- Email: [email protected]
Sumatriptan and ischaemic heart disease
THE LANCET
CASEWISE CONCORDANCE FOR DISEASES REPORTED BY AUSTRALIAN TWINS
As
of
I re-examined data from questionnaire surveys of 3808 adult twin pairs.3 These data included zygosity, reported birthweight, and occurrence of several common diseases. Single placenta MZ twins (of whom 75% will be monochorionic) did not differ significantly from DZ twins in prevalence of any of these diseases or in concordance between single and double placenta MZ twins (table). I conclude that effects such as those Phillips is concerned about are not large enough to alter the broad conclusions made in the classic twin design about genetic a test
Phillips’ hypothesis,
determination. Epidemiology Unit, Queensland Institute of Medical Research, Herston, Queensland 4029, Australia
affected. Methodologically sound studies that address this issue in adult disorders are rare, but, for example, for psychiatric disorders4 over-representation of twins is not generally the case. Schizophrenia, in single surviving twins may be the exception,and some studies suggest that prenatal factors may be related to discordance for schizophrenia in monozygotic twins.5 Twin study methods have far wider application than the estimation of heritability.’ The comparison of results obtained by various genetic epidemiological methods7 enables information to be gathered about both the genetic and environmental influences on disease; proof of genetic aetiology is not what is being sought. There are other issues relating to classic twin studies of adult disease, such as ascertainment and investigator biases, and methods of analysis, which may have far greater impact on the validity of results and receive considerably less attention. Twins remain an important natural experimental model and the destructive divisions of the nature-nurture controversy are meaningless in modem medical research. The paradigm is no longer "nature or nurture?" but "how do genes and environment operate?". Twin studies continue to have much to offer in finding the answers. Genetics Section, Institute of Psychiatry, De Crespigny Park, London SE5 8AF, UK
ALISON M. MACDONALD
MacGillivray I, Campbell D, Thompson B, eds. Twinning and twins. Ch 9-12. Chichester: John Wiley, 1988. 2. Melnick M, Myrianthopoulos NC. The effects of chorion type on normal and abnormal developmental variation in monozygous twins. Am J Med Genet 1979; 4: 1.
147-56. 3.
4.
D. L. DUFFY 5.
1. Vlietinck R, Derom R, Neale MC, et al. Genetic and environmental variation in the birth weight of twins. Behav Genet 1989; 19: 151-61. 2. Reed T, Christian JC, Wood PD, Schaefer EJ. Influence of placentation on high density lipoprotein in adult males: the NHLBI twin study. Acta Genet Med Gemellol 1991; 40: 353-59. 3. Duffy DL, Battistutta D, Martin NG, Hopper JL, Matthews JD. Genetics of asthma and hayfever in Australian twins. Am Rev Respir Dis 1990 142: 1351-58.
1419
6. 7.
Corney G, MacGillivray I, Campbell DM, Thompson B, Little J. Congenital anomalies in twins in Aberdeen and North-East Scotland. Acta Genet Med Gemellol 1983; 32: 31-35. Chitkara B, Macdonald AM, Reveley AM. Twin birth and adult psychiatric disorder: an examination of the case records of the Maudsley Hospital. Br J Psychiatry 1988; 152: 391-98. Lewis SW, Chitkara B, Reveley AM, Murray RM. Family history and birth weight in monozygotic twins concordant and discordant for psychosis. Acta Genet Med Gemellol 1987; 36: 267-73. McGuffin P, Katz R. Who believes in estimating heritability as an end in itself? Behav Brain Sci 1990; 13: 141-42. Morton NE, Rao DC, Lalouel J-M. Methods in genetic epidemiology. Basel: Karger, 1983.
