254
tested again for HBV serum markers. A serum sample from donor (26-year-old man) proved positive for HBsAb and HBcAb. He had been admitted to a nearby hospital because of liver function abnormalities two months after he gave blood. The hospital record clearly showed that he had contracted typical acute hepatitis B infection. These findings show that, as predicted,2 post-transfusion fulminant hepatitis B infection can occur even after the introduction of blood screening for high-titre HBcAb, through blood from donors who, although acutely infected with HBV, are both HBsAg and HBcAb negative at the time of donation. were
one
M. YOSHIBA K. SEKIYAMA F. SUGATA
Division of Gastroenterology, Showa University Fujigaoka Hospital, Midori-Ku, Yokohama 227, Japan
Y. KAWAMOTO
H. MURAOKA M. AOYAMA
Japanese Red Cross, Kawasaki Blood Centre
M, Yamada H, Yoshikawa Y, et al. Hemodiafiltration treatment of deep hepatic coma by protein passing membrane case report. Artif Organs 1986; 10:
1. Yoshiba
used since it was prepared by Distillers Biochemicals from penicillin which, on hydrolysis, yields the D isomer only. In the United States penicillamine for clinical trials was synthesised from D,L-valine and the D and L forms were not separated, hence the case of optic neuritis described by Tu et al.’ After this report D,L-penicillamine was withdrawn and now only the D isomer is available. However, even D-penicillamine can mimic the antipyridoxine action of the L isomer if given in large doses and under certain specific circumstances such as pregnancy or during a growth spurt.4 In 1991 a case of optic neuritis induced by D-penicillamine was reported.s Kean et al seem to imply that the penicillamine-copper chelate is a simple 2-to-lI combination. The evidence suggests many penicillamine molecules binding copper (I) and (II) in a ratio of 14 to 12, though it does not follow that the chelate formed in vitro is was
that which occurs in vivo,7 where other aminoacids and albumin may be involved. Department of Neurology, Middlesex Hospital, London W1 N 8AA, UK
J. M. WALSHE
417-21. 2.
Hoofnagle JH. Editorial: posttransfusion hepatitis B Transfusion 1990; 30:
384-86.
Cyclosporin and muscle SIR,-Arellano and Krupp/ in their review of muscle disorders associated with cyclosporin, referred to three cases of possible muscular toxicity published by our group in 1989.2 In the table they state that biopsy was not done in two of our patients and that all had received prednisone. This is incorrect. Our report noted that biopsy, with examination by electronmicroscopy, was done in all cases, and all three had abnormalities. Moreover prednisone had been discontinued several months before myopathy developed. The statement that electromyograms were normal in our three cases is not true either, since our first case had an abnormal electromyogram. Arellano and Krupp have not read our report carefully, and it is very regrettable that Sandoz’ drug monitoring centre shall have recorded incorrect information, with the risk of the possible myotoxicity of cyclosporin being underrated. Division of
Cardiology and Pathology, CHU Vaudois,
J.-J. GOY J. P. DERUAZ
1011 Lausanne, Switzerland 1. Arellano
F, Krupp
P. Muscular disorders associated with
cyclosporin
Lancet 1991;
337: 915 2.
Goy JJ, Stauffer SC, Deruaz JP, et al. Myopathy cyclosporin. Lancet 1989, i: 1446-47.
as a
possible side-effect of
***This letter has been shown to Dr Arellano and Professor Krupp, reply follows.-ED. L. SIR,-We regret the misquotation of Dr Goy and colleagues’ report. Despite this oversight, we feel that the conclusions in our letter were not inappropriate. Indeed, our review has resulted in amendments to international product information on ’Sandimmun’. Myopathy is now mentioned as a side-effect and colchicine and lovastatin are noted as possible drug interactions leading to muscular toxicity. It was never our intention to underrate the myotoxicity of cyclosporin.
