- Email: [email protected]
GASTROINTESTINAL BLEEDING DUE TO VITAMIN K DEFICIENCY IN PATIENTS ON PARENTERAL CEFAMANDOLE
39 tom-free until then and had taken. only small amounts of antacids. Both had also had bleeding ulcers before the start of the one-year maintenance programme, and the patient treated with placebo had had an acute bleeding ulcer during the trial. He was operated on when he bled after the trial. The other patient refused surgery and was given a short course of cimetidine. He remained well but towards the end of the second year he was put on cimetidine again for dyspeptic symptoms, this time without bleeding. When moderate or severe symptoms developed after the end of the trial, they were well controlled by short courses of cimetidine or, in 3 patients, almost continuous treatment (table n). In patients referred for elective surgery because of the severity of symptoms when ulcers recurred after the trial-the symptoms were clearly relieved by cimetidine. In a few patients with moderate dyspeptic symptoms antacid-anticholinergic therapy was given. In patients with mild symptoms only small amounts of antacids were needed. We have reported that, two days and three and a half months after treatment with cimetidine, acid output fell slightly in comparison with pre-treatment output in patients with duodenal ulcers and increased slightly in patients with prepyloric ulcers.1.2 Our follow-up study shows that ulcer-disease is not aggravated when maintenance treatment is stopped. Others have reported that there is the same high relapse-rate in peptic ulcers after one year treatment as there is after shortterm treatment with cimetidine in patients with chronic peptic ulcer disease.33 At the end of the one year trial we did endoscopies only on patients with symptoms and so had no information on the real relapse rate. However, our two-year follow-up findings suggests that peptic-ulcer disease is less severe in patients treated with cimetidine for a year than in those given with placebo. 22 of the 36 patients in the placebo group have been referred for surgery during or after the end of the trial (table II). If the 3 patients who have needed almost continuous treatment with cimetidine and the 1 patient with an acute bleeding ulcer after the trial are deemed to have required surgery, only 7 of the 32 patients in the cimetidine group were (or should have been) operated on (X2=21.4, p<0.001). 8 patients in the cimetidine group are still completely symptom-free two years after the trial compared with 4 patients in the placebo-group
(x,2=6.34, p<0.02).
All patients were advised not to smoke or to take or to abuse any tablet containing aspirin, alcohol, so these factors cannot explain the difference in the severity of peptic-ulcer disease between placebo and cimetidine treated patients. Few patients have changed their working conditions after the end of the study. In our placebo group there was a mean rise of 50% (p<0.05) in non-stimulated acid secretion (BAO) after treatment for seven months with placebo, and basal acid output was 33% higher (not significant) at the end of the twelve months’ treatment. BAO in the cimetidine treated patients with duodenal ulcer was lower 2 days and 3tmonths after the trial lasting a year. One of us also found, in an earlier study, that in patients with chronic peptic ulcer disease there was a mean rise in BAO of 68% (p<0.10) at the end of treatment with placebo for several years but not after treatment with an anticholinergic drug during the same time period.4 These results suggest that, during the years when patients have active peptic ulcer disease, BAO rises steadily. This could explain the high recurrence rate in peptic ulcers in patients
2. Bodemar G, Walan A, Lundquist G. Gastrin concentrations at the end of and following long term treatment with cimetidine and placebo. In: Cretzfeldt W, ed. Cimetidine. Amsterdam: Excerpta Medica, 1978:233-35. 3. Hansky J, Korman MG, Hetzel DJ, Schearman DJC. Relapse rate after cessation of 12 months Cimetidine in duodenal ulcer. Gastroenterology
1979, 76: 1151. on peptic ulcer disease with special reference of 1-hyoscyamine. Acta Med Scand 1970; suppl 516.
