1549
had HLA
HIV-1 viraemia and influenza SIR,-Cellular activation of CD4 lymphocytes is required for HIV-1 replication in vitro and it is often assumed that immune stimulation due to infection, vaccination, or allogeneic exposure can result in enhanced HIV-1replication in vivo. I have examined this possibility by serial measurement of viral burden in blood from three HIV-1seropositive, symptomless individuals who had acute influenza infection or influenza vaccination. Infectious titres of HIV-1in plasma and peripheral blood mononuclear cells (PBMC), measured by end-point-dilution cultures,! increased transiently in patient 1 who had an acute illness clinically diagnosed as influenza. The levels of infectious HIV-1in plasma increased by 20-fold in the two given influenza A vaccine:
antigens that were rarely seen in the Japanese group. In particular, the class II antigen DRI was absent in the Japanese patients with insulin autoimmune syndrome, and HLA-DR2 was found in as few as 3 of them. Thus, the high prevalence of HLA-DR4 (DRB1*0406) in Japanese patients may, at least in part, indicate the high prevalence of that antigen in this ethnic group. Division of
Endocrinology and Metabolism, University Hospital, 4031 Basel, Switzerland
1. Sklenar I, Wilkin TJ, Diaz JL, Erb P, Keller U. Spontaneous hypoglycemia associated with autoimmunity specific to human insulin. Diabetes Care 1987; 10: 152-59.
Mazzotti test for onchocerciasis
HIV- 7 titre
Case
Day
G. A. SPINAS U. KELLER
Plasma
PBMC
(TCID/ml)
(’7’C/D//0"ce//
SjR,—The Mazzotti test is a dangerous, provocative test that has been used to detect onchocerciasis but is condemned by the World Health Organisation expert committee on onchocerciasis.l It came into vogue 40 years ago as an easy way to detect this disease. In developing areas, one could always give a patient some diethylcarbamazine (DEC), and if they became sick they had "oncho". In many ways it was easier and cheaper than doing skin
snips.
TCID=tissue-culture mfective doses.
This rise in viraemia was most prominent in the first week following immunisation, and viral titres had returned to or toward baseline by 2-3 weeks. An increase in HIV-1titre in PBMC was seen in only one of two cases. The transiently higher levels of viraemia were not associated with a dramatic decline in the CD4 cell counts. These findings are preliminary and should not be used to make clinical decisions about influenza vaccinations in HIV-1-infected persons. However, they should trigger more systematic studies of the acute effect of different infections or immunisations on HIV-1I replication in vivo. The impact of allogeneic exposures from blood transfusions on viral burden should also be determined. Aaron Diamond AIDS Research Center, NYU School of Medicine, New York, NY 10016, USA
DAVID D. Ho
1. Ho DD, Moudgil T, Alam M. Quantitation of human immunodeficiency in the blood of infected persons. N Engl J Med 1989; 321: 1621-25.