Sumatriptan and ischaemic heart disease SIR,-Dr Phillips says that twin studies may have misled us into believing a genetic origin of many diseases. His basic premise that the adverse prenatal environment of monozygotic twins relative to that of dizygotic twins will result in increased monozygotic concordance and hence to overestimates of genetic aetiology is fallacious, even by his own evidence. Phillips notes that monozygotic twins have greater within-pair variability of birthweight than dizygotic twins-ie, they are more discordant. The relation of birthweight to later development in twins is complex and only where the intra-pair difference is extreme have differences in longer term outcome been detected.1 In discussing the higher rates of congenital malformations he does not mention that the most recent data2,3 show no difference in rates in monochorionic and dichorionic monozygotic twins, nor that discordance is the norm for congenital malformations in surviving twins, be they monozygotic or dizygotic.1 Perinatal mortality also frequently results in the survival of a single infant;1 such single survivors, who might be regarded as most at risk under Phillips’ hypothesis, are rarely included in classic twin studies, and concordant pairs are never so. Overwhelmingly, differences in the prenatal environment of twins lead to discordance, and make monozygotic twins less alike. On present evidence, the classic twin study will not lead to erroneous overestimation of genetic influence on adult diseases in which prenatal factors are important in aetiology, but, if anything, to underestimation. It is, as Phillips notes, difficult to determine whether prenatal factors affecting twins influence the development of adult disease. However, if this is the case, it follows that twins per se should have higher rates of the diseases in question than the general population. If monozygotic twins are more adversely affected than dizygotic twins, a disproportionate number of monozygotic pairs should be
SIR,-Dr Ottervanger and colleagues report (April 3, p 861) the incidence ofa transmural myocardial infarction after administration of 6 mg sumatriptan subcutaneously in a patient with cluster headache. They point out that the effect of sumatriptan on the human coronary vasculature has been reported to be mild, but offer no explanation of how this 5-hydroxytryptamine-like selective agonist could induce an acute coronary event severe enough to cause a myocardial infarction. Data from our laboratory have also indicated that contractile effects mediated by 5-HTB-like receptors are only weak in non-atherosclerotic human coronary arteries.1 However, further studies demonstrated that the effect of serotonin, acting specifically at 5-HTB-like receptors in healthy human coronary arteries, could be substantially enhanced in the presence of the thromboxane mimetic U46619.2 We have been able to confirm that a similar interaction occurs between U46619 and sumatriptan in human isolated non-atherosclerotic coronary arteries (figure). Dr Maclean and colleagues (April 24, p 1092) propose an explanation for the anomaly between the weak constrictor properties of sumatriptan in apparently normal coronary arteries and the occurrence of myocardial infarction. They suggest that basal release of nitric oxide in the healthy coronary artery is sufficient to inhibit the vessels’ response to sumatriptan, but in smokers, in whom there is a loss of endothelial function, the effect of sumatriptan is unmasked. We believe that stimulation of 5-HT l-like receptors by serotonin, under certain conditions, may cause flow-limiting stenosis resulting in myocardial ischaemia and infarction. This could arise if the effect of selective 5-HTclike receptor agonists, such as sumatriptan, is enhanced by the presence ofthromboxane- as well as conditions
1420
THE LANCET
AND Emax EXPRESSED AS PERCENTAGE OF MAXIMUM CONTRACTILE RESPONSE TO POTASSIUM,
EC50 (nmol/L)
100 mmol/L*
*Mean
1101
Contractile effect of sumatriptan in absence of U46619 (open circles, n = 4), and in presence of EC,o concentration (filled circles, n = 4), EC,o concentration (open triangles), or ECso concentration (filled triangles) of U46619 on nonatherosclerotic epicardial coronary arteries.
Specimens were obtained from 4 patients aged 25, 29, 43, and receiving heart transplantation for cardiomyopathy. Experiments were done in duplicate in vessels from each patient. (Initial response U46619concentrations shown by different starting points for each curve on the y axis.) M mol/L.
54 years,
=
(SE) shown, n=6-11, tsubcutaneous.
The recommended dose of subcutaneously administered ergotamine is 10 to 48 times lower than for subuctaneous sumatriptan2-4 (table). However, the 100-fold higher potency, combined with the 2-fold higher maximum effect, as well as the much longer duration of action of ergotamine, suggests that, in general, the cardiac liability of ergotamine is higher than that of sumatriptan. Both drugs are contraindicated in patients with known coronary artery disease. Possibly, the oral administration of sumatriptan-if feasible-would lead to a less rapid rise in plasma concentrations, which could be safer in patients who may be at risk from coronary side-effects. Department of Pharmacology, Faculty of Medicine and Health Sciences, Erasmus University Rotterdam, 3000 DR Rotterdam, Netherlands
WILLEM A. BAX PRAMOD R. SAXENA
of abrogated nitric oxide release. These mechanisms could be of pathophysiological importance in vessels affected by atherosclerotic disease, since we have shown that areas immediately distal to atheromas are hyperreactive to sumatriptan.1 The possibility exists that some patients receiving sumatriptan could have newly formed yet undetected atheroma, with an associated area of vascular hyperreactivity. Care should be taken when giving this drug to patients with coronary risk factors. The physiological role of 5-HT1-like receptors remains unclear, but in certain circumstances these receptors might mediate pathophysiological events in the coronary circulation.
