- Email: [email protected]
Hyperthermic liver perfusion chemotherapy in the foregut carcinoid syndrome
568
LETTERS
to
Vitamin A
the
EDITOR
supplementation
SIR,-Dr West and his colleagues’ study of vitamin A supplementation in Nepal (July 13, p 67) answers the sceptics up to a point. West et al show that full coverage of a malnourished population of children with high-dose (200 000 IU) capsules given every 4 months may reduce death rates in children aged 1-4 years by up to 30%. But what are the implications of this for health planners in countries like Nepal? What is the likely impact of supplementation on total under-5 mortality rates under existing conditions of primary health care? If we assume that supplementation will not reduce deaths in infants under 6 months, which account for 80% of the infant mortality rate,l,2 the impact in Nepal, which in 1989 had an under-5 mortality of 193 per 1000 (infant mortality 125 per 10003), will be a reduction of 27-9 deaths per 1000 ([0-3 x 0.2 x 125] + [03 x 68]) or 14.5%. This is a maximum reduction and assumes full coverage. West et al conducted a rigorous but expensive aid-assisted study which probably used fieldworkers and administrative staff paid and managed outside the government system. Details of the costs were not provided and the reproducibility of this intervention was not discussed. In Nepal nearly 40% of government health expenditure is externally financed from aid; annual per caput expenditure on health is less than$2, of which less than 45 % is spent on district level primary health care.4 There is one village health worker (VHW) for every 4000 population and these workers are usually unsupported and receive inadequate daily allowances and incentives for fieldwork. The household coverage by VHWs is therefore low, and in many poor areas it is non-existent. West et al did their study in a terai (plains) district, where population density is greater than in the logistically difficult middle hill region where more than 50% of the population live. The section of the population most likely to experience vitamin A deficiency (the very poor and low-caste families) are the groups who consistently fail to receive or demand services even when they are moderately accessible. So in terms of sustainable impact on mortality a decision to implement 4-monthly vitamin A supplementation at household level within the existing government health system may achieve at best perhaps 30% of the mortality reduction West et al achievedie, a 5% reduction in under-5 deaths or 9 deaths per 1000. Health planners must decide therefore where vitamin A supplementation sits in the queue of other primary health care interventions and "magic bullets"-before or behind antibiotics for acute respiratory infections, oral rehydration sachets for dehydration, bed-nets for malaria, clean delivery kits for traditional birth attendants, road-tohealth cards, immunisation, weaning education, or promotion of breastfeeding? Certainly vitamin A supplements should be liberally supplied and promoted at health post level and given to VHWs. But international agencies should think carefully before supporting vertically run national vitamin A supplementation programmes. Money might be spent more effectively and sustainably on initiatives to improve an integrated primary health care service through better management and training, and the development of an organisational culture which makes these services more userfriendly for poor and disadvantaged families, one of whose problems is vitamin A deficiency. Centre for International Child Health, Institute of Child Health, London WC1 N 1 EH, UK
ANTHONY M. de L. COSTELLO
1. Ashworth A, Waterlow JC. Infant mortality in developing countries Arch Dis Child 1982; 57: 882-84. 2. Costello AM de L. GOBI and infant mortality. Lancet 1988; i: 186. 3. UNICEF. State of the world’s children. Oxford: Oxford University Press, 1991 4. World Bank. Nepal social sector strategy review report no 7498-NEP New York: World Bank, 1989.
