125
attempt external version of the second twin or internal podalic version with total breech extraction. The first stage of labour was completed in 8 h. The first fetus delivered spontaneously in the occiput anterior position after 40 min. The second twin remained transverse. External version under ultrasound monitoring was unsuccessful. Delivery by total breech extraction was started. The amniotic sac was left intact. When the feet were grasped to begin internal podalic version the uterus contracted down upon the operator’s forearm. The anaesthetist administered sublingual glyceryl trinitrate by aerosol in two boluses of 400 Ag each. The uterus relaxed within about 30 s, permitting internal version and total breech extraction without difficulty. The placentas delivered spontaneously and the uterus contracted normally after the third stage of labour was over. The blood loss was about 500 ml. There was no uterine atony. The twins, both female, weighed 2889 and 3089 g and 1 min/5 min Apgar scores were 9/9 and 8/9, respectively. Glyceryl trinitrate has been used to obtain uterine relaxation to permit delivery of retained placenta.3,4The sublingual route permits rapid relaxation of the uterus in 30 to 60 s, and avoids the need for endotracheal intubation and general anaesthesia with their attendant risks. Glyceryl trinitrate can be administered quickly. Although an intravenous route is not needed, venous access is prudent to ensure that the patient is normovolaemic and in the event that other medications must be administered. No adverse effects on the fetus were noted when glyceryl trinitrate continuous infusion (albeit at lower doses) was used for several hours in patients with severe pre-eclampsia.5 Short-term use to assist breech delivery would not be expected to affect the fetus
adversely. In an out-of-hospital setting,
a trial of glyceryl trinitrate to achieve uterine relaxation and release an entrapped aftercoming head might be less traumatic than Duhrssen’s incisions, and could be attempted while preparations for general anaesthesia are being made. Furthermore, glyceryl trinitrate may be useful in the management of uterine inversion because it is easier to adminster than the accepted medications 6 Although glyceryl trinitrate sublingual spray is not approved for marketing for obstetric applications, its use for obstetric emergencies deserves further evaluation.
Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Cedars-Sinai Medical Center, Los Angeles, California 90048, USA
JEFFREY S. GREENSPOON ANITA KOVACIC
FA, Johnson RE, Youcha S, Hobbins JC, Berkowitz RL. Intrapartum management of twin gestation. Obstet Gynecol 1985; 65: 119-214. 2. Rabinovici J, Barkai G, Reichman B, Serr DM, Mashiach S. Internal podalic version 1. Chervenak
with unruptured membranes for the second twin in transverse lie. Obstet Gynecol 1988; 71: 428-30. 3. Peng ATC, Gorman RS, Shulman SM, Demarchis R, Nyunt K, Blancato LS. Intravenous nitroglycerin for uterine relaxation in the postpartum patient with retained placenta. Anesthesiology 1989; 71: 172-73. 4. Desimone CA, Norris MC, Leighton BL. Intravenous nitroglycerin aids in manual extraction of a retained placenta. Anesthesiology 1990; 73: 787. 5. Cotton DB, Jones MM, Longmire S, Dorman KF, Tessem J, Joyce TH III. Role of intravenous nitroglycerin in the treatment of severe pregnancy-induced hypertension complicated by pulmonary edema. Am J Obstet Gynecol 1986; 154: 91-93. 6. Brar HS, Greenspoon JS, Platt LD, Paul RH. Acute puerperal uterine inversion. New approaches to management. J Reprod Med 1989; 34: 173-77.
Self-help for childbirth SiR,—The Safe Motherhood Initiative encourages developing countries to focus on maternal health. Training of birth attendants and the high-risk approach are underlined as important strategies. Self-delivery in the bush is common in Papua New Guinea and in islands in the South Pacific, where vaginal blood and amniotic fluid are seen as having evil effects. A woman in labour must stay away from her home and any person or object coming in contact with her is believed to be contaminated and subject to misfortune. There is, therefore, a reluctance to assist a woman in labour; even to offer her transport is thought to be inviting bad luck. Many women therefore give birth by themselves in special huts built in the jungle by the husbands or in the open air. She cannot be seen in the village for one week after delivery and to get food during this period she has to get a
vegetable garden ready well in advance. Water has to be obtained from springs; this may be difficult since she cannot use her usual source.