BS, Lipton RB, Solomona S, et al. Myocardial ischemia related to ergot alkaloids: a case report and literature review. Headache 1991; 31: 446-50. 2. Perrin VL. Clinical pharmacokinetics of ergotamine in migraine and cluster headache. Clin Pharmacokin 1985; 10: 334-52. 3. Farmacotherapeutisch Kompas 1992. Amstelveen, The Netherlands: Ziekenfondsraad, 1992: 288-301. 4. The Subcutaneous Sumatriptan International Study Group. Treatment of migraine attacks with sumatriptan. N Engl J Med 1991; 325: 316-21.
Department of Cardiothoracic Surgery, National Heart and Lung Institute,
SiR,—The HER2 gene, located on the long arm of chromosome 17, codes for a protein with the characteristics of a growth factor
Heart Science Centre, Harefield Hospital, Middlesex UB9 6JH, UK
A. H. CHESTER G. S. O’NEIL M. H. YACOUB
1. Chester AH, Martin GR, Bodelsson, et al. 5-hydroxytryptamine receptor profile in healthy and diseased human epicardial coronary arteries. Cardiovasc Res 1990; 24: 932-37. 2. Chester AH, Allen SP, Tadjkarimi S, Yacoub MH. Interaction between thromboxane A2 and 5-hydroxytryptamine receptor subtypes in human coronary arteries. Circulation 1993; 87: 874-80.
SIR,-Dr Ottervanger and colleagues prudently advise caution in the use of sumatriptan in patients with chest pain or "tightness of the chest" after use of sumatriptan. As they mention, myocardial ischaemia and infarction have also been recorded after ergotamine,l an antimigraine drug that has been used for many decades. However, the fact that their particular patient had not shown signs of myocardial ischaemia after ergotamine, prompts us to put in a word of caution. We compared the contractile effects in the human isolated coronary artery of sumatriptan and the non-selective 5-HTreceptor agonist ergotamine, and also of serotonin. The right coronary artery segments were obtained from 16 heart-beating organ donors who died of non-cardiac disorders (9 male, 7 female; aged 1-49 years). Hearts were provided by the Rotterdam Heart Valve Bank (Bio Implant Services Foundation/Eurotransplant Foundation) after removal of the aortic and pulmonary valves for homograft valve transplantation. Tension was recorded isometrically in organ baths containing oxygenated Kreb’s solution. The ECo and maximum effect (Emax) of the two antimigraine drugs and of serotonin show (table) that ergotamine was about 100 times more potent than sumatriptan. Moreover, the Emax of ergotamine was twice that of sumatriptan.
1. Galer
Tumour HER2 protein in breast cancer and
family history receptor.1,2 The ligand may belong to a family of proteins called heregulins.3 HER2 protein overexpression in breast cancers is associated with poor prognosis.4 Moreover, HER2 overexpression is positively correlated with inflammatory tumours, higher tumour mitotic activity, high tumour grade, large tumour size, and young age. HER2 overexpression is negatively correlated with tumour oestrogen and progesterone receptor content. We report that high tumour HER2 protein is also associated with a family history of breast cancer. We measured HER2 protein in 20 women who had breast cancers excised in our centre. HER2 protein in the tumours was measured by enzyme-linked immunosorbent assay (ELISA)6 which, being quantitative, may be preferable to immunohistochemistry or western blotting, which are semiquantitative .6 All assays were done by Dianon Systems, Stratford, Connecticut. HER2 concentrations were significantly higher in women with a family history of breast cancer (p=00025, two-tailed MannWhitney U, figure). The 10 women with family history were
predominantly post-menopausal (mean age 64-4 [SE 3 1] vs 57 [4-7] for the other 10; not significantly different, t-test). There was no significant difference in tumour DNA index between the two groups (1-13 [0’08] vs 1-28 [0-12]). Of the 10 women with a family history of breast cancer, 4 had a first-degree relative with the disease (mother or sister); the other 6 had a second-degree relative with breast cancer (aunts, cousins, or a niece). A familial breast-ovarian cancer gene, BRCAI, has been localised to the long arm of chromosome 17.7,8 This gene, associated with breast cancer in women under age 45, is not the HER2 gene.2 In addition, there is evidence implicating at least two genes on the short arm of chromosome 17 in breast cancer