Hyperthermic liver perfusion chemotherapy in the foregut carcinoid syndrome SIR,-Foregut carcinoid tumours may lead to an atypical syndrome (facial oedema, long-lasting generalised red flushes, diarrhoea, bronchoconstriction, and hypotension that can be fatal). This syndrome may be due to hypersecretion of histamine, the serotonin (5-HT) precursor 5-hydroxytryptophan (5-HTP), kinin-related peptides, or as yet unidentified substances. Individuals with these lesions often have a deficiency of aromatic acid decarboxylase that normally converts 5-HTP to 5-HT. 5-HTP, secreted by the tumour, can be partly decarboxylated in the kidney and further oxidised to 5-hydroxyindoleacetic acid (5-HIAA). Urinary analyses of 5-HIAA and the main histamine metabolite, methylimidazole acetic acid (MelmAA), are therefore of value in establishing the identity of foregut carcinoid tumours. Patients with disseminated disease may have life-threatening carcinoid
symptoms that are difficult to treat. Carcinoid tumours have a low sensitivity to systemic treatment with cytotoxic drugs and a variable sensitivity to immunomodulatory treatment.Z.3 Patients with the foregut carcinoid syndrome can be treated with histamine receptor blockers, cortisone, and octreotide. Patients with other advanced endocrine tumours can be effectively palliated by ischaemic treatment of the metastatic liver (dearterialisation or embolisation).’ These procedures might be dangerous for the patient with a severe foregut carcinoid syndrome because of the possibility of uncontrollable release of histamine or vasoactive peptides. We now report the outcome of a patient who was treated with cytotoxic drugs delivered by regional hyperthermic liver perfusion. During perfusion the venous effluent from the liver, together with any vasoactive substances, was shunted from the systemic circulation
thereby avoiding vasomotor sequelae. A 48-year-old man underwent pulmonary lobectomy due to a bronchial carcinoid tumour. No regional or distant dissemination was found during surgery. Histological examination showed tumour cells, uniform in shape but with nuclear atypia, growing in irregular strands, trabeculae, or broad anastomosing cords. No mitotic activity was seen. The tumour was classified as a well-differentiated neuroendocrine carcinoma of the intermediate sized cells. A few cells were immunopositive for chromogranin A, but argyrophilia was found in most tumour cells. Some cells were also immunopositive for 5-HT and adrenocorticotrophic hormone (ACTH). Two years after thoracic surgery the patient had severe attacks of bronchoconstriction associated with protracted brightred facial flushing. Shortly thereafter he became severely dyspnoeic and hypotensive with facial oedema. An enlarged nodular liver was palpable. We suspected a foregut carcinoid syndrome and H, and H2 histamine-receptor antagonists, cortisone acetate, and octreotide were started. The patient’s critical state was completely reversed within two days. Screening for vasoactive intestinal polypeptide, tachykinins, proinsulin, insulin, gastrin, glucagon, human chorionic gonadotrophin, pancreatic polypeptide, and ACTH was negative, but urinary concentrations of 5-HIAA (4900 JlIl101/24 h; normal range < 70) and MeImAA (30-6 mmol/mol creatinine; normal range <2’4) were strikingly increased. Urinary cortisol concentrations were slightly increased and plasma cortisol showed no diurnal variation. Under protection with octreotide (200 ug subcutaneously four times daily) and cortisone acetate (100 mg intravenously four times daily), the patient underwent regional hyperthermic (40-41°C) perfusion of the right lobe of the liver with melphalan (0-5 mg/kg body weight) and cisplatinum (0-5 mg/kg). The left lobe had less tumour and was managed by ligation of its arterial blood supply. Cytostatic perfusion of the isolated liver, simultaneous filtration of portal vein blood, and a maintained systemic circulation were made possible with a special liver perfusion catheter (’Perfufix’, B. Braun,
569
Melsungen, Germany). This technique involves isolation of the hepatic artery, portal vein, and caval vein inferior and superior to the liver, together with a temporary (60-90 min) portocaval shunt that allows blood flow rates to be maintained in both the hepatic artery and portal lines 56 After the postoperative phase, a low dose of octreotide (100 Ilgjday subcutaneously) was maintained, and the