whose
Drug Monitoring Centre, Sandoz Pharma, 4002 Basle, Switzerland
F. ARELLANO P. KRUPP
Chirality of penicillamine SIR,-In their review on chirality and antirheumatic drugs Dr Kean and colleagues (Dec 21/28, p 1565) pay special attention to the D (or S) and the L (or R) enantiomers of penicillamine and refer to their work to illustrate optic neuritis induced by using D,L-penicillamine. That manifestation of L-penicillamine toxicity had been known since 1963/ and the original description of L-penicillamine toxicity can be traced back to the work of Wilson and Du Vigneaud in 1950/* In my 1956 article on the therapeutic potential of penicillamine for Wilson’s disease I stressed the differential toxicity of the D and L forms and recommended that the D form be used in clinical trials.’ In the UK D-penicillamine only
J-b, Blackwell RQ, Lee P-F. DL-penicillamine as a cause of optic axial neuritis JAMA 1963; 185: 83-86. 2. Wilson JE, Du Vigneaud V. Inhibition of growth in the rat by 1-penicillamine and its prevention by aminoethanol and related compounds J Biol Chem 1950; 312: 63-70. 3. Walshe JM. Penicillamine, a new oral therapy for Wilson’s disease Am J Med 1956, 1. Tu
21: 487-95. 4. Gibbs K, Walshe JM. Interruption of the tryptophan-nicotinic acid pathway by penicillamine induced pyridoxine deficiency in patients with Wilson’s disease and in experimental animals Ann NY Acad Sci 1969; 166: 158-69. 5. Lee A. Lawton NF. Penicillamine treatment of Wilson’s disease and optic neuropathy J Neurol Neurosurg Psychiatry 1991; 58: 746 6. Laurie SH, Prime DN. The formation and nature of the mixed valence copper-dpenicillamine-chloride cluster in aqueous solution and its relevance to the treatment of Wilson’s disease. J Inorgan Biochem 1979; 11: 229-39.
Two types of translucent membrane of caesarean section scar tissue SIR,-Pedowitz and Schwartz have reported that translucent membrane of transverse caesarean section scar tissue and incomplete uterine rupture occurred in 8-3% (22/266) patients at repeat caesarean section.’ We have graded findings at repeat caesarean section as follows: grade 1, neither thinning nor loss of continuity of lower uterine segment; grade II, thinning and loss of continuity but fetal hair not visible; grade III thinning or absence of lower uterine segment and fetal hair visible. The frequency of grade III was 9-1% (18/197) from September, 1982, until now 2-4 a frequency much the same as that recorded by Pedowitz and Schwartz. However, complete rupture of the lower transverse uterine scar is reported in 0-2-3% of planned trials of labour,s so what type of translucent membrane is it that ruptures during trials of labour-and can ultrasound scans distinguish one type from another? Since February, 1989, we have scanned 45 patients with previous caesarean section by high-resolution ultrasound (SSA-250A, Toshiba). 38 patients showed good healing with a thickness more than 1-2 mm throughout and 7 patients showed poor healing with reduced thickness and/or loss of continuity. Of 38 patients with good healing 10 were delivered vaginally and 28 patients had repeat caesarean section for other obstetric indications. Of 7 patients with poor healing 3 showed grade III operative findings (translucent membrane). The thinnest thickness of the lower uterine segment in these 3 patients near term by ultrasound is shown in the figure. Patient I showed great variability of thickness in the short term (5-10 min) and also over time (37-39 weeks) but patient III showed little variability (static thinning). Before repeat caesarean section in patient I thickness was never less than 0-5 mm and we predicted grade II; the thickness, by ophthalmic calipers, was 0-5 mm at operation. In patient III, with static thinning, thickness was 0-3 mm by ultrasound (7-5 MHz linear probe) before operation and 0-2 mm by ophthalmic calipers at operation. The thinnest portion of the lower uterine segment in patient I (variable thinning) showed good preservation of smooth muscle fibres and a little stromal fibrosis while that of the patient with static thinning showed prominent degeneration and atrophy of muscle fibres and severe stromal fibrosis with hyalinisation.
255
36w
37w
3ew
39W
4UW
Serial membrane thickness measurements in 3 patients with translucent membranes near term by ultrasound.
