HIV-associated diarrhoea and wasting

HIV-associated diarrhoea and H L DuPont, G D Marshall Report of a meeting at the

University

wasting

of Texas Medical School at Houston

Chronic diarrhoea and weight loss commonly complicate the course of HIV infection and AIDS. The broad range of potential aetiological agents in HIV-infected patients with diarrhoea, the cost of the various tests required to detect these agents, and the variable response to treatment in these patients make the approach to HIVassociated diarrhoea complex and confusing. Weight loss and cachexia may occur with diarrhoea or in association with other HIV-associated infections and disorders. Both acute and chronic diarrhoea are seen commonly in HIVinfected persons living in Europe and the United States. Wasting commonly occurs with or without diarrhoea in advanced HIV disease. A syndrome of chronic diarrhoea (more than 30 days) and important weight loss (reduction of 10% or more of normal weight) is one of the major manifestations of AIDS in Africa; the syndrome is referred to as slim disease. The patient presented here illustrates how difficult it is to establish the cause of diarrhoea in an HIV-infected patient and to reach a satisfactory management plan.

Case-presentation ihe

patient, a 34-year-old Hispanicbisexual male, was found to be HIV positive in May, 1989, during routine clinical examination. In April, 1993, he began to have bothersome diarrhoea without fever. At that time, his medications included zidovudine 500 mg per day, dapsone 100 mg twice weekly, and acyclovir 200 mg three times daily. A stool sample was positive for Giardia lamblia (3+ cysts and 1+ trophozoites) and negative for bacterial enteropathogens and leucocytes. The patient was treated with oral metronidazole 500 mg three times daily for 10 days without improvement. Mucous diarrhoea persisted and direct microscopy of a stool specimen collected in October, 1993, revealed many organisms from the order Microsporidia but no G lamblia. He was treated for a second time with oral metronidazole 500 mg three times daily for 10 days, followed by a course of trimethoprim/sulphamethoxazole (co-trimoxazole) 160 mg/800 mg twice daily for 10 days, again without improvement. His CD4 T-lymphocyte count at this time was 66/(JbL, his CD8 T-lymphocyte count was 528/)JbL (ratio 0- 13). He had lost 9 kg. In January, 1994, oesophageal candidosis was diagnosed endoscopically and he was started on ketoconazole 200 mg daily; and because of his refractory diarrhoea he was prescribed subcutaneous octreotide 100 )Jbg three times daily,

Department of Internal Medicine, University of Texas Medical School at Houston, Houston, Texas, USA (Prof H L DuPont MD, G D Marshall MD) Correspondence to: Dr H L Houston, TX, 77030, USA 352

DuPont, 6720 Bertner Ave, MC1-164,

2 mg after each unformed stool and not to exceed 16 mg/day, and dideoxycytosine 0-75 mg three times daily. There was a slight and temporary improvement in his symptoms. In May, 1994, he was admitted to Hermann Hospital for evaluation of diarrhoea and weight loss. At that time he was passing two to three voluminous unformed stools per day, usually after eating. He had no appetite. On examination he was wasted and had epigastric tenderness. The stool contained occult blood but again no leucocytes were seen; stool cultures for bacterial pathogens and examinations for parasites were negative. Ultrasound examination of the abdomen revealed small (<5 mm) ectogenic nodular foci in the gallbladder wall consistent with small polyps. Endoscopies revealed slight erythema of the gastric mucosa and focal inflammation of the rectosigmoid colon. Duodenal biopsy showed non-specific inflammation; and on rectal biopsy there was moderate with inclusions consistent with inflammation,

loperamide

cytomegalovirus. Endoscopic retrograde cholangiopancreatography suggested papillary stenosis and a 9 mm biliary papillotomy was done, at which time a biopsy specimen was taken from the papillary valve; this contained glandular epithelium, with nuclear and cytoplasmic inclusions characteristic of cytomegalovirus. In bileduct washings, yeast forms and pseudohyphae were identified. The patient was started on intravenous ganciclovir 225 mg every 12 h. Causes of

persistent diarrhoea

There are two leading explanations for this complication of HIV/AIDS-enteric infection by one or more species of pathogenic microbe (opportunistic enteric infection), and gut architectural alteration secondary to local HIV infection (HIV enteropathy).