patient
returned
to
full-time work
at
six weeks without any
hormonally related symptoms. Both 5-HIAA and MelmAA levels were strikingly reduced (940 fimol/24 h and 7.2 mmol/mol creatinine, respectively). Thirteen months later the patient noticed slight facial flushing and diarrhoea, but computed tomography showed no obvious tumour progression. The dose of octreotide was increased and predinisone (5 mg twice daily) was added when urinary concentrations of tumour markers increased. Further episodes of dyspnoea and hypotension four months later, together with raised serum concentrations of 5-HIAA and MelmAA (2900 mol/24 h and 24-2 mmol/mol creatinine, respectively) led to an angiogram that showed occlusion of the right hepatic artery. At surgery, exploration of both the portal vein and hepatic artery were impossible because of tumour growth. Hyperthermic perfusion of the
portal vein with melphalan and cisplatinum again led to symptomatic relief and reduction of tumour markers (1200 pmol/ 24 h and 9-3 mmol/mol creatinine, respectively). Hyperthermic perfusion of the metastatic liver with cytotoxic drugs may be a suitable treatment for patients with severe foregut carcinoid syndrome. Department of Surgery, Sahlgrenska Hospital, S-413 45 Goteborg, Sweden
T. SCHERSTÉN H. AHLMAN B. WÄNGBERG
Department of Clinical Physiology, University of Linkoping
G. GRANÉRUS
Department of Pathology, University of Uppsala
L. GRIMELIUS
ED, Azzopardi JG. Tumours of the lung and the carcinoid syndrome. Thorax 1960; 15: 30-36. 2. Kvols LK. Metastatic carcinoid tumours and the carcinoid syndrome. A selective review of chemotherapy and hormonal therapy. Am J Med 1986; 81: 49-55. 3. Moertel CG, Rubin J, Kvols LK. Therapy of metastatic carcinoid tumour and the malignant carcinoid syndrome with recombinant leucocyte A interferon. J Clin 1. Williams
Oncol 1989; 7: 865-68. 4. Gerterud K, Tylén U, Jansson S, Stenqvist O, Tisell LE, Ahlman H. Hepatic arterial embolisation in the treatment of the midgut carcinoid syndrome and other advanced endocrine tumours metastatic to the liver. J Intervent Radiol 1990; 5: 69-76. 5. Aigner K, Walther H, Tonn J, et al. First experimental and clinical results of isolated liver perfusion with cytotoxics in metastasis from colorectal primary. Recent Results Cancer Res 1983; 86: 99-102. 6. Hafstrom LO, Rudenstam CM, Holmberg S, Scherstén T, Ehrsson H. The pharmacokinetics of melphalan in regional hyperthermic liver perfusion. Reg Cancer Treat 1990; 3: 23-25.
Blood pressure and exercise testing SIR,-Dr Akhras and Dr Jackson (April 13, p 899) report on the blood pressure (BP) response during an exercise test of patients in whom coronary artery disease was subsequently found by arteriography. They suggest that an increase in diastolic BP of 15 mm Hg or more during exercise is excessive and associated with coronary disease and ischaemic left-ventricular dysfunction. The sensitivity and specificity of exercise testing for the detection of coronary disease are less than perfect, and any additional information from the exercise test which helped to predict coronary disease would be of value. However, our experience suggests that similar rises in diastolic BP during exercise are common in healthy people. BP was recorded at rest and during maximum exercise in 1007 men aged 25-67 years (mean 42) attending a health screening clinic; none had chest pain, signs or symptoms of heart failure, or any electrocardiographic features of ischaemic heart disease at rest or on exercise. Diastolic BP rose by 5-10 mm Hg in 214 men, by 10-15 mm Hg in 104, and by more than 15 mm Hg in 197 (196%). Although we cannot exclude coronary disease with certainty because coronary arteriography was not done, the probability of coronary disease (given the absence of chest pain and of electrocardiographic abnormalities on an exercise test) must be low.
This suggests that in unselected patients, the specificity and positive predictive value of diastolic BP rise on exercise will be poor. Even in selected groups with a high pre-test probability of coronary disease, the specificity and positive predictive value are likely to be limited.