virus type 1
Insulin autoimmune syndrome and HLA-DR4 SIR,-Dr Uchigata and colleagues (Feb 15, p 393) demonstrate that all of 27 Japanese patients with insulin autoimmune syndrome were positive for HLA-DR4 (DRBI*0406), suggesting that the development of the syndrome is linked to the presence of specific HLA class II molecules. We have evidence that this may not be so for other ethnic groups. In 1986, a 55-year-old Swiss woman was referred to our hospital for evaluation of recurrent spontaneous hypoglycaemic attacks that usually took place 1-3 h after meal ingestion, were improved by food intake, and were precipitated by exercise. Fasting plasma total insulin (1 -22 nmol/1) and proinsulin (0-48 nmol/1) were raised, and human insulin specific autoantibodies were detected both by radioimmunoasay (RIA) and by an IgG-specific ELISA. In the RIA, the patient’s serum bound 49-3% of human, but only 3-7% of added pork, insulin tracer, and in the ELISA 352% (% of reference serum binding) human insulin and 2-6% pork insulin were bound. Further analysis revealed a monotypic and monoclonal human insulin autoantibody, which showed a restriction to the lambda light chain. On the basis of these findings insulin autoimmune syndrome
diagnosed.’ HLA-typing of this patient by standard microlymphocytic toxicity tests showed the following HLA haplotype: A30, 32; B27, 62; DR 1,2. Thus, apart from the class I antigen B62, which was present in 22 of the 27 patients reported by Uchigata, our patient was
We now know that the use of DEC and provoking a Mazzotti reaction is not without risk. It can cause severe acute illnesses, irreversible long-term sequelae, and at times even death.l-4 These reactions are not manifestations of direct drug toxicity but are related to the death of microfilariae. However, the severity of the reaction is not closely linked to the intensity of infection, nor will every infected person have a reaction. In double-masked, controlled clinical trials, those having a Mazzotti reaction after DEC are clearly different from those who receive placebo therapy, or, for that matter, ivermectin.5 I was surprised to read Dr Keystone and Dr Davies’ report (March 14, p 678) in which they recommend this test and describe a single-blind method for its administration. In patients for whom they consider the diagnosis of onchocerciasis, they should make a definitive diagnosis by taking a set of skin snips. If the skin snips are negative they can be repeated.6 The collection of skin snips is straightforward. It needs only simple equipment and is defmitive when positive. The "trial by fire" approach that the Mazzotti test exemplifies seems hard to justify. Department of Ophthalmology, University of Melbourne, Melbourne, Victoria, Australia 3002
HUGH R. TAYLOR
1. World Health Organisation Expert Committee on Onchocerdasis: third report. Geneva: WHO. Tech Rep Ser 752, 1987: 55-61. 2. Bird AC, El-Sheikh A, Anderson J, Fuglsang H. Changes in visual function and in the posterior segment of the eye during treatment of onchocerciasis with diethylcarbamazine citrate. Br J Ophthalmol 1980; 64: 191-200. 3. Greene BM, Taylor HR, Humphrey RL. Proteinuria associated with diethylcarbamazine treatment of onchocerciasis. Lancet 1980; i: 254-55. 4. Oomen AP. Fatalities after treatment of onchocerdasis with diethylcarbamazine. Trans R Soc Trop Med Hyg 1969; 63: 548. 5. Greene BM, Taylor HR, Cupp EW, et al. Comparison of ivermectin and diethylcarbamazine in the treatment of onchocerciasis. N Engl J Med 1985; 313:
133-38. 6. Aziz MA, Diallo S, Diop IM, Lariviere M, Porta M. Efficacy and tolerance of ivermectin in human onchocerciasis. Lancet 1982; ii: 171-73.
* ’Thisletter has been shown to Dr Keystone and Dr Davies, whose reply follows.-ED L. SIR,-We agree with Dr Taylor that skin snips are the method of choice to diagnose onchocerciasis and thank him for emphasising the potential dangers of the Mazzotti test, especially in patients with moderate to heavy infections. However, we do not agree that the test is without merit. The diagnosis of early onchocerciasis is difficult. Onchocerca volvulus microfilariae can be absent from multiple skin snips in light or early infections.’ Taylor and his colleagues have confirmed this by showing that the sensitivity of skin snips declines strikingly when microfilaria density drops below 1 per mg skin.2 Repeat of skin snips may not always be possible because of patient intolerance of the procedure, and may be negative yet again.