Opportunistic enteric infection An infectious agent

can

be identified in about 50% of

patients with AIDS-associated chronic diarrhoea. The principal enteropathogens in AIDS-associated diarrhoea listed in table 1. Parasites seem to be the most commonly identified causal organisms. In one study, parasites were identified in 19 of 27 (70%) patients with chronic diarrhoea and malnutrition.’ Those with most are

potential as enteropathogens are Cryptosporidium parvum, Isospora belli, members of the order Microsporidia, and Cyclospora. The patient under discussion had G lamblia infection. Although certain immunodeficiency states to infection is it Giardia, predispose by hypogammaglobulinaemia, not AIDS, that seems to increase the risk for giardiasis. If HIV-positive patients are at high risk of Giardia infection, the excess is probably due to the increased risk of infection in homosexual men, which is unrelated to the retroviral infection. This may also be true for Entamoeba histolytica. In contrast,

Table 1: Causes of chronic diarrhoea in

patients with AIDS

Cryptosporidium parvum, which produces a chronic and debilitating illness in patients with AIDS,2 is HIV related: their susceptibility is probably due to cellular immune defects. The

patient under discussion was infected with a member of the order Microsporidia, organisms commonly found in AIDS patients with diarrhoea. Enterocytozoon bieneusi is the most commonly identified microsporidian species found in human beings. The spores can be found in infected patients by light microscopy of faecal samples, small bowel fluid, or mucosal sections. AIDS patients infected with a member of the order Microsporidia generally have lower CD4 counts, Karnofsky scores, and survival prospects than those not infected. Certain antimicrobial agents favourably alter the illness associated with Microsporidia in patients with AIDS; this has been put forth as evidence that the organism is responsible for the diarrhoea. Our patient received metronidazole without effect, but this agent has occasionally been useful in relieving the illness associated with microsoporidiosis. Albendazole, unlicenced for this indication, may be the most effective drug so far evaluated.3 Members of the order Microsporidia have been identified in several studies of HIV-associated diarrhoea, but the existence of intestinal infection by these parasites is not sufficient to establish cause and effect. Despite their favourable clinical effect, antimicrobial drugs do not eliminate parasites of the order Microsporidia and absorptive function of the gut does not always improve.3 The beneficial effects may be due to reduction of enteric bacteria in the small bowel. In a case-control study of AIDS-associated diarrhoea, Rabeneck et al4 found no difference in the infection rate by a parasite of the order Microsporidia in patients with and without diarrhoea. Cyclospora may likewise infect the gut of patients with HIV infection and AIDS; it seems especially common in certain geographic regions such as Nepal and Haiti. The conventional enteric bacterial pathogens such as Salmonella enteritidis, Shigella spp, or Campylobacter jejuni may cause protracted and refractory illness. S enteritidis serotypes can produce bacteraemia without focal infection, as seen in other conditions associated with defects of cellular immunity. Mycobacterium avium complex may also produce enteric infection and chronic diarrhoea; likewise coliform bacteria adherent to intestinal

When CMV produces enteric diseases in conjunction with HIV infection, it characteristically involves the colon. Endoscopy reveals colitis, an ulcer, haemorrhage, a necrotising process, or toxic dilation. By contrast, in patients with non-AIDS-predisposing conditions such as after organ therapeutic immunosuppression transplantation, CMV tends to affect the upper gastrointestinal tract. Many HIV-infected patients with CMV intestinal infection will benefit from antiviral therapy.8 The currently recommended treatment is ganciclovir 10-15 mg/kg per day in two or three divided doses each day for 2-3 weeks. There is a high relapse rate, at least partly explained by the development of resistance to the ganciclovir. Relapsing patients may respond to a course of foscarnet. The patient discussed probably had CMV colitis and CMV cholangitis.