High-resoloution ultrasound may be able to distinguish two types of translucent membrane by the variability of thickness in the lower uterine segment, and a translucent membrane with static thinning may be more prone to rupture than one with variable thinning. Since September, 1982, 67 vaginal deliveries and 23 repeat caesarean section have been preceded by ultrasound examination. There were no complete uterine ruptures but there were 2 incomplete ruptures. For these 2 patients caesarean section was planned because of poor healing according to ultrasound; but when labour pains began repeat caesarean sections were done immediately and incomplete uterine ruptures were observed. Unfortunately, these 2 patients were seen before 1989 and we could not look for variable or static thinning. Fukuda Ladies Clinic, 309 Kariya, Aho, Japan
MISAO FUKUDA KIYOMI FUKUDA
Division of Perinatology, Ottawa General Hospital, Ottawa, Canada
TAKASHI SHIMIZU
Natsuyama Hospital, Natsuyama, Japan Department of Obstetrics and Gynecology, Kobe
University School of Medicine
EIICHI NATSUYAMA
MATSUTO MOCHIZUKI
1 Pedowitz
P, Schwartz RM. The true incidence of silent rupture of caesarean section Am J Obstet Gynecol 1957; 74: 1071-81 2 Fukuda M, Fukuda K, Mochizuki M. Examination of previous caesarean section scars by ultrasound Arch Gynecol Obstet 1988; 243: 221-24 3 Fukuda M, Shimizu T, Ihara Y, Fukuda K, Natsuyama E, Mochizuki M. Ultrasound examination of caesarean section scars during pregnancy. Arch Gynecol Obstet scar.
1991; 248: 129-38. M, Shimizu T, Fukuda K, Natsuyama E, Mochizuki M New ultrasound examinatin of previous caesarean section scars during pregnancy, labor and postpartum (Abstr) Presented at sixth meeting of World Federation for Ultrasound in Medicine and Biology (Copenhagen, Sept 1-6, 1991). 5. Scott JR. Mandatory trial of labor after caesarean delivery an alternative viewpoint Obstet Gynecol 1991; 77: 811-14 4 Fukuda
Long QT and Harvey-ras SIR,-Your Nov 5 editorial suggests that one possible mechanism for the long
QT syndrome may be related to abnormalities in the 1 gene, since its protein product is implicated in transmembrane signalling, and linkage without recombination has been demonstrated between long QT and Harvey-ras in one study.’ There is, however, genetic heterogeneity in the long QT syndrome, and in the autosomal recessive form associated with sensorineural deafness (Jervell and Lange-Nielsen syndrome)2 we have not been able to demonstrate linkage to Harvey-ras. We have investigated a family in which two children have the Jervell and Lange-Nielsen syndrome. One (113) was diagnosed as having a congenital long QT syndrome at 15 months of age when he had episodic loss of consciousness. Electrographic recordings confirmed a long QT interval in sinus rhythm and Torsades de pointes was seen during attacks. He and his elder brother (114) were both known to have bilateral sensorineural deafness since early infancy and a subsequent electrocardiogram confirmed that 114 also Harvey-ras
had a long QT interval. Neither parent has a long QT interval in sinus rhythm and neither has had symptoms suggestive of the condition. The two boys therefore have the typical syndrome originally described by Jervell and Lange-Nielsen. No other family members have a long QT interval on surface electrocardiography. Because of repeated episodes of Torsades de pointes, 113 was treated with propranolol and seemed to be controlled until 8 years of age when he once again presented with recurrent Torsades de pointes, which responded to an increase in propranolol. 114 first had symptoms at 3 years of age. He was treated with propranolol, which was followed by a left stellate ganglionectomy. He has remained free of symptoms since this time on propranolol. Investigation of this family with the probe EJ6-6 showed that the two affected sons had inherited different alleles for this Harvey-ras cDNA probe from their father. For the Harvey-ras gene to be linked to the syndrome in this family there would have had to be a recombination event in one of the children between the disease gene and Harvey-ras in the paternal chromosome 11. Since there were no recombinations seen in the previous linkage study, this recombination suggests that long QT with deafness may not be linked to Harvey-ras 1 and a physiological connection between the Harvev-ras orotein and arrhythmia in this familv is unlikelv.
S. JEFFERY St George’s Hospital Medical School, London SW17 0RE, UK
Royal Brompton National and Lung Hospitals,
R. JAMIESON M. A. PATTON
Heart
J. TILL
London SW3
Keating M, Atkinson D, Dunn C, et al Linkage of a cardiac arrhythmia, the long QT syndrome, and the Harvey-Ras-1 gene. Science 1990; 252: 704-06 2. Fraser G, Froggatt P, James T Congenital deafness associated with electrocardiographic abnormalities, fainting attacks and sudden death. Q J Med 1964; 33: 361-85 1.