1550
comments on the dangers of the Mazzotti reaction are substantiated by Francis and colleagues’3 report showing that the severity of the reaction correlates well with intensity of infection.3 Their results suggest to us that the risk of a moderate or severe reaction would be very low in those who are skin-snip
Taylor’s
not
negative. Taylor points out that the World Health Organisation’s committee on onchocerciasis condemns the use of the Mazzotti test.4 Yet, in that report the authors write that the Mazzotti test "should be used only in individuals whom a diagnosis of onchocerciasis is being considered, but in whom parasites cannot be detected in skin, eyes or blood. In such individuals, the Mazzotti test can be extremely helpful by inducing local pruritis and rash, clinically indicative of the parasite’s presence".’ We agree. Disease Unit, Toronto Hospital, Toronto, Ontario M5G 2C4, Canada, UK; Division of Infectious Diseases, Hospital for Sick Children, Toronto
Tropical
JAY S. KEYSTONE DELE DAVIES
1. Editorial. Diagnosis of early onchocerciasis. Lancet 1991; 337: 1449. 2. Taylor HR, Munoz B, Keyvan-Larijani E, Green BM. Reliability of detection of
microfilarariae in skin snips in the diagnosis of onchocerciasis. Am J Trop Med Hyg 1989; 41: 467-71. 3. Francis H, Awadzi K, Ottesen EA. The Mazzotti reaction following treatment of onchocerciasis with diethyl-carbamazine: clinical severity as a function of infection intensity. Am J Trop Med Hyg 1985; 34: 529-36. 4. WHO Expert Committee on Oncerciasis. Third report. Geneva: World Health Organisation. Tech Rep Ser 1987; 752: 60-61.
Factor VIII and factor IX inhibitors in
haemophiliacs SIR,-Dr Ehrenforth and colleagues report (March 7, p 594) that of 59 haemophiliac children born since 1970 attending their centre (and treated at least once),52% of those with severe and 5% of those with moderate haemophilia A had developed factor VIII inhibitors (antibodies), whereas none of those with haemophilia B had factor IX inhibitors. By contrast, Dr Ciavarella and Dr Schiavoni (May 23, p 1301) report that in 21 previously untreated patients the incidence of inhibitor development after treatment with FVIII concentrates of intermediate or high purity was as low as about 5 %; Dr Verbruggen and colleagues (p 1301) point out that the Bethesda assay Ehrenforth used tended to overestimate antibody activity, and therefore the incidence of FVIII inhibitors. During the period 1970-90 (ie, roughly that covered by Ehrenforth’s report), at the three Swedish centres (which treat all haemophiliacs in the country), 117 newborn babies with severe or moderate haemophilia were registered, 81 with haemophilia A (77 severe, 24 moderate), and 16 with haemophilia B (12 severe, 4 moderate), which is much the same proportion of severity as Ehrenforth’s population. Inhibitors developed in 21% (16/77) of those with severe haemophilia A (and at low titre in 1 child with mild disease, which have disappeared spontaneously), and in 33% (4/12) of those with severe haemophilia B; most of these children were less than 2 years of age at inhibitor development, and all were less than 7. The inhibitors were of the high response type in 8/20 cases, and of the low response type in the remaining 12 cases (titre < 10 Bethesda units, and no anamnestic response). Of the 6 patients with high-response factor VIII antibodies, 2 have been converted to low responders and are now on daily high-dose factor VIII treatment according to the Bonn protocol, and 4 underwent the Malimo treatment protocol for the induction of immune tolerance,2.3 the antibodies disappearing in 2 of them. Of the 10 patients with low-titre factor VIII antibodies, the inhibitors have disappeared in 6 (spontaneously in 5, and after immune tolerance induction in 1), enabling them to be put on prophylactic treatment ; no specific treatment has been prescribed in 2, and the remaining 2 have only just been detected and treatment is not yet decided. The 2 patients with high-titre factor IX antibodies underwent the Malmo treatment protocol, with disappearance of the antibodies in 1. The 2 patients with low-titre factor IX antibodies are siblings, 1 of whom has had anaphylactoid reactions to two factor IX preparations. 1 patient with high-titre factor IX antibodies has died.