HIV

enteropathy

patients with AIDS the inability to correlate gastrointestinal symptoms and malabsorption with known gastrointestinal pathogens suggests the presence of HIV enteropathy. The mechanisms are varied and complex (table 2). The intestinal lamina propria contains CD4 T cells, which are susceptible to infection by HIV. CD4 T cells in the intestine decline just as they do in the peripheral circulation during the course of HIV infection. Indirect evidence that HIV actively infects the intestinal tract and is responsible for a measure of the illness comes from observations that HIV-infected patients with persistent diarrhoea shed HIV nucleic acids in stool9 and frequently show a faecal sIgA response to HIV p24 antigen, with the magnitude of the response relating to

In many

duration of the associated diarrhoea. 10 Alterations in host immunity due to HIV infection contribute substantially to the pathophysiology of AIDS with enteropathy. Although immunodeficiency subsequent opportunistic infections account for many of the clinical findings, an important part is played by the dysfunctional immunity, particularly as it relates to the mucosal immune system (MIS). The MIS consists of focal aggregates of cells and tissue from the epithelium to the lamina propria and includes T cells, B cells, mast cells, and macrophages. Specialised epithelial cells (M cells) function to bind intraluminal antigens and transport them to the enteric lymphoid tissue for processing. HIV infection leads to MIS dysfunction. HIV has been detected in mononuclear leucocytes from intestinal lamina propria of AIDS patients."Thus HIV may enter the MIS via M cells, which transport the virus to the lamina propria and present it to T and B cells with subsequent virus-specific IgA production. With

mucosa.5

Cytomegalovirus (CMV) can infect any level of the gastrointestinal tract. 8% of HIV-infected patients and 13% of patients with AIDS were found to have CMV enteric infection. It causes severe, potentially lifethreatening, enteric disease in 2% of patients with AIDS.7 6

When CMV is found in intestinal tissue, it is not invariably causing the disease; other treatable conditions may be responsible. In an AIDS patient with diarrhoea CMV should be considered an enteropathogen only if there is associated inflammation, vasculitis, or both.

Table 2: Possible immune and neuroendocrine mediated events in AIDS-associated enteropathy

353

HIV causes dysfunction and/or correlate with differences in total faecal IgE in the various T well be destruction of CD4 cells that may a central patient groups. feature responsible for the subsequent enteropathy. A Antigen-initiated events are unlikely to be the whole decline in functional CD4 T cells decreases help to B explanation for diarrhoea of this sort. There exists an cells, manifested in the gut as fewer IgA-secreting plasma important relationship between the neuroendocrine and cells. This would result in decreased secretory IgA 25 the immune systems in the gut. Neuroendocrine mediators

progressive infection,

intestinal immunoglobulin. As AIDS progresses, CDS T cells, responsible for protection against intracellular pathogens, also become dysfunctional-probably as a result of decreased help from the increasingly depeleted CD4 T cells. Thus, the individual becomes increasingly susceptible to intracellular pathogens such as cytomegalovirus. Systemic dysfunction of the immune system leads to additional persistent infections such as those from mycobacteria or protozoa. As B-cell dysfunction becomes generalised, declining antibody responses to various infectious agents (and to administered vaccines) lead to persistent infections. The waning of cell-mediated immunity with progressive HIV-associated disease also contributes to this infectious disease dilemma. In several enteric infections, specific antibody alone is not protective12 but most work in concert with T-cell-mediated mechanisms. 13 Thus, a defect in either or both mechanisms in a particular patient be reason for persistent enteric disease and can subsequent diarrhoea. Finally, depletion of T cells by HIV in the intestinal lamina propria per se may explain the villous atrophy and crypt hyperplasia commonly seen in AIDS enteropathy." Since HIV-infected patients are at increased risk of immune hypersensitivity disorders, might such mechanisms be relevant to AIDS enteropathy? In experimental trichinosis, a defined hypersensitivity defence mechanism has been described.14 Parasite antigen-specific IgE is induced which, when bound to intestinal mast cells and subsequently cross-linked by antigen, leads to a secretory diarrhoea responsible for expelling the worms and thus protecting the host.15 Further, in animal models, non-parasitic antigens such as ovalbumin and (3-lactoglobulin induce specific IgEmediated secretory diarrhoea.16 These activated mast cells secrete products such as histamine, serotonin,

