Digestive system diseases

To the Editor: We report three cases of locally acquired Entamoeba histolytica infection in men who have sex with men (MSM) in Sydney, New South Wales. E. histolytica is an invasive pathogenic amoeba that can cause invasive intestinal and extraintestinal amoebiasis. Entamoeba dispar is morphologically identical but is considered non-pathogenic and non-invasive.1 The three patients presented with a 1–3-week history of diarrhoea and abdominal pain. Routine bacterial cultures were negative for pathogens. Ova, cyst and parasite investigations showed cysts and trophozoites of E. histolytica/dispar complex in permanently stained, fixed faecal smears. Stool samples were tested for E. histolytica and E. dispar by polymerase chain reaction (PCR), using a previously described method.2 All three patients were positive for E. histolytica by PCR; sequencing of the amplicons verified the presence of E. histolytica DNA. The three patients presented within a 12-month period in 2005–2006. All were homosexually active men (ages, 31–53 years) who lived in inner Sydney. None had a history of overseas travel within the previous 5 years, suggesting that the infections were locally acquired. High rates of intestinal parasitism are found in MSM throughout the world. Oral–anal and oral–genital sexual practices are reported to predispose to infection with enteric pathogens, particularly protozoa. A 1991 study reported a higher prevalence (37%) of E. histolytica/dispar complex in a homosexual population in Sydney when compared to non-MSM.3 However, that study did not differentiate between the two species E. histolytica and E. dispar. Amoebiasis has become endemic in MSM in Japan and causes significant morbidity and mortality; complications such as colitis and liver abscesses occur more frequently in homosexual and bisexual men than in heterosexual men.4 Similar findings on amoebiasis are reported from Taiwan, with MSM at increased risk for invasive amoebiasis and intestinal colonisation with E. histolytica.5 The discovery of E. histolytica infection in MSM in Australia is of public health concern and highlights the importance of continued surveillance, as the organism has the potential to become endemic in the gay population and to cause significant morbidity. Clinicians should also be aware that E. histolytica is present in urban settings in Australia and should be included in differential diagnoses.

Damien J Stark · Rashmi Fotedar · John T Ellis · John L Harkness

Eosinophilic oesophagitis (EO), an isolated eosinophilic inflammation of the oesophagus, is the most common of the eosinophilic gastrointestinal disorders. EO is of increasing clinical significance in many developed countries,1 in parallel with the recent increase in food allergic disorders. A retrospective study of Western Australian children reported a dramatic rise in prevalence of EO (from 0.05 cases per 10 000 children in 1995 to 0.89 cases per 10 000 children in 2004) (Level III-2).2 The estimated prevalence in the US population aged 0–19 years is even higher, at 4.3 cases per 10 000 individuals.3 Clinical features. Patients present with symptoms indistinguishable from those of gastro-oesophageal reflux (GOR); however, unlike people with GOR, they are generally unresponsive to treatment with proton pump inhibitors (Level III-2). Infants often have additional clinical features, including feeding difficulties, feeding refusal and/or poor weight gain. In older children and adults, oesophageal food impaction is the most characteristic symptom (Level III-3), and should alert clinicians to include EO in the differential diagnosis. Typically, patients with EO have associated atopic disorders, including asthma and eczema. Mucosal appearance in EO Cause. The cause of EO is not clear, but the condition is closely associated with atopic disorders, and there is evidence that both IgE- and non-IgE-mediated food allergy are involved in its aetiology. Foods commonly implicated include cows milk, soy, wheat and egg.3 Data from animal experiments and anecdotal clinical observations suggest that inhalant allergens may also contribute. In keeping with murine models, recent gene array studies in people with EO have shown increased gene expression for eotaxin-3, a chemokine that promotes the migration of eosinophils into the oesophagus.1 Diagnosis. The diagnostic hallmark of EO is a dense, eosinophilic infiltrate involving the entire oesophageal mucosa, which is normally free of eosinophils. Key diagnostic criteria are basal layer hyperplasia and the presence of more than 20 eosinophils per high power field (× 400 magnification) in gastroscopic biopsies of the lower and upper oesophagus (Level III-3). Although oesophageal eosinophils are also seen in patients with reflux oesophagitis, mucosal eosinophil counts in such patients are lower (< 5 per high power field), and the eosinophils are limited to the lower oesophagus. There is a typical mucosal appearance in many patients with EO (thickened mucosa, with longitudinal furrowing and superficial white plaques [Box]) — although in about a third of patients the mucosa will look macroscopically normal. Skin prick testing (SPT) and atopy patch testing (APT) (the application of food or food extracts to the skin for 48 hours) are thought to be helpful in identifying potential causative food allergens,3 but prospective studies are needed to evaluate their predictive value. Management. To date, no completed randomised controlled trials evaluating the benefits and adverse effects of medical treatments for EO are available. Infants and young children may respond to dietary allergen restriction or an elemental diet. A diagnostic trial of an amino acid-based formula (AAF) for 6–8 weeks may be useful in determining whether a patient is diet-responsive. The treatment response needs to be assessed by repeat endoscopy, as not all patients will improve. If remission is achieved, food items can be gradually reintroduced into the diet, as tolerated, taking SPT and APT findings into consideration.3 In older children and adults, an elemental diet is often impractical and poorly tolerated due to the taste of AAF. These patients can be treated with swallowed topical corticosteroid aerosols (fluticasone) or, in refractory cases, with systemic corticosteroids.1 Other drugs, including cromoglycate and montelukast, have been used in uncontrolled trials, but their efficacy is poorly documented. Novel monoclonal antibodies against interleukin-5 may also be of benefit in treating EO, but prospective studies are not yet available. Evidence-based practice tip Infants and young children with eosinophilic oesophagitis often respond to an elemental diet, while older children and adults have been successfully treated with swallowed topical corticosteroid aerosols (fluticasone) (Level III-2).* * NHMRC levels of evidence. Prognosis. The long-term prognosis for EO is largely unknown. In some infants and young children with food protein-induced EO, the disease may remit due to development of oral tolerance to the offending food protein. However, EO usually follows a chronic relapsing course. To date, no studies have shown an increased risk of malignancy in patients with EO, but there is evidence that uncontrolled chronic eosinophilic inflammation may cause subepithelial fibrosis and remodelling, which eventually may cause obstructive dysphagia, strictures or persistent oesophageal narrowing.