Our findings with respect to age at development of inhibitors and the amount and duration of replacement therapy are consistent with those of Ehrenforth and colleagues, though the frequency of inhibitor development among haemophilia A patients is lower in our series (21% vs 52%). Our patients are not a subgroup in a prospective study of inhibitor development, but represent the entire national population of haemophiliacs bom in 1970-90, thus excluding any selection bias that might have been a confounding factor. Moreover, in all severe and many moderate cases of haemophilia, prophylactic treatment is started at an early age in Sweden (about age 4 in the early 1970s, and at 1-1-1now), and inhibitor testing and factor recovery analysis are done routinely 1-3 times a year 4 The few patients still untreated do not decisively affect our statistics. The mutation has been characterised in about 60% of all haemophilia B families in Sweden, and in all six families in whom a family member has developed inhibitors.s Fifteen families with severe or moderate haemophilia B have a mutation causing gross physical or functional loss of coding information; in six of these families, factor IX inhibitor has developed in a family member. By contrast, none of the severe or moderate haemophilia B patients with mutations resulting in single aminoacid substitutions has developed antibodies. Although the effect of the type of factor VIII mutation on the development of factor VIII inhibitors has yet to be elucidated, data so far suggest that it may be a determinant of antibody development. In all likelihood, where there is a high prevalence of certain mutations in a specific catchment area there will be a correspondingly high prevalence of antibody development. This study was supported by grants from The Medical Faculty, University of Lund, Sweden, and the Swedish Medical Research Council (grant 0087).
Departments of Coagulation Disorders at Malmo General Hospital, S21401 Sweden; Karolinska Hospital, Stockholm; and Sahlgrenska Hospital, Gothenburg
ROLF LJUNG
PIA PETRINI ANN-CHRISTINE LINDGREN LILIAN TENGBORN INGA MARIE NILSSON
1. Brackmann HH. Induced immunotolerance in factor VIII inhibitor patients. Progr Clin Biol Res 1984; 150: 181-85. 2. Nilsson IM, Bemtorp E, Zettervall O. Induction of split tolerance and clinical cure in high-responding hemophiliacs with factor IX antibodies. Proc Natl Acad Sci USA
1986; 83: 9169-73. 3. Nilsson IM, Bemtorp E, Zettervall O. Induction of immune tolerance in patients with hemophilia and antibodies to factor VIII by combined treatment with intravenous IgG, cyclophosphamide, and factor VIII. N Engl J Med 1988; 318: 947-50. 4. Nilsson IM, Bemtorp E, Löfqvist T, Pettersson H. Twenty-five years’ experience of prophylactic treatment in severe haemophilia A and B. J Int Med (in press). 5. Green PM, Montandon AJ, Ljung R, et al. Haemophilia B mutations in a complete Swedish population sample: a test of a new strategy for the genetic counselling of diseases with high mutational heterogeneity. Br J Haematol 1991; 78: 390-97.
SIR,-Dr Ehrenforth and colleagues report a high incidence of to factor VIII (FVIII) in patients with severe and haemophilia. 15 of their 46 haemophilia A patients developed inhibitors. The incidence was 39 per 1000 patient-years and the aged-dependent cumulative risk was 33 % by age 6 years. As Ehrenforth et al state, these figures are higher than those reported previously. We agree that earlier studies have probably underestimated the risk of acquiring FVIII inhibitors and have produced conflicting results because they used prevalence (rather than incidence), data have been retrospective, patients with mild haemophilia have been included, and ages of the study groups have inhibitors moderate