principal protective

prostaglandin

E2,

various

leukotrienes,

bradykinin,

substance P, and vasoactive intestinal peptide. Other intestinal cells can also make many of these mediators. Finally, an "innocent bystander" effect of excessive mast cell mediators, resulting in enteropathy, has been proposed. Thus, it follows that antigen-specific IgE bound to intestinal mast cells could, when cross-linked by antigen, initiate the cascade to promote acute immunemediated diarrhoea. If the antigen was persistent, chronic secretory diarrhoea could also result. 14 HIV-infected patients are reported to have higher total serum IgE than age and sex matched controls. 17 Additionally, the specificity of IgE in HIV-infected patients is directed not only at classic allergens but often at parasites, bacteria, fungi, and viruses, including HIV’8 We were interested to determine whether chronic diarrhoea could occur with chronic antigen stimulation. Accordingly, we studied stool samples from 42 Zambian AIDS patients for the presence of HIV p24-specific IgE.19 62% of patients with chronic diarrhoea had p24-specific faecal IgE vs only 12% with acute diarrhoea and none of those without diarrhoea at the time of sampling. The finding, which deserves study in other populations, was independent of definable enteric pathogens and did not 354

influence

the

luminal

epithelial

absorptive-secretory

mechanisms, priming and/or directly activating mast cells. In animal models, psychological conditioning has positive effects on immune-mediated gut activity. Additionally, neuroendocrine mediator receptors have been reported on the cell surface of immunocompetent cells. Taken together, these findings indicate how stress might be a factor in diarrhoea. Clinically, such relationships are further suggested by knowledge that stress increases basal mast cell activity.2O As in the irritable bowel syndrome,2’ a condition characterised by intermittent bouts of abdominal pain and diarrhoea, psychological counselling and stress management might be beneficial.

AIDS-related

biliary tract

disease

Patients with biliary tract disease complicating HIV infection characteristically present with low-grade fever and abdominal pain. Acalculous cholecystitis, sclerosing cholangitis, dilation of bile ducts, and papillary stenosis are recognised complications of HIV infection, and characteristically these complications are caused by Cryptosporidium parvum, cytomegalovirus, or a member of the order Microsporidia.22,23 The diagnosis is made by examining bile, gallbladder, ductal mucosa, or ampullary tissue. In the patient under discussion, probable papillary stenosis was diagnosed and biopsy of the ampulla revealed nuclear cytoplasmic viral inclusions consistent with CMV infection.

Management Work-up and evaluation

of AIDS-related diarrhoea should cost-benefit considerations.24 Antimicrobial drugs can be useful in pathogen-specific diarrhoea associated with HIV infection, but in many patients a cause is never established and in others several possible causes may be identified. The evaluation should be staged according to findings of a preliminary work-up and

be based

on

specific or symptomatic treatment (figure). In the patient with chronic diarrhoea one starts with noninvasive tests for enteric pathogens. Two freshly passed stool samples are examined for (1) routine bacteria, including Shigella spp, Salmonealla spp, Campylobacter Aeromonas Plesiomonas shigelloides; (2) spp, spp, Clostridium difficile cytotoxin; and (3) parasitic agents, including Cryptosporidium parvum, members of the order Microsporidia, members of the class Cyclosporeae, Isospora belli, Giardia lamblia, and Entamoeba histolytica. A blood culture should be obtained for bacteria and mycobacteria. If an agent is identified, then specific treatment is given. If no agent is identified, a 10-day course of oral fluoroquinolone therapy may be given to adult patients to ensure that treatable bacterial enteropathogens have been excluded as the cause of diarrhoea. Co-trimoxazole is not used in AIDS patients with enteric infection, other than for isospora25 or cyclospora infection,26 because of the risk of cutaneous reactions and because of the importance of the drug in prevention and therapy of Pneumocystis carinii pneumonia. Two difficulties with antimicrobial therapy are recurrence of symptoms and infection after treatment and