Safety of fluconazole in women taking oral hypoglycaemic agents SIR,-Fluconazole is recommended for vulvovaginal candidosis. Early studies suggested that it prolongs the half-life of the sulphonylurea group of oral hypoglycaemic agents.’ There are also reports that its concomitant administration with chlorpropamide, glibenclamide, or glipizide may be associated with hypoglycaemic symptoms, although no alteration in blood glucose control has been recorded.2 These observations have led to the suggestion that careful monitoring of blood glucose is required when fluconazole and an oral hypoglycaemic agent are given together. With ethical committee approval we studied 29 postmenopausal diabetic women with vulvovaginitis (mean age 64, range 49-88). 14 were randomised to fluconazole 50 mg per day for 14 days and 15 to intravaginal clotrimazole 100 mg for the same period. These women were taking gliclazide (17) or glibenclamide (12). Diabetic control was assessed by measurement of serum fructosamine and glycosylated haemoglobin (HbA,) at the beginning of the study, after the 2 week interaction period, and 4 weeks after cessation of antifungal therapy. The women were questioned about hypoglycaemic symptoms at every clinic visit. The two drugs were equally effective in controlling vulvovaginal symptoms but, as previously reportedfluconazole was easier to use. There was no significant difference in blood glucose control between the two treatment groups at baseline or at 2 (or, not shown, 6) weeks: Fluconazole
Clotrlmazole
Nor were there significant changes in control in either oral hypoglycaemic agent group at 2 weeks or 6 weeks compared with baseline (data not shown). No patient had any symptoms suggestive of hypoglycaemia and no side-effects or laboratory abnormalities noted. The recommended dose of fluconazole for vaginal candidosis is 150 mg as a single oral dose. In patients in whom hypoglycaemia may be likely the manufacturers claim that 50 mg per day for 3 days is equally effective.3 Our patients, who were on oral hypoglycaemic were
tested again for HBV serum markers. A serum sample from donor (26-year-old man) proved positive for HBsAb and HBcAb. He had been admitted to a nearby hospital because of liver function abnormalities two months after he gave blood. The hospital record clearly showed that he had contracted typical acute hepatitis B infection. These findings show that, as predicted,2 post-transfusion fulminant hepatitis B infection can occur even after the introduction of blood screening for high-titre HBcAb, through blood from donors who, although acutely infected with HBV, are both HBsAg and HBcAb negative at the time of donation. were
one
M. YOSHIBA K. SEKIYAMA F. SUGATA
Division of Gastroenterology, Showa University Fujigaoka Hospital, Midori-Ku, Yokohama 227, Japan
Y. KAWAMOTO
H. MURAOKA M. AOYAMA
Japanese Red Cross, Kawasaki Blood Centre
M, Yamada H, Yoshikawa Y, et al. Hemodiafiltration treatment of deep hepatic coma by protein passing membrane case report. Artif Organs 1986; 10:
1. Yoshiba
used since it was prepared by Distillers Biochemicals from penicillin which, on hydrolysis, yields the D isomer only. In the United States penicillamine for clinical trials was synthesised from D,L-valine and the D and L forms were not separated, hence the case of optic neuritis described by Tu et al.’ After this report D,L-penicillamine was withdrawn and now only the D isomer is available. However, even D-penicillamine can mimic the antipyridoxine action of the L isomer if given in large doses and under certain specific circumstances such as pregnancy or during a growth spurt.4 In 1991 a case of optic neuritis induced by D-penicillamine was reported.s Kean et al seem to imply that the penicillamine-copper chelate is a simple 2-to-lI combination. The evidence suggests many penicillamine molecules binding copper (I) and (II) in a ratio of 14 to 12, though it does not follow that the chelate formed in vitro is was
that which occurs in vivo,7 where other aminoacids and albumin may be involved. Department of Neurology, Middlesex Hospital, London W1 N 8AA, UK
J. M. WALSHE
417-21. 2.
Hoofnagle JH. Editorial: posttransfusion hepatitis B Transfusion 1990; 30:
384-86.