Because of this unassisted self-delivery in the bush women do not get the chance to see a baby being born and they are inexperienced when it is their turn. Women may make one or more visits to a clinic for antenatal care late in pregnancy but they usually choose to deliver in the traditional fashion, especially if multiparous. The risks of this traditional practice are very high. Perinatal mortality data for rural areas are not available and maternal mortality is grossly underreported. Estimates of maternal mortality are based on those contacting the health services at some time during delivery, and bush deliveries, which account for more than half the total, are not included. Estimates of maternal mortality range from 1to 20 per 1000 in rural areas with an average of 8 per 1000 births.’ In Papua New Guinea the lifetime risk of death in pregnancy and childbirth is very high. In some rural areas the risk of pregnancy-related death is 1 in 15, haemorrhage probably being the most common cause especially since anaemia due to malaria is universal. Mother and child health policies in Papua New Guinea are aimed at increasing antenatal care and institutional deliveries. Traditional birth attendants in the usual sense do not exist and training of local women to this role has not proved successful.’ It takes time to change traditions and attitudes. In the meantime, it is necessary to introduce the concept of "Self-help for childbirth". Self-reliance is accepted in primary health care but it would be a new dimension to the Safe Motherhood Initiative. Because of low literacy rates, visual or audiovisual aids to help women understand pregnancy and childbirth are required. The characters in such audiovisual material must be local, so that women can relate to them. Women need to be encouraged to discuss pregnancy and delivery among themselves to spread simple messages about clean delivery, the avoidance of a standing birth position, simple resuscitation, fresh-cut bamboo to cut the umbilical cord, avoidance of cutting the cord too short or pulling on it to loosen the placenta, abdominal massage to reduce bleeding, and immediate breastfeeding. Although more concerted efforts are essential to reduce maternal mortality by half by the year 2000 (the goal of the Safe Motherhood Inititative) an educational approach to spread basic knowledge of self-help will mean that more women survive childbirth. Department of Gynaecology and Obstetrics, Hospital, University of Oslo, 0514 Oslo 5, Norway;
Aker
and Technical and Evaluation Division, UN Fund for Population Activities, New York, NY, USA
BABILL STRAY-PEDERSEN
USHA SHAH LALAN MUBIALA
JE. The health of women in Papua New Guinea. Papua New Guinea Inst Monogr 1990, no 90: 1-180. 2. Papua New Guinea National Health Plan 1986-1990. Port Moresby: Department of Health, 1986. 1. Gillett
Caesarean section rates SiR,—The Association of Professors of Obstetrics and knows that there have never been any statutory collected data on caesarean section rates for the UK. We wrote that "It has been difficult to get recent data on the caesarean section rates in the United Kingdom" and the phrase Ms Macfarlane (June 15, p 1481) complains of was entered by The Lancet-an attempt at brevity that has introduced confusion. A more important point is that those who are epidemiologically trained should not try to lead us into thinking that the Scottish figures represent the rest of the UK. For two reasons they cannot. Firstly, the number of recorded births in Scotland is small, about the same as that of a large English regional health authority; few would consider such data as characteristic of the UK. Secondly, the of Scotland may have sociobiologically different women characteristics from women in the rest of the UK; patterns of childbirth may differ. For example, Mr Leyland refers to the higher proportion of first births to older women in Scotland, two well-known factors of higher risk for operative delivery. When
Gynaecology
126
comparing the data of Scotland and England, on the rugby football field one may be comparing like with like-this is not so in the perinatal field. The Association of Professors of Obstetrics and Gynaecology admire the prompt and well thought out service given to obstetrics by the Information and Statistics Division of the Scottish Health Service. Would that it were so in England, for we have lagged behind. The Association will consider Macfarlane’s suggestion of recommending to the Royal College of Obstetricians and Gynaecologists that recognition should be withdrawn from English hospitals whose maternity units fail to contribute Komer data on the Maternity Hospital Episode System. However, until the new data are available for England, we consider that the contribution made by our Association was a useful one for people who want to look at the trends of caesarean section rates in the four constituent parts of the UK. Association of Professors of Obstetrics and Gynaecology, 27 Sussex Place, London NW1 4RG, UK
