test and in a dose of 8 N.F. units in the augmented test. The anti-intrinsic-factor antibody test was done by Dr Hugh Fudenberg.
Case-history The patient is a 37-year-old White female admitted for evaluation ofvitamin-B12 deficiency. The patient has a longstanding history of seizure disorder for which she was maintained on phenytoin, (’Epanutin’ [’Dilantin’]) phenobarbitone, and primidone (’Mysoline’). Ansemia was first noted following her second pregnancy in 1959 and was said to have responded to iron therapy. In 1970 she presented with generalised weakness and severe anaemia with Hb 4.3 g/dl. No glossitis or neurological signs were noted. Evaluation revealed a megaloblastic bone-marrow. Reticulocyte count, Coombs’ test, serum-iron/ total iron binding capacity, Hb A2, F, and electrophoresis, and serum-haptoglobin were normal. Increased serum-folate of 35 ng/ml and markedly decreased serum-vitamin-B12 of 70 pg/ml was noted. Schilling test showed less than 1% excretion without intrinsic factor, and 2% excretion with intrinsic factor (normal greater than 7%). Pernicious ansemia was suspected and the patient began monthly injections of vitamin B12 resulting in a brisk reticulocyte response, and progressive increase in Hb and hsematocrit to normal levels. In 1972 haematocrit was 37%, serum-iron/total iron binding capacity was 35/360 mg/dl, and oral iron supplements were added to the regimen. A repeat Schilling test at this time again showed less than 1% excretion with no increase with intrinsic factor. Because of the persistently abnormal Schilling test with intrinsic factor suggesting malabsorption rather than pernicious anaemia the patient was admitted to the University of California Hospital for further evaluation. There was no history of weight-loss, diarrhoea, steatorrhsra, parasitic infection, or abdominal complaints. Physical examination revealed no significant findings. There was normal papillation of the tongue and normal neurological examination. X-ray examinations of the chest and abdomen were within normal limits. Upper gastrointestinal and small bowel series showed no evidence of malabsorption or other significant finding. Laboratory studies revealed Hb 12.1 g/dl, haematocrit 36%, red-blood-cell count 5 210 000/mm3, reticulocytes 0-6%, platelets 245 000/mm3, white-blood-cell count 4300/mm3 with 47% neutrophils, 42% lymphocytes, 2% monocytes, and 1% eosino-
phils.
The
serum-iron/level
iron
binding capacity
was
63/280
On the basis of histamine-fast achlorhydria, positive serum antiparietal-cell, and anti-intrinsic-factor antibodies, and response to an augmented dose of intrinsic factor in the Schilling test a diagnosis of pernicious anaemia was confirmed and the patient discharged to continue monthly injections of vitamin B1.
Discussion This patient was evaluated because a repeatedly abnormal Schilling test with intrinsic factor raised the question of generalised malabsorption as the cause of her vitamin-B12 deficiency rather than pernicious anaemia. The response to intrinsic factor had remained abnormal despite continuous prolonged therapy with parental vitamin B12 ruling out secondary intestinal atrophy as the cause of her persistently abnormal Schilling test. Screening tests for malabsorption including small bowel X-ray series, faecal fat, stool ova, and parasite examination, and determinations of serum carotene, calcium, and cholesterol were all within normal limits. The diagnosis of pernicious anaemia, however, was strongly supported by the findings of histamine-fast achlorhydria and positive serum tests for antiparietalcell and anti-intrinsic-factor antibodies. The finding of serum anti-intrinsic-factor antibody suggested that the failure to respond to oral intrinsic factor in the Schilling test might be due to the presence of secretory anti-intrinsic-factor antibody in the gut. It has been shown that patients with pernicious anaemia may have secretorya well as serum anti-intrinsic-factor antibodies, both of the binding and blocking types, the latter of which blocks the binding of vitamin B12 to intrinsic factor preventing ileal absorption.4An augmented Schilling test was therefore devised, using a larger dose of intrinsi factor in order to saturate any secretory antibody present. Administration of eight times the standard dose c: intrinsic factor resulted in excretion of 13% of t labelled dose of vitamin B12’ whereas administration : the standard dose had resulted in excretion of less thar 1% (table). This response to an augmented Schillings
547
confirmed the diagnosis of pernicious anaemia in this patient. Unfortunately, the. patient refused a second nasogastric aspiration for direct assay of secretory antibody in the gastric fluid. I suggest that the
augmented Schilling test may be a useful diagnostic tool in the occasional patient with features of pernicious anaemia who fails to respond to conventional doses of intrinsic factor. I am grateful to Dr Stephen B. Shohet for kindly reviewing the
,
manuscript. REFERENCES
1. Schilling, lab. clin. Med. 1953, 42, 860 2. Ardeman, S., Chanarin, I. Lancet, 1963, ii, 1350. 3. Schade, S. G., Abels, J., Schilling, R. F. J. clin. Invest. 1967, 46, 615. 4. Fisher, J. M., Rees, C., Taylor, K. B. Lancet, 1966, ii, 88. 5. Bardhan, K. D., Hall, J. R., Spray, G. H., Callender, S. T. E. ibid. 1968, R. F. J.
ii, 62.