Alyson Kakakios MB BS, FRACP · Ralf G Heine MD, FRACP

To the Editor: A nationwide colorectal cancer (CRC) screening program will commence in 2006. It has been shown that general practitioners can influence their patients in the decision to have CRC screening.1,2 There are several screening test options in Australia, and the relative geographical isolation of rural centres may influence attitudes and participation. We sought to determine the attitudes of GPs towards CRC screening and test preferences. Between January and September 2005, all GPs in Western Australia (n = 1837; 1298 metropolitan, 539 rural) were sent a questionnaire, which was completed by 801 (43.6%). Overall, 62.8% of respondents believed that asymptomatic average-risk subjects should have CRC screening (67.1% of metropolitan GPs v 54.2% of rural GPs; P = 0.003). The questionnaire revealed major differences between which test GPs would recommend for their patients and which test they preferred for their own personal screening (Box). These differences were related to the factors GPs believed were most likely to influence choice of screening test. For colonoscopy, accuracy and speed of result were considered most important; for faecal occult blood testing, no need for bowel preparation or time off work and no discomfort were considered the strongest determinants. Previous studies that included patients’ views have found that physicians may incorrectly perceive certain factors in screening to be important to their patients.3 There were no significant differences in choice of test between rural and metropolitan GPs. Although the target age group for the Australian pilot study and the national screening program is 55–74 years,4,5 two thirds of GP respondents felt screening should be offered from the age of 50 years, and a quarter believed it should continue beyond 80 years. Many GPs (65%) indicated they would like further education on CRC screening. In summary, there is good support for CRC screening among Western Australian GPs, but the availability of different screening tests and variations in GPs’ opinions are likely to significantly influence clinical practice. Colorectal cancer screening methods and general practitioners’ recommendations and attitudes Test recommended by GP for patients GPs’ perception of patients’ choice of test GPs’ preferred test for their own screening Faecal occult blood testing 430 (53.7%) 396 (49.4%) 236 (29.5%)* Colonoscopy 285 (35.6%) 278 (34.7%) 479 (59.8%)* Flexible sigmoidoscopy 37 (4.6%) 20 (2.5%) 20 (2.5%) Computed tomography colonography 18 (2.2%) 69 (8.6%) 32 (4.0%) Barium enema 0 2 (0.25%) 3 (0.4%) * P = 0.004 (χ2)