GEOFFREY CHAMBERLAIN, Chairman
Cholesterol lowering and non-cardiac
mortality SIR,-"Statistical correlation carries no conviction as a cause and relationship". So wrote Sir John McMichael 15 years ago,! commenting on the recommendations of a joint working-party of the Royal College of Physicians and British Cardiac Society, which included Prof Michael Oliver. Nowadays most would agree that the relation Sir John was criticising-the role of hypercholesterolaemia in coronary heart disease-has been well and truly established as causal, as summarised in The Lancet by Steinberg.2 effect
Professor Oliver’s
current
contention that
treatment
of
hypercholesterolaemia increases non-cardiac mortality (June 22, p 1529) is based solely on statistical evidence. He provides no data to support his hypothesis that changes in cell-membrane composition occur and in some unspecified way cause an increase in a wide variety of fatal accidents and diseases in patients receiving a whole range of cholesterol-lowering diets and drugs. He cites Muldoon et alwhose analysis included the Minnesota Coronary Survey. In that study inmates of six mental hospitals who were exposed to a lipid-lowering diet for just over 1 year on average had a significant excess of accidental deaths compared with those not on diet. The causes included fractures, drug reactions, bums, foreign bodies, tooth extractions, freezing, heatstroke, drowning, and suicide.4 Until there is evidence of a causal mechanism that explains how and why these and other non-cardiac causes of mortality should increase as a consequence of lipid-lowering therapy I shall continue to treat hypercholesterolaemic patients who have established coronary heart disease or are at high risk of premature death from this cause. However, I agree that careful documentation of non-cardiovascular deaths in
trials of HMG CoA reductase inhibitors is important, especially since the degree of cholesterol reduction these drugs engender is so much greater than that achieved in previous trials. MRC Lipoprotein Team, Hammersmith Hospital, GILBERT R. THOMPSON London W12 0HS, UK current
1. McMichael
2.
J. Prevention of coronary heart disease. Lancet 1976; ii: 569.
Steinberg D. Consensus conference on cholesterol and heart disease. Lancet 1985; ii:
205-07. 3. Muldoon MF, Manuck
SB, Matthews KA. Lowering cholesterol concentrations and mortality: a quantitative review of primary prevention trials. Br Med J 1990; 301:
309-14. 4. Frantz ID, Dawson EA, Ashman PL, et al. Test of effect of lipid cardiovascular risk. Arteriosclerosis 1989; 9: 129-35.
that lower cholesterol by more than the 10% or so observed in the trials examined by Oliver and Muldoon et al. A long-term open-label study of lovastatin, coordinated by Merck Sharp & Dohme Research Laboratories, has just been completed and the data are being analysed. 744 patients, over 90% in North America, with severe hypercholesterolaemia (on-diet baseline plasma cholesterol 9-3 mmol/1) were followed up for an average of five years. In most patients daily doses were titrated up to the maximum recommended 80 mg, and over half received other lipid-lowering agents, usually resins, concomitantly. The average reduction in plasma cholesterol exceeded 30%. There were no deaths from suicide, accidents, or violence. Bradford et aP reported no trauma deaths in a very large study in which over 6000 US patients received lovastatin at various dosages for 48 weeks. The average plasma cholesterol reduction was 17-29%, depending on dosage. In these two studies with almost 10 000 patient-years of vigorous cholesterol-lowering therapy with lovastatin there has been not one death from trauma. By comparison, in the LRC-CPPT study,4also in North America, there were 4 trauma deaths in the placebo group and 11 in the cholestyramine group, each group of 1900 patients contributing about 14 000 patient-years. The apparent increase in such deaths noted by Oliver and Muldoon may be a statistical artifact-or at least not shared by all lipid-lowering therapies. Oliver cites as evidence for membrane changes the in-vitro work of Cutts and Bankhurst,s who reported that lovastatin reduced natural killer cell (NK) cytotoxicity and phytohaemagglutinin(PHA) stimulated proliferation of lymphocytes. We have looked for these effects in a double-blind study in which 51 hypercholesterolaemic patients were randomised to lovastatin 20 mg twice daily or placebo for 8 weeks. We found no effects on NK activity or PHA proliferation in lymphocytes from these patients.