differed. We have retrospectively reviewed data for 57 patients with haemophilia A born in 1975 or after. All patients had been periodically controlled since their diagnosis and treated at least once with commercial FVIII concentrates. Almost all patients were tested for inhibitors at least once a year and whenever their presence was suspected. In patients in whom an inhibitor was detected, antibody levels were measured on each subsequent visit. Sporadic, low titre, inhibitors ( < 5 Bethesda units) were found in 14 patients, but clinical evidence was noted only once or twice, despite repeated exposure to FVIII. We did not include these cases in our incidence calculations, since other similar cases could have been overlooked. Persistent inhibitors developed in 10 of 52 patients with severe or moderate haemophilia (FVIIIC < 0’05 IU/ml), resulting in an incidence of 20 per 1000 patient-years. In all cases, inhibitor
had HLA
HIV-1 viraemia and influenza SIR,-Cellular activation of CD4 lymphocytes is required for HIV-1 replication in vitro and it is often assumed that immune stimulation due to infection, vaccination, or allogeneic exposure can result in enhanced HIV-1replication in vivo. I have examined this possibility by serial measurement of viral burden in blood from three HIV-1seropositive, symptomless individuals who had acute influenza infection or influenza vaccination. Infectious titres of HIV-1in plasma and peripheral blood mononuclear cells (PBMC), measured by end-point-dilution cultures,! increased transiently in patient 1 who had an acute illness clinically diagnosed as influenza. The levels of infectious HIV-1in plasma increased by 20-fold in the two given influenza A vaccine:
antigens that were rarely seen in the Japanese group. In particular, the class II antigen DRI was absent in the Japanese patients with insulin autoimmune syndrome, and HLA-DR2 was found in as few as 3 of them. Thus, the high prevalence of HLA-DR4 (DRB1*0406) in Japanese patients may, at least in part, indicate the high prevalence of that antigen in this ethnic group. Division of
Endocrinology and Metabolism, University Hospital, 4031 Basel, Switzerland
1. Sklenar I, Wilkin TJ, Diaz JL, Erb P, Keller U. Spontaneous hypoglycemia associated with autoimmunity specific to human insulin. Diabetes Care 1987; 10: 152-59.
Mazzotti test for onchocerciasis
HIV- 7 titre
Case
Day
G. A. SPINAS U. KELLER
Plasma
PBMC
(TCID/ml)
(’7’C/D//0"ce//
SjR,—The Mazzotti test is a dangerous, provocative test that has been used to detect onchocerciasis but is condemned by the World Health Organisation expert committee on onchocerciasis.l It came into vogue 40 years ago as an easy way to detect this disease. In developing areas, one could always give a patient some diethylcarbamazine (DEC), and if they became sick they had "oncho". In many ways it was easier and cheaper than doing skin
snips.
TCID=tissue-culture mfective doses.
This rise in viraemia was most prominent in the first week following immunisation, and viral titres had returned to or toward baseline by 2-3 weeks. An increase in HIV-1titre in PBMC was seen in only one of two cases. The transiently higher levels of viraemia were not associated with a dramatic decline in the CD4 cell counts. These findings are preliminary and should not be used to make clinical decisions about influenza vaccinations in HIV-1-infected persons. However, they should trigger more systematic studies of the acute effect of different infections or immunisations on HIV-1I replication in vivo. The impact of allogeneic exposures from blood transfusions on viral burden should also be determined. Aaron Diamond AIDS Research Center, NYU School of Medicine, New York, NY 10016, USA
DAVID D. Ho
1. Ho DD, Moudgil T, Alam M. Quantitation of human immunodeficiency in the blood of infected persons. N Engl J Med 1989; 321: 1621-25.