Royal Free Hospital, London NW3, UK
S. W. DAVIES* G. WANNAMETHEE D. P. LIPKIN
Health Care, London E2
T. M. EMERY M. I. L. WATLING
City
*Present address: Cardiac
Department, London Chest Hospital, London E2 9JX, UK
Duodenal ulcer
recurrence
and Helicobacter
pylori SIR,-Dr Fiocca and colleagues (June 29, p 1614) report a further study showing a low recurrence rate of duodenal ulcer after eradication of Helicobacter pylori. Virtually all of the hitherto published trials of this type used bismuth as a component of the antimicrobial regimens. Since bismuth not only possesses bactericidal activity but also cytoprotective properties, the extent to which cytoprotection contributes to the low relapse rates after treatment of duodenal ulcer with combinations containing bismuth remains unclear. We have randomised 104 patients with recurrent duodenal ulcer to either amoxicillin 750 mg thrice daily plus metronidazole 500 mg thrice daily or placebo for 12 days. All patients received ranitidine 300 mg nocte for six weeks or a further four weeks if healing was not achieved. After healing, patients were seen twice monthly for one year. Endoscopies were completed at time zero, and subsequently at 6 weeks, 2 months, 6 months, and 12 months after healing or whenever symptoms suggested relapse. H pylori status was checked by culture and histology on each occasion. All patients were H pylori positive before treatment. Our study is still running and the randomisation code has not yet been broken. So far 58 patients have completed 1 year and 42 patients have been observed for more than six months. 42 recurrences have occurred. H pylori was found in 41 of these, whereas only 1 patient who relapsed was H pylori negative. The relapse rate in patients who were still H pylori positive after treatment was 79% (41/52), compared with 2% (1/50) in those who cleared H pylori. These preliminary results support the view that eradication of H pylori only, and not any additional cytoprotective effects of bismuth, is responsible for the reduced relapse rate of duodenal ulcer after successful antimicrobial treatment. Hanuschkrankenhaus, A-1140, Vienna, Austria
E. HENTSCHEL H. NEMEC K. SCHÜTZE
Hygieneinstitut der Universität, Vienna
A. HIRSCHL
Ambulatorium Sud, Vienna
B. DRAGOSICS
Landeskrankenhaus Graz, Styria
G. BRANDSTÄTTER M.TAUFER
Lymphopenia in diverticulitis SIR,-Peripheral blood lymphopenia secondary to localised inflammation of the bowel has received little attention, although it has been described in gangrenous appendicitis’ and in active colitis.2 We completed a retrospective review of patients admitted to the Whittington Hospital with a primary diagnosis of diverticular disease between September, 1987, and February, 1991. Only patients with diverticula confirmed by barium enema, colonoscopy, or laparotomy were included. Of 84 diagnoses, 61 met these criteria. Admissions were classified into four groups according to the severity of the inflammatory process: perforated diverticular disease requiring emergency surgery (6 patients), acute diverticulitis treated non-operatively (28 patients), rectal bleeding attributed to diverticular disease (17 patients), and quiescent diverticular disease (10 patients). There was a significant reduction in the peripheral
LETTERS
to
Vitamin A
the
EDITOR
supplementation
SIR,-Dr West and his colleagues’ study of vitamin A supplementation in Nepal (July 13, p 67) answers the sceptics up to a point. West et al show that full coverage of a malnourished population of children with high-dose (200 000 IU) capsules given every 4 months may reduce death rates in children aged 1-4 years by up to 30%. But what are the implications of this for health planners in countries like Nepal? What is the likely impact of supplementation on total under-5 mortality rates under existing conditions of primary health care? If we assume that supplementation will not reduce deaths in infants under 6 months, which account for 80% of the infant mortality rate,l,2 the impact in Nepal, which in 1989 had an under-5 mortality of 193 per 1000 (infant mortality 125 per 10003), will be a reduction of 27-9 deaths per 1000 ([0-3 x 0.2 x 125] + [03 x 68]) or 14.5%. This is a maximum reduction and assumes full coverage. West et al conducted a rigorous but expensive aid-assisted study which probably used fieldworkers and administrative staff paid and managed outside the government system. Details of the costs were not provided and the reproducibility of this intervention was not discussed. In Nepal nearly 40% of government health expenditure is externally financed from aid; annual per caput expenditure on health is less than$2, of which less than 45 % is spent on district level primary health care.4 There is one village health worker (VHW) for every 4000 population and these workers are usually unsupported and receive inadequate daily allowances and incentives for fieldwork. The household coverage by VHWs is therefore low, and in many poor areas it is non-existent. West et al did their study in a terai (plains) district, where population density is greater than in the logistically difficult middle hill region where more than 50% of the population live. The section of the population most likely to experience vitamin A deficiency (the very poor and low-caste families) are the groups who consistently fail to receive or demand services even when they are moderately accessible. So in terms of sustainable impact on mortality a decision to implement 4-monthly vitamin A supplementation at household level within the existing government health system may achieve at best perhaps 30% of the mortality reduction West et al achievedie, a 5% reduction in under-5 deaths or 9 deaths per 1000. Health planners must decide therefore where vitamin A supplementation sits in the queue of other primary health care interventions and "magic bullets"-before or behind antibiotics for acute respiratory infections, oral rehydration sachets for dehydration, bed-nets for malaria, clean delivery kits for traditional birth attendants, road-tohealth cards, immunisation, weaning education, or promotion of breastfeeding? Certainly vitamin A supplements should be liberally supplied and promoted at health post level and given to VHWs. But international agencies should think carefully before supporting vertically run national vitamin A supplementation programmes. Money might be spent more effectively and sustainably on initiatives to improve an integrated primary health care service through better management and training, and the development of an organisational culture which makes these services more userfriendly for poor and disadvantaged families, one of whose problems is vitamin A deficiency. Centre for International Child Health, Institute of Child Health, London WC1 N 1 EH, UK
ANTHONY M. de L. COSTELLO
1. Ashworth A, Waterlow JC. Infant mortality in developing countries Arch Dis Child 1982; 57: 882-84. 2. Costello AM de L. GOBI and infant mortality. Lancet 1988; i: 186. 3. UNICEF. State of the world’s children. Oxford: Oxford University Press, 1991 4. World Bank. Nepal social sector strategy review report no 7498-NEP New York: World Bank, 1989.