carcinogenesis.9
CASEWISE CONCORDANCE FOR DISEASES REPORTED BY AUSTRALIAN TWINS
As
of
I re-examined data from questionnaire surveys of 3808 adult twin pairs.3 These data included zygosity, reported birthweight, and occurrence of several common diseases. Single placenta MZ twins (of whom 75% will be monochorionic) did not differ significantly from DZ twins in prevalence of any of these diseases or in concordance between single and double placenta MZ twins (table). I conclude that effects such as those Phillips is concerned about are not large enough to alter the broad conclusions made in the classic twin design about genetic a test
Phillips’ hypothesis,
determination. Epidemiology Unit, Queensland Institute of Medical Research, Herston, Queensland 4029, Australia
affected. Methodologically sound studies that address this issue in adult disorders are rare, but, for example, for psychiatric disorders4 over-representation of twins is not generally the case. Schizophrenia, in single surviving twins may be the exception,and some studies suggest that prenatal factors may be related to discordance for schizophrenia in monozygotic twins.5 Twin study methods have far wider application than the estimation of heritability.’ The comparison of results obtained by various genetic epidemiological methods7 enables information to be gathered about both the genetic and environmental influences on disease; proof of genetic aetiology is not what is being sought. There are other issues relating to classic twin studies of adult disease, such as ascertainment and investigator biases, and methods of analysis, which may have far greater impact on the validity of results and receive considerably less attention. Twins remain an important natural experimental model and the destructive divisions of the nature-nurture controversy are meaningless in modem medical research. The paradigm is no longer "nature or nurture?" but "how do genes and environment operate?". Twin studies continue to have much to offer in finding the answers. Genetics Section, Institute of Psychiatry, De Crespigny Park, London SE5 8AF, UK
ALISON M. MACDONALD
MacGillivray I, Campbell D, Thompson B, eds. Twinning and twins. Ch 9-12. Chichester: John Wiley, 1988. 2. Melnick M, Myrianthopoulos NC. The effects of chorion type on normal and abnormal developmental variation in monozygous twins. Am J Med Genet 1979; 4: 1.
147-56. 3.
4.
D. L. DUFFY 5.
1. Vlietinck R, Derom R, Neale MC, et al. Genetic and environmental variation in the birth weight of twins. Behav Genet 1989; 19: 151-61. 2. Reed T, Christian JC, Wood PD, Schaefer EJ. Influence of placentation on high density lipoprotein in adult males: the NHLBI twin study. Acta Genet Med Gemellol 1991; 40: 353-59. 3. Duffy DL, Battistutta D, Martin NG, Hopper JL, Matthews JD. Genetics of asthma and hayfever in Australian twins. Am Rev Respir Dis 1990 142: 1351-58.
1419
6. 7.
Corney G, MacGillivray I, Campbell DM, Thompson B, Little J. Congenital anomalies in twins in Aberdeen and North-East Scotland. Acta Genet Med Gemellol 1983; 32: 31-35. Chitkara B, Macdonald AM, Reveley AM. Twin birth and adult psychiatric disorder: an examination of the case records of the Maudsley Hospital. Br J Psychiatry 1988; 152: 391-98. Lewis SW, Chitkara B, Reveley AM, Murray RM. Family history and birth weight in monozygotic twins concordant and discordant for psychosis. Acta Genet Med Gemellol 1987; 36: 267-73. McGuffin P, Katz R. Who believes in estimating heritability as an end in itself? Behav Brain Sci 1990; 13: 141-42. Morton NE, Rao DC, Lalouel J-M. Methods in genetic epidemiology. Basel: Karger, 1983.