response to

preparations. Specific immunomodulatory drug regimens of this sort, and other treatments such as psychological counselling and stress management, may deserve consideration in AIDS patients with refractory diarrhoea. However, formal recommendations must await the results of controlled trials. If symptomatic treatment is not effective in controlling the diarrhoea, then endoscopy is indicated. The patient should initially undergo flexible sigmoidoscopic examination with biopsy of normal lesions. Biopsy material should be examined for cytomegalovirus M avium complex, adenovirus, fungi, and herpes simplex virus. The patient should also undergo gastroduodenoscopy and tissue biopsy. Biopsy material should be examined for cytomegalovirus, mycobacteria, fungi, and parasites. Again, treatable forms of enteric infection should have specific therapy.

Conclusion The patient described in this grand round illustrates many of the clinical dilemmas of HIV-associated diarrhoea and wasting. Firstly, the illness characteristically is protracted, lasting more than a year in this case. Secondly, the cause of the diarrhoea may be obscure and more often than not will involve a complex interaction between infection by well-defined enteric pathogens and intestinal immune defects. Because of the variation in cause, multiple evaluations may be necessary in attempting to improve the quality of life; and treatment will often require not only specific antimicrobial agents but also non-specific drugs for symptomatic relief. References Kotler D, Francisco A, Clayton F, Scholes JV, Orenstein JM. Small intestinal injury and parasitic diseases in AIDS. Ann Intern Med 1990; 113: 444-49. 2 Soave R, Johnson W Jr. Cryptosporidium and Isospora belli infections. J Infect Dis 1988; 157: 225-29. 3 Dieterich D, Lew E, Kotler D, Poles M, Orenstein J. Treatment with albendazole for intestinal disease due to Enterocytozoon bieneusi in patients with AIDS. J Infect Dis 1994; 169: 178-83. 4 Rabeneck L, Gyorkey F, Genta R, Gyorkey P, Foote L, Risser J. The role of Microsporidia in the pathogenesis of HIV-related chronic diarrhea. Ann Intern Med 1993; 119: 895-99. 5 Mathewson JJ, Jiang ZD, Zumla A, et al. HEp-2 cell-adherent Escherichia coli in patients with human immunodeficiency virusassociated diarrhea. J Infect Dis 1995; 171: 1636-39. 6 Frances N, Boylston A, Roberts A, Parkin J, Pinching A. Cytomegalovirus infection in gastrointestinal tracts of patients infected with HIV-1 or AIDS. J Clin Pathol 1989; 42: 1055-64. 7 Jacobson M, Mills J. Serious cytomegalovirus disease in acquired immunodeficiency syndrome (AIDS). Clinical findings, diagnosis, and treatment. Ann Intern Med 1988; 108: 585-94. 8 Chachoua A, Dieterich D, Krasinski K, et al. 9-(1,2-dihydroxy-2propoxymethyl) guanine (ganciclovir) in the treatment of cytomegalovirus gastrointestinal disease with the acquired immunodeficiency syndrome. Ann Intern Med 1987; 107: 133-37. 9 Yolken RH, Shuojia L, Perman J, Viscidi R. Persistent diarrhea and fecal shedding of retroviral nucleic acids in children infected with human immunodeficiency virus. J Infect Dis 1991; 164: 61-66. 10 Mathewson J, Jiang Z, DuPont H, Chintu C, Luo N, Zumla A. Intestinal secretory IgA immune response against human immunodeficiency virus among infected patients with acute and chronic diarrhea. J Infect Dis 1994; 169: 614-17. 11 Ullrich R, Zeitz M, Heise W, L’age M, Hoffken G, Riecken E. Small intestinal structure and function in patients infected with human immunodeficiency virus (HIV): evidence for HIV-induced enteropathy. Ann Intern Med 1989; 111: 15-21. 12 Quinn TC, Piot P, McCormick JB, et al. Serologic and immunologic studies in patients with AIDS in North America and Africa. JAMA 1987; 257: 2617-21. 13 Morgan DR, DuPont HL, Gonik B, Kohl S. Cytotoxicity of human peripheral blood and colostral leukocytes against Shigella species. Infect Immun 1984; 46: 25-33. 1