lowering by diet on
SIR,-Professor Oliver (and Muldoon et all) express concern that cholesterol-lowering therapy may increase the risk of non-cardiac mortality-notably death due to suicide, accidents, and violence. When Wysowski and Gross2 analysed on a case-by-case basis trauma deaths in two of the most important studies in the meta-analysis of Muldoon et al, they found a causal relation implausible. If lowering blood cholesterol does increase the risk of trauma death, a more pronounced effect would be expected with therapies
Merck Sharp & Dohme Research Laboratories, PO Box 2000, Rahway, New Jersey 07065, USA
JONATHAN A. TOBERT
1. Muldoon MF, Manuck
SB, Matthews KA. Lowering cholesterol concentrations and mortality: a quantitative review of primary prevention trials. Br Med J 1990; 301: 309-14.
Wysowski DK, Gross TP. Deaths due to accidents and violence in two recent trials of cholesterol-lowering drugs. Arch Intern Med 1990; 150: 2169-72. 3. Bradford RH, Shear CL, Chremos AN, et al. Expanded clinical evaluation of lovastatin (EXCEL) study results. Arch Intern Med 1991; 151: 43-49. 4. Lipid Research Clinics Program. The lipid research clinics coronary primary prevention trial results. JAMA 1984; 251: 351-64. 5. Cutts JL, Bankhurst AD. Suppression of lymphoid cell function in vitro by inhibition of 3-hydroxy-3-methylglutaryl coenzyme A reductase by lovastatin. Int J Immunopharmacol 1989; 11: 863-69. 2.
Linoleic acid inhibition of adhesion of enteropathogenic Escherichia coli to HEp-2 cells SIR,-Enteropathogenic Escherichia coli (EPEC) consist of serogroups traditionally associated with outbreaks of infantile enteritis1 and constitute an important cause of diarrhoea in infants in developing countries.2,3 The mechanism by which strains of EPEC cause diarrhoea is unknown; classic virulence properties, such as currently recognised toxins and colonisation factors, do not seem to be involved. The ability of EPEC to adhere to certain cells in vitro correlates with the attachment of EPEC to the intestinal mucosa in vivo,4 and this is one possible virulence mechanism. EPEC adhere to HEp-2 cells either as small foci (localised adherence) or over the entire surface of the cells (diffuse). Localised adhesion is dependent on a 50-70 MDa plasmid and strains carrying this plasmid (EAF) express an outer membrane protein of 94 kDa.6,7 We investigated the ability of EPEC carrying EAF plasmid to adhere to HEp-2 cells in the presence of a range of fatty acids. Twenty E coli strains of recognised EPEC serotypes were used in a HEp-2 cell adhesion assay.’ Over a 3 h attachment period strain E20513 (0111:H2) was consistently found to attach, in a localised manner, to 60% or so of HEp-2 cells. However, in the presence of a mixture of 1 nl of each of five fatty acids (palmitic, stearic, oleic, linoleic, and linolenic; Sigma Chemical) adhesion was completely inhibited. When the fatty acids were tested separately only linoleic acid inhibited adhesion. Tests with the other nineteen strains gave
attempt external version of the second twin or internal podalic version with total breech extraction. The first stage of labour was completed in 8 h. The first fetus delivered spontaneously in the occiput anterior position after 40 min. The second twin remained transverse. External version under ultrasound monitoring was unsuccessful. Delivery by total breech extraction was started. The amniotic sac was left intact. When the feet were grasped to begin internal podalic version the uterus contracted down upon the operator’s forearm. The anaesthetist administered sublingual glyceryl trinitrate by aerosol in two boluses of 400 Ag each. The uterus relaxed within about 30 s, permitting internal version and total breech extraction without difficulty. The placentas delivered spontaneously and the uterus