Cyclosporin and muscle SIR,-Arellano and Krupp/ in their review of muscle disorders associated with cyclosporin, referred to three cases of possible muscular toxicity published by our group in 1989.2 In the table they state that biopsy was not done in two of our patients and that all had received prednisone. This is incorrect. Our report noted that biopsy, with examination by electronmicroscopy, was done in all cases, and all three had abnormalities. Moreover prednisone had been discontinued several months before myopathy developed. The statement that electromyograms were normal in our three cases is not true either, since our first case had an abnormal electromyogram. Arellano and Krupp have not read our report carefully, and it is very regrettable that Sandoz’ drug monitoring centre shall have recorded incorrect information, with the risk of the possible myotoxicity of cyclosporin being underrated. Division of
Cardiology and Pathology, CHU Vaudois,
J.-J. GOY J. P. DERUAZ
1011 Lausanne, Switzerland 1. Arellano
F, Krupp
P. Muscular disorders associated with
cyclosporin
Lancet 1991;
337: 915 2.
Goy JJ, Stauffer SC, Deruaz JP, et al. Myopathy cyclosporin. Lancet 1989, i: 1446-47.
as a
possible side-effect of
***This letter has been shown to Dr Arellano and Professor Krupp, reply follows.-ED. L. SIR,-We regret the misquotation of Dr Goy and colleagues’ report. Despite this oversight, we feel that the conclusions in our letter were not inappropriate. Indeed, our review has resulted in amendments to international product information on ’Sandimmun’. Myopathy is now mentioned as a side-effect and colchicine and lovastatin are noted as possible drug interactions leading to muscular toxicity. It was never our intention to underrate the myotoxicity of cyclosporin.
whose
Drug Monitoring Centre, Sandoz Pharma, 4002 Basle, Switzerland
F. ARELLANO P. KRUPP
Chirality of penicillamine SIR,-In their review on chirality and antirheumatic drugs Dr Kean and colleagues (Dec 21/28, p 1565) pay special attention to the D (or S) and the L (or R) enantiomers of penicillamine and refer to their work to illustrate optic neuritis induced by using D,L-penicillamine. That manifestation of L-penicillamine toxicity had been known since 1963/ and the original description of L-penicillamine toxicity can be traced back to the work of Wilson and Du Vigneaud in 1950/* In my 1956 article on the therapeutic potential of penicillamine for Wilson’s disease I stressed the differential toxicity of the D and L forms and recommended that the D form be used in clinical trials.’ In the UK D-penicillamine only
J-b, Blackwell RQ, Lee P-F. DL-penicillamine as a cause of optic axial neuritis JAMA 1963; 185: 83-86. 2. Wilson JE, Du Vigneaud V. Inhibition of growth in the rat by 1-penicillamine and its prevention by aminoethanol and related compounds J Biol Chem 1950; 312: 63-70. 3. Walshe JM. Penicillamine, a new oral therapy for Wilson’s disease Am J Med 1956, 1. Tu
21: 487-95. 4. Gibbs K, Walshe JM. Interruption of the tryptophan-nicotinic acid pathway by penicillamine induced pyridoxine deficiency in patients with Wilson’s disease and in experimental animals Ann NY Acad Sci 1969; 166: 158-69. 5. Lee A. Lawton NF. Penicillamine treatment of Wilson’s disease and optic neuropathy J Neurol Neurosurg Psychiatry 1991; 58: 746 6. Laurie SH, Prime DN. The formation and nature of the mixed valence copper-dpenicillamine-chloride cluster in aqueous solution and its relevance to the treatment of Wilson’s disease. J Inorgan Biochem 1979; 11: 229-39.