Figure: Stepwise approach to investigation of HIV/AIDS-associated chronic diarrhoea

development of drug resistance among the enteric pathogens, especially during chronic suppressive treatment. Resistant Shigella spp, Campylobacter spp, cytomegalovirus, and herpes simplex virus have been particularly important. If no treatable condition is identified, fever is absent, and stools do not show gross evidence of blood, the patient is treated symptomatically, usually with loperamide (4 mg initially followed by 2 mg after each unformed stool passed to a maximum of 16 mg daily) or diphenoxylate hydrochloride (4 mg four times a day as needed). Other useful palliative drugs are tincture of opium and attapulgite. Bismuth subslicylate should be avoided, in view of the hazard of bismuth intoxication from excessive doses. Octreotide has been used successfully in some cases of microsporidiosis and may have value in chronic diarrhoea refractory to other symptomatic treatments. When all other therapeutic measures have failed, experimental approaches may offer further options. In a possible model of hypersensitivity diarrhoea, Zhang et a}27 reported that pretreatment of parasite-immune animals with an antihistamine (diphenhydramine) and a prostaglandin inhibitor (indomethacin) completely prevented the diarrhoea induced by exposure to the gut lumen to the specific parasite, both in situ and in isolated intestinal

355

14 Castro GA, Powell DW. The physiology of mucosal immune system and immune-mediated responses in the gastrointestinal tract. In: Johnson LR, ed. Physiology of the gastrointestinal tract. 3rd ed. New York: Raven, 1994: 707-50. 15 Zhang S, Castro GA. Involvement of type I hypersensitivity in rapid rejection of Trichinella spiralis from adult rats. Int Arch Allergy Appl Immunol 1990; 93: 272-79. 16 Baird AW, Cuthbert AW, Pierce FL. Immediate hypersensitivity reactions in epithelia from rats infected with Nippostrongylus brasiliensis. Br J Pharmacol 1985; 85: 787-95. 17 Wright DN, Nelson RP, Ledford DK, Fernandez-Calsdas E, Trudeau WL, Lockey RF. Serum IgE and human immunodeficiency virus (HIV) infection. J Allergy Clin Immunol 1990; 85: 445-52. 18 Pedersen M, Nielsen CM, Permin H. HIV antigen stimulates basophil leucocytes from AIDS patients to release histamine due to type I allergy. Agents Actions 1989; 27: 55-57. 19 Marshall GD, Gibbons AC, Gould S, et al. HIV-specific IgE in patients with diarrhea. Clin Res 1994; 42: 279A (abstr). 20 Djuric VJ, Djordjevic I, Lazarevic M, Marcovic BM, Janovic BD. Stress and anaphylactic shock. Int J Neurosci 1990; 51: 231-32.