contracted normally after the third stage of labour was over. The blood loss was about 500 ml. There was no uterine atony. The twins, both female, weighed 2889 and 3089 g and 1 min/5 min Apgar scores were 9/9 and 8/9, respectively. Glyceryl trinitrate has been used to obtain uterine relaxation to permit delivery of retained placenta.3,4The sublingual route permits rapid relaxation of the uterus in 30 to 60 s, and avoids the need for endotracheal intubation and general anaesthesia with their attendant risks. Glyceryl trinitrate can be administered quickly. Although an intravenous route is not needed, venous access is prudent to ensure that the patient is normovolaemic and in the event that other medications must be administered. No adverse effects on the fetus were noted when glyceryl trinitrate continuous infusion (albeit at lower doses) was used for several hours in patients with severe pre-eclampsia.5 Short-term use to assist breech delivery would not be expected to affect the fetus
adversely. In an out-of-hospital setting,
a trial of glyceryl trinitrate to achieve uterine relaxation and release an entrapped aftercoming head might be less traumatic than Duhrssen’s incisions, and could be attempted while preparations for general anaesthesia are being made. Furthermore, glyceryl trinitrate may be useful in the management of uterine inversion because it is easier to adminster than the accepted medications 6 Although glyceryl trinitrate sublingual spray is not approved for marketing for obstetric applications, its use for obstetric emergencies deserves further evaluation.
Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Cedars-Sinai Medical Center, Los Angeles, California 90048, USA
JEFFREY S. GREENSPOON ANITA KOVACIC
FA, Johnson RE, Youcha S, Hobbins JC, Berkowitz RL. Intrapartum management of twin gestation. Obstet Gynecol 1985; 65: 119-214. 2. Rabinovici J, Barkai G, Reichman B, Serr DM, Mashiach S. Internal podalic version 1. Chervenak
with unruptured membranes for the second twin in transverse lie. Obstet Gynecol 1988; 71: 428-30. 3. Peng ATC, Gorman RS, Shulman SM, Demarchis R, Nyunt K, Blancato LS. Intravenous nitroglycerin for uterine relaxation in the postpartum patient with retained placenta. Anesthesiology 1989; 71: 172-73. 4. Desimone CA, Norris MC, Leighton BL. Intravenous nitroglycerin aids in manual extraction of a retained placenta. Anesthesiology 1990; 73: 787. 5. Cotton DB, Jones MM, Longmire S, Dorman KF, Tessem J, Joyce TH III. Role of intravenous nitroglycerin in the treatment of severe pregnancy-induced hypertension complicated by pulmonary edema. Am J Obstet Gynecol 1986; 154: 91-93. 6. Brar HS, Greenspoon JS, Platt LD, Paul RH. Acute puerperal uterine inversion. New approaches to management. J Reprod Med 1989; 34: 173-77.
Self-help for childbirth SiR,—The Safe Motherhood Initiative encourages developing countries to focus on maternal health. Training of birth attendants and the high-risk approach are underlined as important strategies. Self-delivery in the bush is common in Papua New Guinea and in islands in the South Pacific, where vaginal blood and amniotic fluid are seen as having evil effects. A woman in labour must stay away from her home and any person or object coming in contact with her is believed to be contaminated and subject to misfortune. There is, therefore, a reluctance to assist a woman in labour; even to offer her transport is thought to be inviting bad luck. Many women therefore give birth by themselves in special huts built in the jungle by the husbands or in the open air. She cannot be seen in the village for one week after delivery and to get food during this period she has to get a
vegetable garden ready well in advance. Water has to be obtained from springs; this may be difficult since she cannot use her usual source.