Two types of translucent membrane of caesarean section scar tissue SIR,-Pedowitz and Schwartz have reported that translucent membrane of transverse caesarean section scar tissue and incomplete uterine rupture occurred in 8-3% (22/266) patients at repeat caesarean section.’ We have graded findings at repeat caesarean section as follows: grade 1, neither thinning nor loss of continuity of lower uterine segment; grade II, thinning and loss of continuity but fetal hair not visible; grade III thinning or absence of lower uterine segment and fetal hair visible. The frequency of grade III was 9-1% (18/197) from September, 1982, until now 2-4 a frequency much the same as that recorded by Pedowitz and Schwartz. However, complete rupture of the lower transverse uterine scar is reported in 0-2-3% of planned trials of labour,s so what type of translucent membrane is it that ruptures during trials of labour-and can ultrasound scans distinguish one type from another? Since February, 1989, we have scanned 45 patients with previous caesarean section by high-resolution ultrasound (SSA-250A, Toshiba). 38 patients showed good healing with a thickness more than 1-2 mm throughout and 7 patients showed poor healing with reduced thickness and/or loss of continuity. Of 38 patients with good healing 10 were delivered vaginally and 28 patients had repeat caesarean section for other obstetric indications. Of 7 patients with poor healing 3 showed grade III operative findings (translucent membrane). The thinnest thickness of the lower uterine segment in these 3 patients near term by ultrasound is shown in the figure. Patient I showed great variability of thickness in the short term (5-10 min) and also over time (37-39 weeks) but patient III showed little variability (static thinning). Before repeat caesarean section in patient I thickness was never less than 0-5 mm and we predicted grade II; the thickness, by ophthalmic calipers, was 0-5 mm at operation. In patient III, with static thinning, thickness was 0-3 mm by ultrasound (7-5 MHz linear probe) before operation and 0-2 mm by ophthalmic calipers at operation. The thinnest portion of the lower uterine segment in patient I (variable thinning) showed good preservation of smooth muscle fibres and a little stromal fibrosis while that of the patient with static thinning showed prominent degeneration and atrophy of muscle fibres and severe stromal fibrosis with hyalinisation.
255
36w
37w
3ew
39W
4UW
Serial membrane thickness measurements in 3 patients with translucent membranes near term by ultrasound.
High-resoloution ultrasound may be able to distinguish two types of translucent membrane by the variability of thickness in the lower uterine segment, and a translucent membrane with static thinning may be more prone to rupture than one with variable thinning. Since September, 1982, 67 vaginal deliveries and 23 repeat caesarean section have been preceded by ultrasound examination. There were no complete uterine ruptures but there were 2 incomplete ruptures. For these 2 patients caesarean section was planned because of poor healing according to ultrasound; but when labour pains began repeat caesarean sections were done immediately and incomplete uterine ruptures were observed. Unfortunately, these 2 patients were seen before 1989 and we could not look for variable or static thinning. Fukuda Ladies Clinic, 309 Kariya, Aho, Japan
MISAO FUKUDA KIYOMI FUKUDA
Division of Perinatology, Ottawa General Hospital, Ottawa, Canada
TAKASHI SHIMIZU
Natsuyama Hospital, Natsuyama, Japan Department of Obstetrics and Gynecology, Kobe
University School of Medicine
EIICHI NATSUYAMA
MATSUTO MOCHIZUKI
1 Pedowitz
P, Schwartz RM. The true incidence of silent rupture of caesarean section Am J Obstet Gynecol 1957; 74: 1071-81 2 Fukuda M, Fukuda K, Mochizuki M. Examination of previous caesarean section scars by ultrasound Arch Gynecol Obstet 1988; 243: 221-24 3 Fukuda M, Shimizu T, Ihara Y, Fukuda K, Natsuyama E, Mochizuki M. Ultrasound examination of caesarean section scars during pregnancy. Arch Gynecol Obstet scar.