4. Walan A. Studies
to
the effect
with chronic disease. Our data suggest that treatment with cimetidine blocks this rise in BAO and favourably alters the course of the peptic ulcer disease when maintenance treatment is ended. Department of Internal Medicine,
University Hospital, S-581 85 Linköping, Sweden
G. BODEMAR A. WALAN
GASTROINTESTINAL BLEEDING DUE TO VITAMIN K DEFICIENCY IN PATIENTS ON PARENTERAL CEFAMANDOLE
SIR, The combination of low vitamin K intake and limited of the vitamin due to suppression of intestinal bacterial flora can produce severe prothrombin deficiency.’ De-
synthesis
creased vitamin K as the cause of a bleeding diathesis has been noted with several broad-spectrum parenteral antibiotics,2-4 and it can be lethal.5 For patients on antibiotics whose oral intake of nutrients is restricted vitamin K therapy may be warranted. During a study of cefamandole sodium in the management of peritoneal and soft tissue sepsis thirty-seven patients with mixed-flora infections were treated by drainage or excision and with cefamandole sodium (average dose 40 mg/kg body-weight 6-hourly, range 500-4500 mg per injection) for an average of 7 days (range 3-16 days). Cefamandole controlled the sepsis in thirty-four of the thirty-six cases with bacterial pathogens. In three patients, however, the prothrombin time was prolonged. All three patients had major gastrointestinal bleeds, and one massive bleed was fatal. All three patients had presented with clinical signs suggestive of peritonitis, but the underlying diagnoses were found to be (A) perforated gastric ulcer, (B) left colon gangrene due to acute mesenteric arterial thrombosis, and (C) abdominal carcinomatosis. Patients A and B underwent emergency laparotomy and were put on cefamandole before surgery. Patient C was treated by nasogastric decompression and cefamandole; surgery was delayed until the third week. All three patients had been maintained for at least 7 days without food or drink by mouth. Their prothrombin times were normal when antibiotic
therapy began. Patients B and C had gastrointestinal bleeding on days 5 and 12, respectively, of antibiotic therapy. Prothrombin times had been rising, culminating, at the onset of haemorrhage, in a twofold (24.5 s, control 10.3) increase in B and a fourfold (46-8 s, control 10.3) increase in C. Neither patient ever had any other significantly altered clotting indices. When cefamandole was withdrawn and vitamin K was given parenterally, all evidence of fresh bleeding stopped within 24 h. Prothrombin times did not return to normal for several days. Patient A had been on haemodialysis for 5 years. 7 days after emergency surgery for a perforated benign gastric ulcer she had a sudden massive upper gastrointestinal bleed. Her prothrombin time was normal until the day of haemorrhage (21-4 s, control 10-5) but no other clotting defect was detected. Parenteral vitamin K and attempts at angiographic localisation and embolisation of the gastric focus were unsuccessful, and she bled to death within a few hours. An antibiotic-associated deficit in vitamin K seems the most likely explanation for these bleeding episodes. The mechanism of antibiotic interference with synthesis and/or utilisation of vitamin K is uncertain. However, with parenteral antibiotics, 1. Fineo GF, Gallus AS, Hirsh J. Unexpected vitamin K deficiency in hospitalized patients. Can Med Assoc J 1973; 109: 880-83. 2. Ham JM. Hypoprothrombinæmia in patients undergoing prolonged intensive care. Med J Aust 1971; ii: 716. 3. Yudis M. Bleeding problems with carbenicillin. Lancet 1972; ii: 599. 4. Registry of adverse reactions. Medicines and Food Division, D.H.S.S., London. 5. Natelson EA, Brown CH III, Bradshaw MW, Alfrey CP Jr, Williams TW Jr. Influence of cephalosporin antibiotics on blood coagulation and platelet function. Antimicrob Agents Chemother 1976; 9: 91-93.
40 excretion in bile seems to be important. Pineo et al.’ in their series of postoperative patients with unexpected vitamin K deficiencies, found fifteen with prolonged prothrombin times of whom nine had a significant gastrointestinal haemorrhage; and all but one of these nine had been maintained on intravenous fluids with no oral intake and had received antibiotics known to be excreted in the bile. Bleeding usually stopped when vitamin K was given. Cefamandole is excreted in high concentrations in bile6 and sufficient antibiotic could thus reach the gut lumen and alter intestinal bacterial flora, especially if dosages were large, as they were in our study. Renal failure may also be important,’1 and the only death due to bleeding in our series of three was in a patient in chronic renal failure. The postoperative patient will be at risk of vitamin K deficiency whenever long-term parenteral nutrition and antibiotic therapy are both needed. We are not suggesting that antibiotics should be avoided or that the postoperative patient must begin eating immediately: that would be absurd. Nevertheless, hypoprothrombinaemia can occur in this setting and can easily be remedied by timely injections of vitamin K-and a problem anticipated is a problem avoided. l-
Supported in part by polis, Indiana.