21 Read NW. Irritable bowel

syndrome-definition and pathophysiology. Scand J Gastroentrol 1987; 130 (suppl): 7-13. 22 Cello J. Acquired immunodeficiency syndrome cholangiopathy: spectrum of disease. Am J Med 1989; 86: 539-46. 23 Pol S, Romana C, Richard S, et al. Microsporidia infection in patients with the human immunodeficiency virus and unexplained cholangitis. N Engl J Med 1993; 328: 95-99. 24 Johanson JF, Sonnenberg A. Efficient management of diarrhea in the acquired immunodefciency syndrome (AIDS): a medical decision 25

26

27

analysis. Ann Intern Med 1990; 112: 942-48. Pape JW, Verdier R-I, Johnson WD Jr. Treatment and prophylaxis of

Isospora belli infection in patients with the acquired immunodeficiency syndrome. N Engl J Med 1989; 320: 1044-47. Hoge CW, Shlim DR, Ghimire M, et al. Placebo-controlled trial of co-trimoxazole for cyclospora infections among travellers and foreign residents in Nepal. Lancet 1995; 345: 691-93. Zhang S, Myers S, Castro GA. Inhibition of anaphylaxis-evoked intestinal fluid secrection by the dual application of an H1 antagonist and cyclooxygenase inhibitor. Gastroenterology 1991; 100: 922-28.

Attitudes of

Japanese and Japanese-American physicians towards life-sustaining treatment

Summary

Introduction

Doctors in different countries have different approaches to bioethical problems. We studied how attitudes to lifesustaining treatment for terminally ill patients differ in Japan and the USA by administering a questionnaire to Japanese (136) and Japanese-American (77) physicians. In a series of clinical scenarios the questionnaire asked what life-sustaining interventions the doctors would recommend to a patient with metastatic gastric cancer. Most Japanese physicians would recommend blood transfusions for gastrointestinal bleeding (74%), total nutrition for malnutrition and parenteral (67%), vasopressors for life-threatening hypotension (61%) when the patient did not know of his diagnosis and outlook. Significantly fewer Japanese physicians would want these interventions for themselves: 29% would want transfusion, 36% would want total parenteral nutrition, and 25% would want vasopressors. 36% of Japanese physicians would override the explicit request of a competent moribund cancer patient to withdraw all life-support. By contrast, among Japanese-American physicians only 42% would recommend blood transfusions, 33% total parenteral nutrition, and 34% vasopressors to a terminally ill cancer patient who did not know of his diagnosis or outlook. Cross-cultural studies in medical ethics can help physicians and the public in different countries to take a fresh look at accepted practices and the ethical reasons behind them.

People

Program in Medical Ethics, Department of General Internal Medicine, School of Medicine, University of California, 521 Parnassus Avenue, Suite C-126, Box 0903, San Francisco, CA 94143-0903, USA (A Asai MD, Prof B Lo MD); and Division of Internal Health, Faculty of Medicine, University of Tokyo, Japan (S Fukuhara MD) Correspondence 356

to: Prof Bernard Lo

in different countries respond to bioethical dilemmas in different ways. Although disclosure of the diagnosis of cancer and informed consent are standard practice in the USA and northern European countries, they are not common in Japan and southern Europe.1-3 Different peoples may also have different attitudes towards life-sustaining interventions. In the USA, patient self-determination is prized, and terminally ill patients often choose to forego life-sustaining interventions. Little is known about the life-sustaining interventions that physicians recommend in countries that do not place such high regard on self-determination, informed therefore and disclosure. We consent, surveyed to determine which interventions Japanese physicians they would recommend to terminally ill adult patients with cancer and which they would want for themselves in this situation. We also administered the same survey to Japanese-American physicians in the USA to ascertain what recommendations would be made by physicians with a similar heritage but in a different cultural setting.

Subjects and methods Subjects July and November, 1994, we sent an eight-page questionnaire to 192 Japanese physicians in Japan and 182 Japanese-American physicians in the USA. The questionnaires used in Japan were in Japanese. In Japan, physicians came from all specialties caring for terminally ill adult cancer patients at six university hospitals and three community hospitals in six areas in Japan. In the USA, physicians were drawn from internal medicine, family practice, and general practice because such doctors are likely to care for terminally ill adult cancer patients. American subjects were all physicians with a Japanese surname who were listed in the Physicians’ Directory published by the American Medical Association, and who lived in the 12 cities with the largest Japanese-American population. The study protocol was approved by the committee on human research at the University of California San Francisco. Between