Hyperthermic liver perfusion chemotherapy in the foregut carcinoid syndrome SIR,-Foregut carcinoid tumours may lead to an atypical syndrome (facial oedema, long-lasting generalised red flushes, diarrhoea, bronchoconstriction, and hypotension that can be fatal). This syndrome may be due to hypersecretion of histamine, the serotonin (5-HT) precursor 5-hydroxytryptophan (5-HTP), kinin-related peptides, or as yet unidentified substances. Individuals with these lesions often have a deficiency of aromatic acid decarboxylase that normally converts 5-HTP to 5-HT. 5-HTP, secreted by the tumour, can be partly decarboxylated in the kidney and further oxidised to 5-hydroxyindoleacetic acid (5-HIAA). Urinary analyses of 5-HIAA and the main histamine metabolite, methylimidazole acetic acid (MelmAA), are therefore of value in establishing the identity of foregut carcinoid tumours. Patients with disseminated disease may have life-threatening carcinoid
symptoms that are difficult to treat. Carcinoid tumours have a low sensitivity to systemic treatment with cytotoxic drugs and a variable sensitivity to immunomodulatory treatment.Z.3 Patients with the foregut carcinoid syndrome can be treated with histamine receptor blockers, cortisone, and octreotide. Patients with other advanced endocrine tumours can be effectively palliated by ischaemic treatment of the metastatic liver (dearterialisation or embolisation).’ These procedures might be dangerous for the patient with a severe foregut carcinoid syndrome because of the possibility of uncontrollable release of histamine or vasoactive peptides. We now report the outcome of a patient who was treated with cytotoxic drugs delivered by regional hyperthermic liver perfusion. During perfusion the venous effluent from the liver, together with any vasoactive substances, was shunted from the systemic circulation
thereby avoiding vasomotor sequelae. A 48-year-old man underwent pulmonary lobectomy due to a bronchial carcinoid tumour. No regional or distant dissemination was found during surgery. Histological examination showed tumour cells, uniform in shape but with nuclear atypia, growing in irregular strands, trabeculae, or broad anastomosing cords. No mitotic activity was seen. The tumour was classified as a well-differentiated neuroendocrine carcinoma of the intermediate sized cells. A few cells were immunopositive for chromogranin A, but argyrophilia was found in most tumour cells. Some cells were also immunopositive for 5-HT and adrenocorticotrophic hormone (ACTH). Two years after thoracic surgery the patient had severe attacks of bronchoconstriction associated with protracted brightred facial flushing. Shortly thereafter he became severely dyspnoeic and hypotensive with facial oedema. An enlarged nodular liver was palpable. We suspected a foregut carcinoid syndrome and H, and H2 histamine-receptor antagonists, cortisone acetate, and octreotide were started. The patient’s critical state was completely reversed within two days. Screening for vasoactive intestinal polypeptide, tachykinins, proinsulin, insulin, gastrin, glucagon, human chorionic gonadotrophin, pancreatic polypeptide, and ACTH was negative, but urinary concentrations of 5-HIAA (4900 JlIl101/24 h; normal range < 70) and MeImAA (30-6 mmol/mol creatinine; normal range <2’4) were strikingly increased. Urinary cortisol concentrations were slightly increased and plasma cortisol showed no diurnal variation. Under protection with