a
grant from Eli
Department of Surgery, Emory University School of Medicine, Atlanta, Georgia 30303, USA
Lilly and Company, IndianaC. ANN HOOPER BONNIE B. HANEY H. HARLAN STONE
CHLAMYDIÆ IN REITER’S SYNDROME
SIR,-Your Sept. 15 editorial on Reiter’s syndrome (RS) stated that the Chlamydia trachomatis isolates obtained by Schachter seem to have been due to "contamination in the laboratory." I disagree. Wherever studied, the patients’ sera contained homotypic antibody corresponding to the isolates obtained. Some of the chlamydial strains recovered from these men with RS were later shown to be serotypes of Chlamydia that had not been identified in the 1960s. There is no question that many, and perhaps most, acute untreated cases of RS are accompanied by chlamydial infection. A study in Finland found that 69% of patients with RS and genital tract findings had current chlamydial infection of the urethra. Thus, your editorial was concerned only with isolates obtained from the joints of patients with RS. Our early successes in recovering Chlamydia from the joints of a small number of patients with RS were followed by a decreasing recovery rate.2 When these studies were performed in the 1960s the results were difficult to evaluate because many of the features of RS were not consistent with features of chlamydial infections. Thus, the results were discussed cautiously in all our papers on the subject. More recent findings concerning the HLA B27 association with RS have provided grounds for speculation but no certain data concerning a possible role of infection in initiating RS. The role of chlamydial infections in RS remains to be elucidated. However, the fact that a laboratory finding does not lend itself to a simple interpretation does not render it invalid. Hooper Foundation, University of California, G. W.
RAPID IMPROVEMENT OF SKIN LESIONS IN GLUCAGONOMAS WITH INTRAVENOUS SOMATOSTATIN INFUSION
SIR,-Dr Long and colleagues (Oct. 13, p. 764) describe the effect of a long-acting somatostatin analogue on pancreatic endocrine secretion and suggest its therapeutic use in patients with pancreatic endocrine tumours. We have studied the effect of a prolonged, continuous infusion of cyclic somatostatin (kindly provided by Centre de Recherches Clin-Midy, Paris) in a patient with a glucagonoma. This 49-year-old woman presented with diffuse superficial erosions with peripheral scaling and circinate reddish-brown crusts on an erythematous base. Somatostatin was infused intravenously in a dose of 250--g/h for 48 h. Blood samples were taken for determination of glucagon (30 K antiserum). Somatostatin induced a significant reduction in glucagon values (2500 pg/ml to 1000 pg/ml), but plasma concentrations did not return to normal. At the end of the infusion, the skin lesions had improved dramatically: the erosions had resolved and the erythema disappeared. This beneficial effect persisted for 7 weeks, despite the return of glucagon concentrations to pretreatment levels. We have just completed a 36 h infusion of somatostatin in a second patient with a glucagonoma and severe cutaneous lesions. The glucagon results are not yet available. Improvement in the skin condition was observed within 24 h of infusion and was still present 2 weeks later. In these two patients with glucagonomas, somatostatin infusion induced a rapid improvement of cutaneous lesions, possibly by means of an acute reduction in hyperglucagonaemia. To our knowledge this effect has not been previously reported and suggests trials of long-term somatostatin in patients with glucagonomas when other treatments may not be justified.
plasma
Dermatology Clinic and Nutrition-Endocrinology Service, Hôpital Saint Louis, 75475 Paris, France
San Francisco, California
94143, USA.
JULIUS SCHACHTER
DISTORTED HLA SEGREGATION OR BIASED ASCERTAINMENT?
SIR,-Dr Cudworth and his colleagues (Aug. 25,
p. 389) preferential zygotic assortment of certain HLA haplotypes in both type-1 diabetes mellitus families and in healthy families. They report that an excessive number of children in diabetic families inherit the maternal A2-B15-Cw3 haplotype and that children from all families inherit preferentially the paternal A1-B8 haplotype. Type-1 diabetes mellitus has a positive association with the HLA haplotypes A1-B8 and A2-Bl5,’ so that by selecting families through diabetic children, Cudworth et al. are ascertaining preferentially families with AI-B8 or A2-B15-Cw3 children. The 22 A2-B15-Cw3 positive mothers, who were selected on the basis of having at least one diabetic child, have
claim
43 children who inherit A2-B15 and 18 children who do not. If 22 diabetic probands with A2-B15 (the exact number is not provided) were the basis for ascertainment, then the correct segregation ratio in offspring is not 43:18 as given, but 21 (43-22):18, not different from expectation. Similarly, correction for ascertainment bias should be made in analysis of inheritance of A1-B8 in diabetic families. For 28 fathers with the A1-B8 hapoltype, the ratio of Al-B8 : not A1-B8 children is 44:26. The number of A1-B8 diabetic probands is not given, but could approach 28, with resulting corrected ratio of 16:26 which is not different from the expected proportions of 1:with p< 0 O 1. In our own series of diabetes mellitus families with HLAtyping available (unpublished) 8 parents (5 fathers and 3 mothers) are heterozygous carriers of the A1,B8 haplotype.