virus type 1
Insulin autoimmune syndrome and HLA-DR4 SIR,-Dr Uchigata and colleagues (Feb 15, p 393) demonstrate that all of 27 Japanese patients with insulin autoimmune syndrome were positive for HLA-DR4 (DRBI*0406), suggesting that the development of the syndrome is linked to the presence of specific HLA class II molecules. We have evidence that this may not be so for other ethnic groups. In 1986, a 55-year-old Swiss woman was referred to our hospital for evaluation of recurrent spontaneous hypoglycaemic attacks that usually took place 1-3 h after meal ingestion, were improved by food intake, and were precipitated by exercise. Fasting plasma total insulin (1 -22 nmol/1) and proinsulin (0-48 nmol/1) were raised, and human insulin specific autoantibodies were detected both by radioimmunoasay (RIA) and by an IgG-specific ELISA. In the RIA, the patient’s serum bound 49-3% of human, but only 3-7% of added pork, insulin tracer, and in the ELISA 352% (% of reference serum binding) human insulin and 2-6% pork insulin were bound. Further analysis revealed a monotypic and monoclonal human insulin autoantibody, which showed a restriction to the lambda light chain. On the basis of these findings insulin autoimmune syndrome
diagnosed.’ HLA-typing of this patient by standard microlymphocytic toxicity tests showed the following HLA haplotype: A30, 32; B27, 62; DR 1,2. Thus, apart from the class I antigen B62, which was present in 22 of the 27 patients reported by Uchigata, our patient was
We now know that the use of DEC and provoking a Mazzotti reaction is not without risk. It can cause severe acute illnesses, irreversible long-term sequelae, and at times even death.l-4 These reactions are not manifestations of direct drug toxicity but are related to the death of microfilariae. However, the severity of the reaction is not closely linked to the intensity of infection, nor will every infected person have a reaction. In double-masked, controlled clinical trials, those having a Mazzotti reaction after DEC are clearly different from those who receive placebo therapy, or, for that matter, ivermectin.5 I was surprised to read Dr Keystone and Dr Davies’ report (March 14, p 678) in which they recommend this test and describe a single-blind method for its administration. In patients for whom they consider the diagnosis of onchocerciasis, they should make a definitive diagnosis by taking a set of skin snips. If the skin snips are negative they can be repeated.6 The collection of skin snips is straightforward. It needs only simple equipment and is defmitive when positive. The "trial by fire" approach that the Mazzotti test exemplifies seems hard to justify. Department of Ophthalmology, University of Melbourne, Melbourne, Victoria, Australia 3002
HUGH R. TAYLOR
1. World Health Organisation Expert Committee on Onchocerdasis: third report. Geneva: WHO. Tech Rep Ser 752, 1987: 55-61. 2. Bird AC, El-Sheikh A, Anderson J, Fuglsang H. Changes in visual function and in the posterior segment of the eye during treatment of onchocerciasis with diethylcarbamazine citrate. Br J Ophthalmol 1980; 64: 191-200. 3. Greene BM, Taylor HR, Humphrey RL. Proteinuria associated with diethylcarbamazine treatment of onchocerciasis. Lancet 1980; i: 254-55. 4. Oomen AP. Fatalities after treatment of onchocerdasis with diethylcarbamazine. Trans R Soc Trop Med Hyg 1969; 63: 548. 5. Greene BM, Taylor HR, Cupp EW, et al. Comparison of ivermectin and diethylcarbamazine in the treatment of onchocerciasis. N Engl J Med 1985; 313:
133-38. 6. Aziz MA, Diallo S, Diop IM, Lariviere M, Porta M. Efficacy and tolerance of ivermectin in human onchocerciasis. Lancet 1982; ii: 171-73.
* ’Thisletter has been shown to Dr Keystone and Dr Davies, whose reply follows.-ED L. SIR,-We agree with Dr Taylor that skin snips are the method of choice to diagnose onchocerciasis and thank him for emphasising the potential dangers of the Mazzotti test, especially in patients with moderate to heavy infections. However, we do not agree that the test is without merit. The diagnosis of early onchocerciasis is difficult. Onchocerca volvulus microfilariae can be absent from multiple skin snips in light or early infections.’ Taylor and his colleagues have confirmed this by showing that the sensitivity of skin snips declines strikingly when microfilaria density drops below 1 per mg skin.2 Repeat of skin snips may not always be possible because of patient intolerance of the procedure, and may be negative yet again.