Sumatriptan and ischaemic heart disease SIR,-Dr Phillips says that twin studies may have misled us into believing a genetic origin of many diseases. His basic premise that the adverse prenatal environment of monozygotic twins relative to that of dizygotic twins will result in increased monozygotic concordance and hence to overestimates of genetic aetiology is fallacious, even by his own evidence. Phillips notes that monozygotic twins have greater within-pair variability of birthweight than dizygotic twins-ie, they are more discordant. The relation of birthweight to later development in twins is complex and only where the intra-pair difference is extreme have differences in longer term outcome been detected.1 In discussing the higher rates of congenital malformations he does not mention that the most recent data2,3 show no difference in rates in monochorionic and dichorionic monozygotic twins, nor that discordance is the norm for congenital malformations in surviving twins, be they monozygotic or dizygotic.1 Perinatal mortality also frequently results in the survival of a single infant;1 such single survivors, who might be regarded as most at risk under Phillips’ hypothesis, are rarely included in classic twin studies, and concordant pairs are never so. Overwhelmingly, differences in the prenatal environment of twins lead to discordance, and make monozygotic twins less alike. On present evidence, the classic twin study will not lead to erroneous overestimation of genetic influence on adult diseases in which prenatal factors are important in aetiology, but, if anything, to underestimation. It is, as Phillips notes, difficult to determine whether prenatal factors affecting twins influence the development of adult disease. However, if this is the case, it follows that twins per se should have higher rates of the diseases in question than the general population. If monozygotic twins are more adversely affected than dizygotic twins, a disproportionate number of monozygotic pairs should be
SIR,-Dr Ottervanger and colleagues report (April 3, p 861) the incidence ofa transmural myocardial infarction after administration of 6 mg sumatriptan subcutaneously in a patient with cluster headache. They point out that the effect of sumatriptan on the human coronary vasculature has been reported to be mild, but offer no explanation of how this 5-hydroxytryptamine-like selective agonist could induce an acute coronary event severe enough to cause a myocardial infarction. Data from our laboratory have also indicated that contractile effects mediated by 5-HTB-like receptors are only weak in non-atherosclerotic human coronary arteries.1 However, further studies demonstrated that the effect of serotonin, acting specifically at 5-HTB-like receptors in healthy human coronary arteries, could be substantially enhanced in the presence of the thromboxane mimetic U46619.2 We have been able to confirm that a similar interaction occurs between U46619 and sumatriptan in human isolated non-atherosclerotic coronary arteries (figure). Dr Maclean and colleagues (April 24, p 1092) propose an explanation for the anomaly between the weak constrictor properties of sumatriptan in apparently normal coronary arteries and the occurrence of myocardial infarction. They suggest that basal release of nitric oxide in the healthy coronary artery is sufficient to inhibit the vessels’ response to sumatriptan, but in smokers, in whom there is a loss of endothelial function, the effect of sumatriptan is unmasked. We believe that stimulation of 5-HT l-like receptors by serotonin, under certain conditions, may cause flow-limiting stenosis resulting in myocardial ischaemia and infarction. This could arise if the effect of selective 5-HTclike receptor agonists, such as sumatriptan, is enhanced by the presence ofthromboxane- as well as conditions
1420
THE LANCET
AND Emax EXPRESSED AS PERCENTAGE OF MAXIMUM CONTRACTILE RESPONSE TO POTASSIUM,
EC50 (nmol/L)
100 mmol/L*
*Mean
1101
Contractile effect of sumatriptan in absence of U46619 (open circles, n = 4), and in presence of EC,o concentration (filled circles, n = 4), EC,o concentration (open triangles), or ECso concentration (filled triangles) of U46619 on nonatherosclerotic epicardial coronary arteries.
Specimens were obtained from 4 patients aged 25, 29, 43, and receiving heart transplantation for cardiomyopathy. Experiments were done in duplicate in vessels from each patient. (Initial response U46619concentrations shown by different starting points for each curve on the y axis.) M mol/L.
54 years,
=
(SE) shown, n=6-11, tsubcutaneous.
The recommended dose of subcutaneously administered ergotamine is 10 to 48 times lower than for subuctaneous sumatriptan2-4 (table). However, the 100-fold higher potency, combined with the 2-fold higher maximum effect, as well as the much longer duration of action of ergotamine, suggests that, in general, the cardiac liability of ergotamine is higher than that of sumatriptan. Both drugs are contraindicated in patients with known coronary artery disease. Possibly, the oral administration of sumatriptan-if feasible-would lead to a less rapid rise in plasma concentrations, which could be safer in patients who may be at risk from coronary side-effects. Department of Pharmacology, Faculty of Medicine and Health Sciences, Erasmus University Rotterdam, 3000 DR Rotterdam, Netherlands
WILLEM A. BAX PRAMOD R. SAXENA
of abrogated nitric oxide release. These mechanisms could be of pathophysiological importance in vessels affected by atherosclerotic disease, since we have shown that areas immediately distal to atheromas are hyperreactive to sumatriptan.1 The possibility exists that some patients receiving sumatriptan could have newly formed yet undetected atheroma, with an associated area of vascular hyperreactivity. Care should be taken when giving this drug to patients with coronary risk factors. The physiological role of 5-HT1-like receptors remains unclear, but in certain circumstances these receptors might mediate pathophysiological events in the coronary circulation.