J. SOHIER M. JEANMOUGIN P. LOMBRAIL PH. PASSA
6. Ratzan KR, Baker HB, Lauredo I. Excretion of cefamandole, cefazolin, and cephalotin into T-tube bile. Antimicrob Agents Chemother 1978; 13: 985-87.
1. Kousa dial
M, Saikku P, Richmond S, Lassus A. Frequent association of chlamyinfection with Reiter’s syndrome. Sexually Transm Dis 1978, 5: 57-61. 2. Schachter J. Can chlamydial infections cause rheumatic disease? In Dumonde DC ed. Infection and immunology in the rheumatic diseases. Oxford: Blackwell Scientific Publications, 1976: 151-57. 3. Cudworth AG, Festenstein H. HLA genetic heterogeneity in diabetes mellitus. Br Med Bull 1978; 34: 285-89.
(x,2=6.34, p<0.02).
All patients were advised not to smoke or to take or to abuse any tablet containing aspirin, alcohol, so these factors cannot explain the difference in the severity of peptic-ulcer disease between placebo and cimetidine treated patients. Few patients have changed their working conditions after the end of the study. In our placebo group there was a mean rise of 50% (p<0.05) in non-stimulated acid secretion (BAO) after treatment for seven months with placebo, and basal acid output was 33% higher (not significant) at the end of the twelve months’ treatment. BAO in the cimetidine treated patients with duodenal ulcer was lower 2 days and 3tmonths after the trial lasting a year. One of us also found, in an earlier study, that in patients with chronic peptic ulcer disease there was a mean rise in BAO of 68% (p<0.10) at the end of treatment with placebo for several years but not after treatment with an anticholinergic drug during the same time period.4 These results suggest that, during the years when patients have active peptic ulcer disease, BAO rises steadily. This could explain the high recurrence rate in peptic ulcers in patients
2. Bodemar G, Walan A, Lundquist G. Gastrin concentrations at the end of and following long term treatment with cimetidine and placebo. In: Cretzfeldt W, ed. Cimetidine. Amsterdam: Excerpta Medica, 1978:233-35. 3. Hansky J, Korman MG, Hetzel DJ, Schearman DJC. Relapse rate after cessation of 12 months Cimetidine in duodenal ulcer. Gastroenterology
1979, 76: 1151. on peptic ulcer disease with special reference of 1-hyoscyamine. Acta Med Scand 1970; suppl 516.
4. Walan A. Studies
to
the effect
with chronic disease. Our data suggest that treatment with cimetidine blocks this rise in BAO and favourably alters the course of the peptic ulcer disease when maintenance treatment is ended. Department of Internal Medicine,
University Hospital, S-581 85 Linköping, Sweden
G. BODEMAR A. WALAN
GASTROINTESTINAL BLEEDING DUE TO VITAMIN K DEFICIENCY IN PATIENTS ON PARENTERAL CEFAMANDOLE
SIR, The combination of low vitamin K intake and limited of the vitamin due to suppression of intestinal bacterial flora can produce severe prothrombin deficiency.’ De-
synthesis
creased vitamin K as the cause of a bleeding diathesis has been noted with several broad-spectrum parenteral antibiotics,2-4 and it can be lethal.5 For patients on antibiotics whose oral intake of nutrients is restricted vitamin K therapy may be warranted. During a study of cefamandole sodium in the management of peritoneal and soft tissue sepsis thirty-seven patients with mixed-flora infections were treated by drainage or excision and with cefamandole sodium (average dose 40 mg/kg body-weight 6-hourly, range 500-4500 mg per injection) for an average of 7 days (range 3-16 days). Cefamandole controlled the sepsis in thirty-four of the thirty-six cases with bacterial pathogens. In three patients, however, the prothrombin time was prolonged. All three patients had major gastrointestinal bleeds, and one massive bleed was fatal. All three patients had presented with clinical signs suggestive of peritonitis, but the underlying diagnoses were found to be (A) perforated gastric ulcer, (B) left colon gangrene due to acute mesenteric arterial thrombosis, and (C) abdominal carcinomatosis. Patients A and B underwent emergency laparotomy and were put on cefamandole before surgery. Patient C was treated by nasogastric decompression and cefamandole; surgery was delayed until the third week. All three patients had been maintained for at least 7 days without food or drink by mouth. Their prothrombin times were normal when antibiotic