Because of this unassisted self-delivery in the bush women do not get the chance to see a baby being born and they are inexperienced when it is their turn. Women may make one or more visits to a clinic for antenatal care late in pregnancy but they usually choose to deliver in the traditional fashion, especially if multiparous. The risks of this traditional practice are very high. Perinatal mortality data for rural areas are not available and maternal mortality is grossly underreported. Estimates of maternal mortality are based on those contacting the health services at some time during delivery, and bush deliveries, which account for more than half the total, are not included. Estimates of maternal mortality range from 1to 20 per 1000 in rural areas with an average of 8 per 1000 births.’ In Papua New Guinea the lifetime risk of death in pregnancy and childbirth is very high. In some rural areas the risk of pregnancy-related death is 1 in 15, haemorrhage probably being the most common cause especially since anaemia due to malaria is universal. Mother and child health policies in Papua New Guinea are aimed at increasing antenatal care and institutional deliveries. Traditional birth attendants in the usual sense do not exist and training of local women to this role has not proved successful.’ It takes time to change traditions and attitudes. In the meantime, it is necessary to introduce the concept of "Self-help for childbirth". Self-reliance is accepted in primary health care but it would be a new dimension to the Safe Motherhood Initiative. Because of low literacy rates, visual or audiovisual aids to help women understand pregnancy and childbirth are required. The characters in such audiovisual material must be local, so that women can relate to them. Women need to be encouraged to discuss pregnancy and delivery among themselves to spread simple messages about clean delivery, the avoidance of a standing birth position, simple resuscitation, fresh-cut bamboo to cut the umbilical cord, avoidance of cutting the cord too short or pulling on it to loosen the placenta, abdominal massage to reduce bleeding, and immediate breastfeeding. Although more concerted efforts are essential to reduce maternal mortality by half by the year 2000 (the goal of the Safe Motherhood Inititative) an educational approach to spread basic knowledge of self-help will mean that more women survive childbirth. Department of Gynaecology and Obstetrics, Hospital, University of Oslo, 0514 Oslo 5, Norway;
Aker
and Technical and Evaluation Division, UN Fund for Population Activities, New York, NY, USA
BABILL STRAY-PEDERSEN
USHA SHAH LALAN MUBIALA
JE. The health of women in Papua New Guinea. Papua New Guinea Inst Monogr 1990, no 90: 1-180. 2. Papua New Guinea National Health Plan 1986-1990. Port Moresby: Department of Health, 1986. 1. Gillett
Caesarean section rates SiR,—The Association of Professors of Obstetrics and knows that there have never been any statutory collected data on caesarean section rates for the UK. We wrote that "It has been difficult to get recent data on the caesarean section rates in the United Kingdom" and the phrase Ms Macfarlane (June 15, p 1481) complains of was entered by The Lancet-an attempt at brevity that has introduced confusion. A more important point is that those who are epidemiologically trained should not try to lead us into thinking that the Scottish figures represent the rest of the UK. For two reasons they cannot. Firstly, the number of recorded births in Scotland is small, about the same as that of a large English regional health authority; few would consider such data as characteristic of the UK. Secondly, the of Scotland may have sociobiologically different women characteristics from women in the rest of the UK; patterns of childbirth may differ. For example, Mr Leyland refers to the higher proportion of first births to older women in Scotland, two well-known factors of higher risk for operative delivery. When
Gynaecology
126
comparing the data of Scotland and England, on the rugby football field one may be comparing like with like-this is not so in the perinatal field. The Association of Professors of Obstetrics and Gynaecology admire the prompt and well thought out service given to obstetrics by the Information and Statistics Division of the Scottish Health Service. Would that it were so in England, for we have lagged behind. The Association will consider Macfarlane’s suggestion of recommending to the Royal College of Obstetricians and Gynaecologists that recognition should be withdrawn from English hospitals whose maternity units fail to contribute Komer data on the Maternity Hospital Episode System. However, until the new data are available for England, we consider that the contribution made by our Association was a useful one for people who want to look at the trends of caesarean section rates in the four constituent parts of the UK. Association of Professors of Obstetrics and Gynaecology, 27 Sussex Place, London NW1 4RG, UK