Hypothesis INTERACTION OF HLA MOLECULES WITH NON-IMMUNOLOGICAL LIGANDS AS AN EXPLANATION OF HLA AND DISEASE ASSOCIATIONS A.
L. P. RYDER SVEJGAARD Tissue Typing Laboratory, Blood Grouping Department, State University Hospital (Rigshospitalet), Copenhagen, Denmark
It is postulated that alloantigens (HLA in particular) may interfere with ligand/ receptor interactions not directly involved in immune reactions. This interference may be regarded as a sideeffect of the creation of new mutant HLA alleles which are maintained in the population for other reasons. As an example, it is suggested that some HLA molecules may have structures resembling those on receptors for certain hormones which could cause competition between HLA and receptor molecules for the hormone. Facilitation of ligand/receptor interaction may also be envisaged. Such interactions could under certain conditions lead to disease and might explain some associations between HLA and non-immunological diseases (e.g., manic-depressive disease and hæmachromatosis).
Summary
INTRODUCTION
MANY diseases are more common in individuals carrying certain antigens of the HLA systeml 2-i.e., the major histocompatibility system in man. Genetic systems resembling the HLA system have been found in all vertebrates investigated so far. Such systems control (i) strong alloantigens, (ii) many immune responses, and iti) some components of the complement cascade.3-6 We have called them ARMS (antigen/receptor/mediator system) systems6 because immune-response (Ir) determinants may be regarded as receptors for antigens. So far, the associations between HLA and disease have mainly been demonstrated for some HLA-B and HLA-D (mixed lymphocyte culture [M.L.C.]) determinants, but there is no proof that these antigens are causally responsible for disease susceptibilities. The as-
sociations found may be secondary to primary associations with other HLA factors for which we cannot type at the moment-e.g., Ir determinants, and which are thought to be in strong linkage disequilibrium with factors which have been studied so far-HLA-A, B, and D determinants. Several theories have been put forward to explain the associations. 1 7-9 They could be due to (i) the adverse effects of specific Ir genes’ or "lacunar immunodeficiencies";10 (ii) molecular mimicry7 (i.e., resemblance between some microbial antigens and some HLA antigens); (iii) a receptor function of HLA factors for certain viruses; (iv) an abnormal function of one or more of the complement components coded by the HLA system; and/or (v) inadequately operating differentiation antigens8 (some HLA antigens [e.g., HLA-D which may be identical to B-lymphocyte antigens] are differentiation antigens in the sense that they are present only on some cell types). The first of these theories is likely to be the explanation for many of the associations observed so far. Molecular mimicry has not yet been shown to operate, while the receptor theory is receiving increasing attention due to the observation of Doherty and Zinkernagel11 that virus-induced alterations of selfantigens are important for the immunological elimination of virus-infected cells. Complement deficiencies may cause associations, and our knowledge of differentiation antigens is yet too limited to allow more than speculation. We suggested that ARMS, apart from dealing with foreign matter, may interfere with the interaction between hormones and receptors in the body.69 Increasing evidence supports this assumption which gives rise to the following hypothesis which could be tested. HYPOTHESISS
Within the term HLA
we
include all factors coded for
genes in linkage disequilibrium with the HLA system-i.e., both the classic HLA-A, B, and C antigens; the HLA-D antigens; some complement factors; the B-cell antigens; and other as yet unknown factors (e.g., homologues of the Tia, Ly, and other H-2 linked antigens in mice). Some HLA determinants-and perhaps other surface and fluid phase molecules---may interfere with the interaction between ligands (e.g., hormones) and their receptors on cell surfaces. One reason to believe this hypothesis is that HLA is itself a ligand/receptor system,
by
and with the considerable number (probably over 100) different HLA factors in the population, it is quite possible that some of these would be non-neutral in relation to other ligand/receptor interactions. Interference with the interactions may arise by various mechanisms, and one of the more simple ones is outlined in the accompanying figure. If an HLA antigen has some incidental resemblance to the binding site of a cell-surface receptor molecule for a given hormone, there could be competition between the receptor and the HLA antigen molecule. Naturally, the affinity between the ligand and its receptor must be much stronger than that between ligand and the HLA antigen, but since the receptor is present only on the specific target cells, whereas many HLA antigens are present on all cells and even in body fluids, the affinity of HLA for ligand need perhaps not