1550
comments on the dangers of the Mazzotti reaction are substantiated by Francis and colleagues’3 report showing that the severity of the reaction correlates well with intensity of infection.3 Their results suggest to us that the risk of a moderate or severe reaction would be very low in those who are skin-snip
Taylor’s
not
negative. Taylor points out that the World Health Organisation’s committee on onchocerciasis condemns the use of the Mazzotti test.4 Yet, in that report the authors write that the Mazzotti test "should be used only in individuals whom a diagnosis of onchocerciasis is being considered, but in whom parasites cannot be detected in skin, eyes or blood. In such individuals, the Mazzotti test can be extremely helpful by inducing local pruritis and rash, clinically indicative of the parasite’s presence".’ We agree. Disease Unit, Toronto Hospital, Toronto, Ontario M5G 2C4, Canada, UK; Division of Infectious Diseases, Hospital for Sick Children, Toronto
Tropical
JAY S. KEYSTONE DELE DAVIES
1. Editorial. Diagnosis of early onchocerciasis. Lancet 1991; 337: 1449. 2. Taylor HR, Munoz B, Keyvan-Larijani E, Green BM. Reliability of detection of
microfilarariae in skin snips in the diagnosis of onchocerciasis. Am J Trop Med Hyg 1989; 41: 467-71. 3. Francis H, Awadzi K, Ottesen EA. The Mazzotti reaction following treatment of onchocerciasis with diethyl-carbamazine: clinical severity as a function of infection intensity. Am J Trop Med Hyg 1985; 34: 529-36. 4. WHO Expert Committee on Oncerciasis. Third report. Geneva: World Health Organisation. Tech Rep Ser 1987; 752: 60-61.
Factor VIII and factor IX inhibitors in
haemophiliacs SIR,-Dr Ehrenforth and colleagues report (March 7, p 594) that of 59 haemophiliac children born since 1970 attending their centre (and treated at least once),52% of those with severe and 5% of those with moderate haemophilia A had developed factor VIII inhibitors (antibodies), whereas none of those with haemophilia B had factor IX inhibitors. By contrast, Dr Ciavarella and Dr Schiavoni (May 23, p 1301) report that in 21 previously untreated patients the incidence of inhibitor development after treatment with FVIII concentrates of intermediate or high purity was as low as about 5 %; Dr Verbruggen and colleagues (p 1301) point out that the Bethesda assay Ehrenforth used tended to overestimate antibody activity, and therefore the incidence of FVIII inhibitors. During the period 1970-90 (ie, roughly that covered by Ehrenforth’s report), at the three Swedish centres (which treat all haemophiliacs in the country), 117 newborn babies with severe or moderate haemophilia were registered, 81 with haemophilia A (77 severe, 24 moderate), and 16 with haemophilia B (12 severe, 4 moderate), which is much the same proportion of severity as Ehrenforth’s population. Inhibitors developed in 21% (16/77) of those with severe haemophilia A (and at low titre in 1 child with mild disease, which have disappeared spontaneously), and in 33% (4/12) of those with severe haemophilia B; most of these children were less than 2 years of age at inhibitor development, and all were less than 7. The inhibitors were of the high response type in 8/20 cases, and of the low response type in the remaining 12 cases (titre < 10 Bethesda units, and no anamnestic response). Of the 6 patients with high-response factor VIII antibodies, 2 have been converted to low responders and are now on daily high-dose factor VIII treatment according to the Bonn protocol, and 4 underwent the Malimo treatment protocol for the induction of immune tolerance,2.3 the antibodies disappearing in 2 of them. Of the 10 patients with low-titre factor VIII antibodies, the inhibitors have disappeared in 6 (spontaneously in 5, and after immune tolerance induction in 1), enabling them to be put on prophylactic treatment ; no specific treatment has been prescribed in 2, and the remaining 2 have only just been detected and treatment is not yet decided. The 2 patients with high-titre factor IX antibodies underwent the Malmo treatment protocol, with disappearance of the antibodies in 1. The 2 patients with low-titre factor IX antibodies are siblings, 1 of whom has had anaphylactoid reactions to two factor IX preparations. 1 patient with high-titre factor IX antibodies has died.