octreotide (200 ug subcutaneously four times daily) and cortisone acetate (100 mg intravenously four times daily), the patient underwent regional hyperthermic (40-41°C) perfusion of the right lobe of the liver with melphalan (0-5 mg/kg body weight) and cisplatinum (0-5 mg/kg). The left lobe had less tumour and was managed by ligation of its arterial blood supply. Cytostatic perfusion of the isolated liver, simultaneous filtration of portal vein blood, and a maintained systemic circulation were made possible with a special liver perfusion catheter (’Perfufix’, B. Braun,
569
Melsungen, Germany). This technique involves isolation of the hepatic artery, portal vein, and caval vein inferior and superior to the liver, together with a temporary (60-90 min) portocaval shunt that allows blood flow rates to be maintained in both the hepatic artery and portal lines 56 After the postoperative phase, a low dose of octreotide (100 Ilgjday subcutaneously) was maintained, and the
patient
returned
to
full-time work
at
six weeks without any
hormonally related symptoms. Both 5-HIAA and MelmAA levels were strikingly reduced (940 fimol/24 h and 7.2 mmol/mol creatinine, respectively). Thirteen months later the patient noticed slight facial flushing and diarrhoea, but computed tomography showed no obvious tumour progression. The dose of octreotide was increased and predinisone (5 mg twice daily) was added when urinary concentrations of tumour markers increased. Further episodes of dyspnoea and hypotension four months later, together with raised serum concentrations of 5-HIAA and MelmAA (2900 mol/24 h and 24-2 mmol/mol creatinine, respectively) led to an angiogram that showed occlusion of the right hepatic artery. At surgery, exploration of both the portal vein and hepatic artery were impossible because of tumour growth. Hyperthermic perfusion of the
portal vein with melphalan and cisplatinum again led to symptomatic relief and reduction of tumour markers (1200 pmol/ 24 h and 9-3 mmol/mol creatinine, respectively). Hyperthermic perfusion of the metastatic liver with cytotoxic drugs may be a suitable treatment for patients with severe foregut carcinoid syndrome. Department of Surgery, Sahlgrenska Hospital, S-413 45 Goteborg, Sweden
T. SCHERSTÉN H. AHLMAN B. WÄNGBERG
Department of Clinical Physiology, University of Linkoping
G. GRANÉRUS
Department of Pathology, University of Uppsala
L. GRIMELIUS
ED, Azzopardi JG. Tumours of the lung and the carcinoid syndrome. Thorax 1960; 15: 30-36. 2. Kvols LK. Metastatic carcinoid tumours and the carcinoid syndrome. A selective review of chemotherapy and hormonal therapy. Am J Med 1986; 81: 49-55. 3. Moertel CG, Rubin J, Kvols LK. Therapy of metastatic carcinoid tumour and the malignant carcinoid syndrome with recombinant leucocyte A interferon. J Clin 1. Williams
Oncol 1989; 7: 865-68. 4. Gerterud K, Tylén U, Jansson S, Stenqvist O, Tisell LE, Ahlman H. Hepatic arterial embolisation in the treatment of the midgut carcinoid syndrome and other advanced endocrine tumours metastatic to the liver. J Intervent Radiol 1990; 5: 69-76. 5. Aigner K, Walther H, Tonn J, et al. First experimental and clinical results of isolated liver perfusion with cytotoxics in metastasis from colorectal primary. Recent Results Cancer Res 1983; 86: 99-102. 6. Hafstrom LO, Rudenstam CM, Holmberg S, Scherstén T, Ehrsson H. The pharmacokinetics of melphalan in regional hyperthermic liver perfusion. Reg Cancer Treat 1990; 3: 23-25.