BS, Lipton RB, Solomona S, et al. Myocardial ischemia related to ergot alkaloids: a case report and literature review. Headache 1991; 31: 446-50. 2. Perrin VL. Clinical pharmacokinetics of ergotamine in migraine and cluster headache. Clin Pharmacokin 1985; 10: 334-52. 3. Farmacotherapeutisch Kompas 1992. Amstelveen, The Netherlands: Ziekenfondsraad, 1992: 288-301. 4. The Subcutaneous Sumatriptan International Study Group. Treatment of migraine attacks with sumatriptan. N Engl J Med 1991; 325: 316-21.
Department of Cardiothoracic Surgery, National Heart and Lung Institute,
SiR,—The HER2 gene, located on the long arm of chromosome 17, codes for a protein with the characteristics of a growth factor
Heart Science Centre, Harefield Hospital, Middlesex UB9 6JH, UK
A. H. CHESTER G. S. O’NEIL M. H. YACOUB
1. Chester AH, Martin GR, Bodelsson, et al. 5-hydroxytryptamine receptor profile in healthy and diseased human epicardial coronary arteries. Cardiovasc Res 1990; 24: 932-37. 2. Chester AH, Allen SP, Tadjkarimi S, Yacoub MH. Interaction between thromboxane A2 and 5-hydroxytryptamine receptor subtypes in human coronary arteries. Circulation 1993; 87: 874-80.
SIR,-Dr Ottervanger and colleagues prudently advise caution in the use of sumatriptan in patients with chest pain or "tightness of the chest" after use of sumatriptan. As they mention, myocardial ischaemia and infarction have also been recorded after ergotamine,l an antimigraine drug that has been used for many decades. However, the fact that their particular patient had not shown signs of myocardial ischaemia after ergotamine, prompts us to put in a word of caution. We compared the contractile effects in the human isolated coronary artery of sumatriptan and the non-selective 5-HTreceptor agonist ergotamine, and also of serotonin. The right coronary artery segments were obtained from 16 heart-beating organ donors who died of non-cardiac disorders (9 male, 7 female; aged 1-49 years). Hearts were provided by the Rotterdam Heart Valve Bank (Bio Implant Services Foundation/Eurotransplant Foundation) after removal of the aortic and pulmonary valves for homograft valve transplantation. Tension was recorded isometrically in organ baths containing oxygenated Kreb’s solution. The ECo and maximum effect (Emax) of the two antimigraine drugs and of serotonin show (table) that ergotamine was about 100 times more potent than sumatriptan. Moreover, the Emax of ergotamine was twice that of sumatriptan.
1. Galer
Tumour HER2 protein in breast cancer and
family history receptor.1,2 The ligand may belong to a family of proteins called heregulins.3 HER2 protein overexpression in breast cancers is associated with poor prognosis.4 Moreover, HER2 overexpression is positively correlated with inflammatory tumours, higher tumour mitotic activity, high tumour grade, large tumour size, and young age. HER2 overexpression is negatively correlated with tumour oestrogen and progesterone receptor content. We report that high tumour HER2 protein is also associated with a family history of breast cancer. We measured HER2 protein in 20 women who had breast cancers excised in our centre. HER2 protein in the tumours was measured by enzyme-linked immunosorbent assay (ELISA)6 which, being quantitative, may be preferable to immunohistochemistry or western blotting, which are semiquantitative .6 All assays were done by Dianon Systems, Stratford, Connecticut. HER2 concentrations were significantly higher in women with a family history of breast cancer (p=00025, two-tailed MannWhitney U, figure). The 10 women with family history were
predominantly post-menopausal (mean age 64-4 [SE 3 1] vs 57 [4-7] for the other 10; not significantly different, t-test). There was no significant difference in tumour DNA index between the two groups (1-13 [0’08] vs 1-28 [0-12]). Of the 10 women with a family history of breast cancer, 4 had a first-degree relative with the disease (mother or sister); the other 6 had a second-degree relative with breast cancer (aunts, cousins, or a niece). A familial breast-ovarian cancer gene, BRCAI, has been localised to the long arm of chromosome 17.7,8 This gene, associated with breast cancer in women under age 45, is not the HER2 gene.2 In addition, there is evidence implicating at least two genes on the short arm of chromosome 17 in breast cancer
carcinogenesis.9