Our findings with respect to age at development of inhibitors and the amount and duration of replacement therapy are consistent with those of Ehrenforth and colleagues, though the frequency of inhibitor development among haemophilia A patients is lower in our series (21% vs 52%). Our patients are not a subgroup in a prospective study of inhibitor development, but represent the entire national population of haemophiliacs bom in 1970-90, thus excluding any selection bias that might have been a confounding factor. Moreover, in all severe and many moderate cases of haemophilia, prophylactic treatment is started at an early age in Sweden (about age 4 in the early 1970s, and at 1-1-1now), and inhibitor testing and factor recovery analysis are done routinely 1-3 times a year 4 The few patients still untreated do not decisively affect our statistics. The mutation has been characterised in about 60% of all haemophilia B families in Sweden, and in all six families in whom a family member has developed inhibitors.s Fifteen families with severe or moderate haemophilia B have a mutation causing gross physical or functional loss of coding information; in six of these families, factor IX inhibitor has developed in a family member. By contrast, none of the severe or moderate haemophilia B patients with mutations resulting in single aminoacid substitutions has developed antibodies. Although the effect of the type of factor VIII mutation on the development of factor VIII inhibitors has yet to be elucidated, data so far suggest that it may be a determinant of antibody development. In all likelihood, where there is a high prevalence of certain mutations in a specific catchment area there will be a correspondingly high prevalence of antibody development. This study was supported by grants from The Medical Faculty, University of Lund, Sweden, and the Swedish Medical Research Council (grant 0087).
Departments of Coagulation Disorders at Malmo General Hospital, S21401 Sweden; Karolinska Hospital, Stockholm; and Sahlgrenska Hospital, Gothenburg
ROLF LJUNG
PIA PETRINI ANN-CHRISTINE LINDGREN LILIAN TENGBORN INGA MARIE NILSSON
1. Brackmann HH. Induced immunotolerance in factor VIII inhibitor patients. Progr Clin Biol Res 1984; 150: 181-85. 2. Nilsson IM, Bemtorp E, Zettervall O. Induction of split tolerance and clinical cure in high-responding hemophiliacs with factor IX antibodies. Proc Natl Acad Sci USA
1986; 83: 9169-73. 3. Nilsson IM, Bemtorp E, Zettervall O. Induction of immune tolerance in patients with hemophilia and antibodies to factor VIII by combined treatment with intravenous IgG, cyclophosphamide, and factor VIII. N Engl J Med 1988; 318: 947-50. 4. Nilsson IM, Bemtorp E, Löfqvist T, Pettersson H. Twenty-five years’ experience of prophylactic treatment in severe haemophilia A and B. J Int Med (in press). 5. Green PM, Montandon AJ, Ljung R, et al. Haemophilia B mutations in a complete Swedish population sample: a test of a new strategy for the genetic counselling of diseases with high mutational heterogeneity. Br J Haematol 1991; 78: 390-97.
SIR,-Dr Ehrenforth and colleagues report a high incidence of to factor VIII (FVIII) in patients with severe and haemophilia. 15 of their 46 haemophilia A patients developed inhibitors. The incidence was 39 per 1000 patient-years and the aged-dependent cumulative risk was 33 % by age 6 years. As Ehrenforth et al state, these figures are higher than those reported previously. We agree that earlier studies have probably underestimated the risk of acquiring FVIII inhibitors and have produced conflicting results because they used prevalence (rather than incidence), data have been retrospective, patients with mild haemophilia have been included, and ages of the study groups have inhibitors moderate
differed. We have retrospectively reviewed data for 57 patients with haemophilia A born in 1975 or after. All patients had been periodically controlled since their diagnosis and treated at least once with commercial FVIII concentrates. Almost all patients were tested for inhibitors at least once a year and whenever their presence was suspected. In patients in whom an inhibitor was detected, antibody levels were measured on each subsequent visit. Sporadic, low titre, inhibitors ( < 5 Bethesda units) were found in 14 patients, but clinical evidence was noted only once or twice, despite repeated exposure to FVIII. We did not include these cases in our incidence calculations, since other similar cases could have been overlooked. Persistent inhibitors developed in 10 of 52 patients with severe or moderate haemophilia (FVIIIC < 0’05 IU/ml), resulting in an incidence of 20 per 1000 patient-years. In all cases, inhibitor