1991; 248: 129-38. M, Shimizu T, Fukuda K, Natsuyama E, Mochizuki M New ultrasound examinatin of previous caesarean section scars during pregnancy, labor and postpartum (Abstr) Presented at sixth meeting of World Federation for Ultrasound in Medicine and Biology (Copenhagen, Sept 1-6, 1991). 5. Scott JR. Mandatory trial of labor after caesarean delivery an alternative viewpoint Obstet Gynecol 1991; 77: 811-14 4 Fukuda
Long QT and Harvey-ras SIR,-Your Nov 5 editorial suggests that one possible mechanism for the long
QT syndrome may be related to abnormalities in the 1 gene, since its protein product is implicated in transmembrane signalling, and linkage without recombination has been demonstrated between long QT and Harvey-ras in one study.’ There is, however, genetic heterogeneity in the long QT syndrome, and in the autosomal recessive form associated with sensorineural deafness (Jervell and Lange-Nielsen syndrome)2 we have not been able to demonstrate linkage to Harvey-ras. We have investigated a family in which two children have the Jervell and Lange-Nielsen syndrome. One (113) was diagnosed as having a congenital long QT syndrome at 15 months of age when he had episodic loss of consciousness. Electrographic recordings confirmed a long QT interval in sinus rhythm and Torsades de pointes was seen during attacks. He and his elder brother (114) were both known to have bilateral sensorineural deafness since early infancy and a subsequent electrocardiogram confirmed that 114 also Harvey-ras
had a long QT interval. Neither parent has a long QT interval in sinus rhythm and neither has had symptoms suggestive of the condition. The two boys therefore have the typical syndrome originally described by Jervell and Lange-Nielsen. No other family members have a long QT interval on surface electrocardiography. Because of repeated episodes of Torsades de pointes, 113 was treated with propranolol and seemed to be controlled until 8 years of age when he once again presented with recurrent Torsades de pointes, which responded to an increase in propranolol. 114 first had symptoms at 3 years of age. He was treated with propranolol, which was followed by a left stellate ganglionectomy. He has remained free of symptoms since this time on propranolol. Investigation of this family with the probe EJ6-6 showed that the two affected sons had inherited different alleles for this Harvey-ras cDNA probe from their father. For the Harvey-ras gene to be linked to the syndrome in this family there would have had to be a recombination event in one of the children between the disease gene and Harvey-ras in the paternal chromosome 11. Since there were no recombinations seen in the previous linkage study, this recombination suggests that long QT with deafness may not be linked to Harvey-ras 1 and a physiological connection between the Harvev-ras orotein and arrhythmia in this familv is unlikelv.
S. JEFFERY St George’s Hospital Medical School, London SW17 0RE, UK
Royal Brompton National and Lung Hospitals,
R. JAMIESON M. A. PATTON
Heart
J. TILL
London SW3
Keating M, Atkinson D, Dunn C, et al Linkage of a cardiac arrhythmia, the long QT syndrome, and the Harvey-Ras-1 gene. Science 1990; 252: 704-06 2. Fraser G, Froggatt P, James T Congenital deafness associated with electrocardiographic abnormalities, fainting attacks and sudden death. Q J Med 1964; 33: 361-85 1.
Safety of fluconazole in women taking oral hypoglycaemic agents SIR,-Fluconazole is recommended for vulvovaginal candidosis. Early studies suggested that it prolongs the half-life of the sulphonylurea group of oral hypoglycaemic agents.’ There are also reports that its concomitant administration with chlorpropamide, glibenclamide, or glipizide may be associated with hypoglycaemic symptoms, although no alteration in blood glucose control has been recorded.2 These observations have led to the suggestion that careful monitoring of blood glucose is required when fluconazole and an oral hypoglycaemic agent are given together. With ethical committee approval we studied 29 postmenopausal diabetic women with vulvovaginitis (mean age 64, range 49-88). 14 were randomised to fluconazole 50 mg per day for 14 days and 15 to intravaginal clotrimazole 100 mg for the same period. These women were taking gliclazide (17) or glibenclamide (12). Diabetic control was assessed by measurement of serum fructosamine and glycosylated haemoglobin (HbA,) at the beginning of the study, after the 2 week interaction period, and 4 weeks after cessation of antifungal therapy. The women were questioned about hypoglycaemic symptoms at every clinic visit. The two drugs were equally effective in controlling vulvovaginal symptoms but, as previously reportedfluconazole was easier to use. There was no significant difference in blood glucose control between the two treatment groups at baseline or at 2 (or, not shown, 6) weeks: Fluconazole
Clotrlmazole
Nor were there significant changes in control in either oral hypoglycaemic agent group at 2 weeks or 6 weeks compared with baseline (data not shown). No patient had any symptoms suggestive of hypoglycaemia and no side-effects or laboratory abnormalities noted. The recommended dose of fluconazole for vaginal candidosis is 150 mg as a single oral dose. In patients in whom hypoglycaemia may be likely the manufacturers claim that 50 mg per day for 3 days is equally effective.3 Our patients, who were on oral hypoglycaemic were