therapy began. Patients B and C had gastrointestinal bleeding on days 5 and 12, respectively, of antibiotic therapy. Prothrombin times had been rising, culminating, at the onset of haemorrhage, in a twofold (24.5 s, control 10.3) increase in B and a fourfold (46-8 s, control 10.3) increase in C. Neither patient ever had any other significantly altered clotting indices. When cefamandole was withdrawn and vitamin K was given parenterally, all evidence of fresh bleeding stopped within 24 h. Prothrombin times did not return to normal for several days. Patient A had been on haemodialysis for 5 years. 7 days after emergency surgery for a perforated benign gastric ulcer she had a sudden massive upper gastrointestinal bleed. Her prothrombin time was normal until the day of haemorrhage (21-4 s, control 10-5) but no other clotting defect was detected. Parenteral vitamin K and attempts at angiographic localisation and embolisation of the gastric focus were unsuccessful, and she bled to death within a few hours. An antibiotic-associated deficit in vitamin K seems the most likely explanation for these bleeding episodes. The mechanism of antibiotic interference with synthesis and/or utilisation of vitamin K is uncertain. However, with parenteral antibiotics, 1. Fineo GF, Gallus AS, Hirsh J. Unexpected vitamin K deficiency in hospitalized patients. Can Med Assoc J 1973; 109: 880-83. 2. Ham JM. Hypoprothrombinæmia in patients undergoing prolonged intensive care. Med J Aust 1971; ii: 716. 3. Yudis M. Bleeding problems with carbenicillin. Lancet 1972; ii: 599. 4. Registry of adverse reactions. Medicines and Food Division, D.H.S.S., London. 5. Natelson EA, Brown CH III, Bradshaw MW, Alfrey CP Jr, Williams TW Jr. Influence of cephalosporin antibiotics on blood coagulation and platelet function. Antimicrob Agents Chemother 1976; 9: 91-93.
40 excretion in bile seems to be important. Pineo et al.’ in their series of postoperative patients with unexpected vitamin K deficiencies, found fifteen with prolonged prothrombin times of whom nine had a significant gastrointestinal haemorrhage; and all but one of these nine had been maintained on intravenous fluids with no oral intake and had received antibiotics known to be excreted in the bile. Bleeding usually stopped when vitamin K was given. Cefamandole is excreted in high concentrations in bile6 and sufficient antibiotic could thus reach the gut lumen and alter intestinal bacterial flora, especially if dosages were large, as they were in our study. Renal failure may also be important,’1 and the only death due to bleeding in our series of three was in a patient in chronic renal failure. The postoperative patient will be at risk of vitamin K deficiency whenever long-term parenteral nutrition and antibiotic therapy are both needed. We are not suggesting that antibiotics should be avoided or that the postoperative patient must begin eating immediately: that would be absurd. Nevertheless, hypoprothrombinaemia can occur in this setting and can easily be remedied by timely injections of vitamin K-and a problem anticipated is a problem avoided. l-
Supported in part by polis, Indiana.
a
grant from Eli
Department of Surgery, Emory University School of Medicine, Atlanta, Georgia 30303, USA
Lilly and Company, IndianaC. ANN HOOPER BONNIE B. HANEY H. HARLAN STONE
CHLAMYDIÆ IN REITER’S SYNDROME
SIR,-Your Sept. 15 editorial on Reiter’s syndrome (RS) stated that the Chlamydia trachomatis isolates obtained by Schachter seem to have been due to "contamination in the laboratory." I disagree. Wherever studied, the patients’ sera contained homotypic antibody corresponding to the isolates obtained. Some of the chlamydial strains recovered from these men with RS were later shown to be serotypes of Chlamydia that had not been identified in the 1960s. There is no question that many, and perhaps most, acute untreated cases of RS are accompanied by chlamydial infection. A study in Finland found that 69% of patients with RS and genital tract findings had current chlamydial infection of the urethra. Thus, your editorial was concerned only with isolates obtained from the joints of patients with RS. Our early successes in recovering Chlamydia from the joints of a small number of patients with RS were followed by a decreasing recovery rate.2 When these studies were performed in the 1960s the results were difficult to evaluate because many of the features of RS were not consistent with features of chlamydial infections. Thus, the results were discussed cautiously in all our papers on the subject. More recent findings concerning the HLA B27 association with RS have provided grounds for speculation but no certain data concerning a possible role of infection in initiating RS. The role of chlamydial infections in RS remains to be elucidated. However, the fact that a laboratory finding does not lend itself to a simple interpretation does not render it invalid. Hooper Foundation, University of California, G. W.