GEOFFREY CHAMBERLAIN, Chairman
Cholesterol lowering and non-cardiac
mortality SIR,-"Statistical correlation carries no conviction as a cause and relationship". So wrote Sir John McMichael 15 years ago,! commenting on the recommendations of a joint working-party of the Royal College of Physicians and British Cardiac Society, which included Prof Michael Oliver. Nowadays most would agree that the relation Sir John was criticising-the role of hypercholesterolaemia in coronary heart disease-has been well and truly established as causal, as summarised in The Lancet by Steinberg.2 effect
Professor Oliver’s
current
contention that
treatment
of
hypercholesterolaemia increases non-cardiac mortality (June 22, p 1529) is based solely on statistical evidence. He provides no data to support his hypothesis that changes in cell-membrane composition occur and in some unspecified way cause an increase in a wide variety of fatal accidents and diseases in patients receiving a whole range of cholesterol-lowering diets and drugs. He cites Muldoon et alwhose analysis included the Minnesota Coronary Survey. In that study inmates of six mental hospitals who were exposed to a lipid-lowering diet for just over 1 year on average had a significant excess of accidental deaths compared with those not on diet. The causes included fractures, drug reactions, bums, foreign bodies, tooth extractions, freezing, heatstroke, drowning, and suicide.4 Until there is evidence of a causal mechanism that explains how and why these and other non-cardiac causes of mortality should increase as a consequence of lipid-lowering therapy I shall continue to treat hypercholesterolaemic patients who have established coronary heart disease or are at high risk of premature death from this cause. However, I agree that careful documentation of non-cardiovascular deaths in
trials of HMG CoA reductase inhibitors is important, especially since the degree of cholesterol reduction these drugs engender is so much greater than that achieved in previous trials. MRC Lipoprotein Team, Hammersmith Hospital, GILBERT R. THOMPSON London W12 0HS, UK current
1. McMichael
2.
J. Prevention of coronary heart disease. Lancet 1976; ii: 569.
Steinberg D. Consensus conference on cholesterol and heart disease. Lancet 1985; ii:
205-07. 3. Muldoon MF, Manuck
SB, Matthews KA. Lowering cholesterol concentrations and mortality: a quantitative review of primary prevention trials. Br Med J 1990; 301:
309-14. 4. Frantz ID, Dawson EA, Ashman PL, et al. Test of effect of lipid cardiovascular risk. Arteriosclerosis 1989; 9: 129-35.
that lower cholesterol by more than the 10% or so observed in the trials examined by Oliver and Muldoon et al. A long-term open-label study of lovastatin, coordinated by Merck Sharp & Dohme Research Laboratories, has just been completed and the data are being analysed. 744 patients, over 90% in North America, with severe hypercholesterolaemia (on-diet baseline plasma cholesterol 9-3 mmol/1) were followed up for an average of five years. In most patients daily doses were titrated up to the maximum recommended 80 mg, and over half received other lipid-lowering agents, usually resins, concomitantly. The average reduction in plasma cholesterol exceeded 30%. There were no deaths from suicide, accidents, or violence. Bradford et aP reported no trauma deaths in a very large study in which over 6000 US patients received lovastatin at various dosages for 48 weeks. The average plasma cholesterol reduction was 17-29%, depending on dosage. In these two studies with almost 10 000 patient-years of vigorous cholesterol-lowering therapy with lovastatin there has been not one death from trauma. By comparison, in the LRC-CPPT study,4also in North America, there were 4 trauma deaths in the placebo group and 11 in the cholestyramine group, each group of 1900 patients contributing about 14 000 patient-years. The apparent increase in such deaths noted by Oliver and Muldoon may be a statistical artifact-or at least not shared by all lipid-lowering therapies. Oliver cites as evidence for membrane changes the in-vitro work of Cutts and Bankhurst,s who reported that lovastatin reduced natural killer cell (NK) cytotoxicity and phytohaemagglutinin(PHA) stimulated proliferation of lymphocytes. We have looked for these effects in a double-blind study in which 51 hypercholesterolaemic patients were randomised to lovastatin 20 mg twice daily or placebo for 8 weeks. We found no effects on NK activity or PHA proliferation in lymphocytes from these patients.