Blood pressure and exercise testing SIR,-Dr Akhras and Dr Jackson (April 13, p 899) report on the blood pressure (BP) response during an exercise test of patients in whom coronary artery disease was subsequently found by arteriography. They suggest that an increase in diastolic BP of 15 mm Hg or more during exercise is excessive and associated with coronary disease and ischaemic left-ventricular dysfunction. The sensitivity and specificity of exercise testing for the detection of coronary disease are less than perfect, and any additional information from the exercise test which helped to predict coronary disease would be of value. However, our experience suggests that similar rises in diastolic BP during exercise are common in healthy people. BP was recorded at rest and during maximum exercise in 1007 men aged 25-67 years (mean 42) attending a health screening clinic; none had chest pain, signs or symptoms of heart failure, or any electrocardiographic features of ischaemic heart disease at rest or on exercise. Diastolic BP rose by 5-10 mm Hg in 214 men, by 10-15 mm Hg in 104, and by more than 15 mm Hg in 197 (196%). Although we cannot exclude coronary disease with certainty because coronary arteriography was not done, the probability of coronary disease (given the absence of chest pain and of electrocardiographic abnormalities on an exercise test) must be low.
This suggests that in unselected patients, the specificity and positive predictive value of diastolic BP rise on exercise will be poor. Even in selected groups with a high pre-test probability of coronary disease, the specificity and positive predictive value are likely to be limited.
Royal Free Hospital, London NW3, UK
S. W. DAVIES* G. WANNAMETHEE D. P. LIPKIN
Health Care, London E2
T. M. EMERY M. I. L. WATLING
City
*Present address: Cardiac
Department, London Chest Hospital, London E2 9JX, UK
Duodenal ulcer
recurrence
and Helicobacter
pylori SIR,-Dr Fiocca and colleagues (June 29, p 1614) report a further study showing a low recurrence rate of duodenal ulcer after eradication of Helicobacter pylori. Virtually all of the hitherto published trials of this type used bismuth as a component of the antimicrobial regimens. Since bismuth not only possesses bactericidal activity but also cytoprotective properties, the extent to which cytoprotection contributes to the low relapse rates after treatment of duodenal ulcer with combinations containing bismuth remains unclear. We have randomised 104 patients with recurrent duodenal ulcer to either amoxicillin 750 mg thrice daily plus metronidazole 500 mg thrice daily or placebo for 12 days. All patients received ranitidine 300 mg nocte for six weeks or a further four weeks if healing was not achieved. After healing, patients were seen twice monthly for one year. Endoscopies were completed at time zero, and subsequently at 6 weeks, 2 months, 6 months, and 12 months after healing or whenever symptoms suggested relapse. H pylori status was checked by culture and histology on each occasion. All patients were H pylori positive before treatment. Our study is still running and the randomisation code has not yet been broken. So far 58 patients have completed 1 year and 42 patients have been observed for more than six months. 42 recurrences have occurred. H pylori was found in 41 of these, whereas only 1 patient who relapsed was H pylori negative. The relapse rate in patients who were still H pylori positive after treatment was 79% (41/52), compared with 2% (1/50) in those who cleared H pylori. These preliminary results support the view that eradication of H pylori only, and not any additional cytoprotective effects of bismuth, is responsible for the reduced relapse rate of duodenal ulcer after successful antimicrobial treatment. Hanuschkrankenhaus, A-1140, Vienna, Austria
E. HENTSCHEL H. NEMEC K. SCHÜTZE
Hygieneinstitut der Universität, Vienna
A. HIRSCHL
Ambulatorium Sud, Vienna
B. DRAGOSICS
Landeskrankenhaus Graz, Styria
G. BRANDSTÄTTER M.TAUFER
Lymphopenia in diverticulitis SIR,-Peripheral blood lymphopenia secondary to localised inflammation of the bowel has received little attention, although it has been described in gangrenous appendicitis’ and in active colitis.2 We completed a retrospective review of patients admitted to the Whittington Hospital with a primary diagnosis of diverticular disease between September, 1987, and February, 1991. Only patients with diverticula confirmed by barium enema, colonoscopy, or laparotomy were included. Of 84 diagnoses, 61 met these criteria. Admissions were classified into four groups according to the severity of the inflammatory process: perforated diverticular disease requiring emergency surgery (6 patients), acute diverticulitis treated non-operatively (28 patients), rectal bleeding attributed to diverticular disease (17 patients), and quiescent diverticular disease (10 patients). There was a significant reduction in the peripheral