RAPID IMPROVEMENT OF SKIN LESIONS IN GLUCAGONOMAS WITH INTRAVENOUS SOMATOSTATIN INFUSION
SIR,-Dr Long and colleagues (Oct. 13, p. 764) describe the effect of a long-acting somatostatin analogue on pancreatic endocrine secretion and suggest its therapeutic use in patients with pancreatic endocrine tumours. We have studied the effect of a prolonged, continuous infusion of cyclic somatostatin (kindly provided by Centre de Recherches Clin-Midy, Paris) in a patient with a glucagonoma. This 49-year-old woman presented with diffuse superficial erosions with peripheral scaling and circinate reddish-brown crusts on an erythematous base. Somatostatin was infused intravenously in a dose of 250--g/h for 48 h. Blood samples were taken for determination of glucagon (30 K antiserum). Somatostatin induced a significant reduction in glucagon values (2500 pg/ml to 1000 pg/ml), but plasma concentrations did not return to normal. At the end of the infusion, the skin lesions had improved dramatically: the erosions had resolved and the erythema disappeared. This beneficial effect persisted for 7 weeks, despite the return of glucagon concentrations to pretreatment levels. We have just completed a 36 h infusion of somatostatin in a second patient with a glucagonoma and severe cutaneous lesions. The glucagon results are not yet available. Improvement in the skin condition was observed within 24 h of infusion and was still present 2 weeks later. In these two patients with glucagonomas, somatostatin infusion induced a rapid improvement of cutaneous lesions, possibly by means of an acute reduction in hyperglucagonaemia. To our knowledge this effect has not been previously reported and suggests trials of long-term somatostatin in patients with glucagonomas when other treatments may not be justified.
plasma
Dermatology Clinic and Nutrition-Endocrinology Service, Hôpital Saint Louis, 75475 Paris, France
San Francisco, California
94143, USA.
JULIUS SCHACHTER
DISTORTED HLA SEGREGATION OR BIASED ASCERTAINMENT?
SIR,-Dr Cudworth and his colleagues (Aug. 25,
p. 389) preferential zygotic assortment of certain HLA haplotypes in both type-1 diabetes mellitus families and in healthy families. They report that an excessive number of children in diabetic families inherit the maternal A2-B15-Cw3 haplotype and that children from all families inherit preferentially the paternal A1-B8 haplotype. Type-1 diabetes mellitus has a positive association with the HLA haplotypes A1-B8 and A2-Bl5,’ so that by selecting families through diabetic children, Cudworth et al. are ascertaining preferentially families with AI-B8 or A2-B15-Cw3 children. The 22 A2-B15-Cw3 positive mothers, who were selected on the basis of having at least one diabetic child, have
claim
43 children who inherit A2-B15 and 18 children who do not. If 22 diabetic probands with A2-B15 (the exact number is not provided) were the basis for ascertainment, then the correct segregation ratio in offspring is not 43:18 as given, but 21 (43-22):18, not different from expectation. Similarly, correction for ascertainment bias should be made in analysis of inheritance of A1-B8 in diabetic families. For 28 fathers with the A1-B8 hapoltype, the ratio of Al-B8 : not A1-B8 children is 44:26. The number of A1-B8 diabetic probands is not given, but could approach 28, with resulting corrected ratio of 16:26 which is not different from the expected proportions of 1:with p< 0 O 1. In our own series of diabetes mellitus families with HLAtyping available (unpublished) 8 parents (5 fathers and 3 mothers) are heterozygous carriers of the A1,B8 haplotype.
J. SOHIER M. JEANMOUGIN P. LOMBRAIL PH. PASSA
6. Ratzan KR, Baker HB, Lauredo I. Excretion of cefamandole, cefazolin, and cephalotin into T-tube bile. Antimicrob Agents Chemother 1978; 13: 985-87.
1. Kousa dial
M, Saikku P, Richmond S, Lassus A. Frequent association of chlamyinfection with Reiter’s syndrome. Sexually Transm Dis 1978, 5: 57-61. 2. Schachter J. Can chlamydial infections cause rheumatic disease? In Dumonde DC ed. Infection and immunology in the rheumatic diseases. Oxford: Blackwell Scientific Publications, 1976: 151-57. 3. Cudworth AG, Festenstein H. HLA genetic heterogeneity in diabetes mellitus. Br Med Bull 1978; 34: 285-89.