lowering by diet on
SIR,-Professor Oliver (and Muldoon et all) express concern that cholesterol-lowering therapy may increase the risk of non-cardiac mortality-notably death due to suicide, accidents, and violence. When Wysowski and Gross2 analysed on a case-by-case basis trauma deaths in two of the most important studies in the meta-analysis of Muldoon et al, they found a causal relation implausible. If lowering blood cholesterol does increase the risk of trauma death, a more pronounced effect would be expected with therapies
Merck Sharp & Dohme Research Laboratories, PO Box 2000, Rahway, New Jersey 07065, USA
JONATHAN A. TOBERT
1. Muldoon MF, Manuck
SB, Matthews KA. Lowering cholesterol concentrations and mortality: a quantitative review of primary prevention trials. Br Med J 1990; 301: 309-14.
Wysowski DK, Gross TP. Deaths due to accidents and violence in two recent trials of cholesterol-lowering drugs. Arch Intern Med 1990; 150: 2169-72. 3. Bradford RH, Shear CL, Chremos AN, et al. Expanded clinical evaluation of lovastatin (EXCEL) study results. Arch Intern Med 1991; 151: 43-49. 4. Lipid Research Clinics Program. The lipid research clinics coronary primary prevention trial results. JAMA 1984; 251: 351-64. 5. Cutts JL, Bankhurst AD. Suppression of lymphoid cell function in vitro by inhibition of 3-hydroxy-3-methylglutaryl coenzyme A reductase by lovastatin. Int J Immunopharmacol 1989; 11: 863-69. 2.
Linoleic acid inhibition of adhesion of enteropathogenic Escherichia coli to HEp-2 cells SIR,-Enteropathogenic Escherichia coli (EPEC) consist of serogroups traditionally associated with outbreaks of infantile enteritis1 and constitute an important cause of diarrhoea in infants in developing countries.2,3 The mechanism by which strains of EPEC cause diarrhoea is unknown; classic virulence properties, such as currently recognised toxins and colonisation factors, do not seem to be involved. The ability of EPEC to adhere to certain cells in vitro correlates with the attachment of EPEC to the intestinal mucosa in vivo,4 and this is one possible virulence mechanism. EPEC adhere to HEp-2 cells either as small foci (localised adherence) or over the entire surface of the cells (diffuse). Localised adhesion is dependent on a 50-70 MDa plasmid and strains carrying this plasmid (EAF) express an outer membrane protein of 94 kDa.6,7 We investigated the ability of EPEC carrying EAF plasmid to adhere to HEp-2 cells in the presence of a range of fatty acids. Twenty E coli strains of recognised EPEC serotypes were used in a HEp-2 cell adhesion assay.’ Over a 3 h attachment period strain E20513 (0111:H2) was consistently found to attach, in a localised manner, to 60% or so of HEp-2 cells. However, in the presence of a mixture of 1 nl of each of five fatty acids (palmitic, stearic, oleic, linoleic, and linolenic; Sigma Chemical) adhesion was completely inhibited. When the fatty acids were tested separately only linoleic acid inhibited adhesion. Tests with the other nineteen strains gave