Graham B Turner · Marcus W Chin · Noellene M Foster · Jon Emery · Geoff M Forbes

To the Editor: Australia’s imminent bowel cancer screening program will revolve around the general practitioner,1-3 whereas, in the United Kingdom, the GP will have virtually nothing to do with the national screening program now underway.4 It is curious that two programs with the same evidence base regarding effectiveness should be so fundamentally different. One explanation could be the differing health care systems in each nation. However, they are more alike than not, so the true explanation for the Australian methodology could rest with the outcome of the Australian pilot studies. If that is the case, then perhaps one should be both alert and alarmed. Given the inequity in access to GPs in Australia, it is not surprising that the Final Evaluation Report5 of the pilot national screening program stated that: Some GPs interviewed in Woolcott’s Qualitative Research focus groups . . . expressed concern over access to FOBTs [Faecal Occult Blood Tests] for people without a fixed address. It was mentioned that this group, particularly Aboriginal and Torres Strait Islander people and people in low socioeconomic groups, particularly homeless people, did not receive invitations to participate in the Pilot. Some GPs commented that the information packs, in both English and the translated versions, were too complicated for people with low literacy and those from culturally and linguistically diverse backgrounds.5 The same report noted that 38% of people overall (men, 42%; women, 34%) and 52% of non-English speakers did not visit their GP after a positive FOBT. Nevertheless, the report favours the continued central role of the GP.5 This is not the case in the UK screening program, which has a more direct approach, with program hubs and associated screening centres — all with defined accountabilities. The Australian approach is to simply add to the workload of GPs — a more pragmatic approach in the short term, but less imaginative. Our program will undoubtedly be a step forward in colorectal cancer prevention. The question is how large that step will be. Reliance on the existing system threatens to reinforce existing health care inequities.

Allan D Spigelman

To the Editor: The recent report by the Multicentre Australian Colorectal-neoplasia Screening (MACS) Group1 offered some intriguing findings. Participation in bowel cancer screening was lower than expected, despite a range of tests being offered. In addition, people offered a choice of different faecal occult blood tests (FOBTs) were less likely to participate than those not offered this choice. The accompanying editorial by Salkeld and colleagues concluded that “Informed consumers making smart choices about screening . . . would be a public health success”.2 We believe the available evidence indicates otherwise. As the MACS Group study showed, participation in FOBTs was lower than in the Australian Government FOBT pilot program,3 and participation in screening by colonoscopy was lower than for other studies, including our recent Australian study.4 They suggested this may be because local general practitioners were not engaged in the project. Our study was designed to address this issue, and concluded that involvement of GPs had a small, non-significant effect on participation rates and no effect on response to invitation.4 This is not to say that involvement of GPs is undesirable. The MACS Group study found participation in FOBTs of 27.4% when only an FOBT kit was provided and a significantly lower participation when a choice of four screening modalities was offered, with an FOBT kit provided (18.6%, P = 0.03). These data confirm the findings of a large multicentre study from the SCORE2 Working Group.5 In that study, participation in FOBT was 30.1%, while participation in either FOBT or flexible sigmoidoscopy, when a choice of the two was offered, was 27.1%. The authors did not offer this analysis, but the difference was again significant (P = 0.015, two-tailed Fisher’s exact test). The editorial by Salkeld et al suggested that the Australian bowel cancer screening program should incorporate decision-support systems to allow informed choice of screening options. The “choice paradox” reported by the MACS and SCORE2 studies argues against this. Further, there is no evidence that decision support improves rates of participation in screening, and some explicit evidence that it has no effect.6 This should not be troubling. At this time, most colon cancer screening is still performed after consultation between patient and doctor, and in this setting informed choice is possible and desirable. Decisionmakers such as the Australian Government Department of Health and Ageing use a different process, which is explicit and quantitative,7 in determining screening policy. With the evidence available, the Australian mass-screening program should offer and evaluate a single test modality.

Douglas R Taupin · Mike Corbett

In reply: Taupin and Corbett contend that the “choice paradox” reported by the MACS Group study argues against decision-support systems to allow informed choice of screening options. That would be true if the purpose of informed choice was simply to increase participation in screening.1 Our point is that the purpose of informed choice is to support an ethical basis for individuals’ decisions about screening.2,3 This can occur within the single-test modality (faecal occult blood tests) of the national screening program. It would be desirable to have decision-support systems embedded in a doctor–patient consultation. But this may not be feasible in terms of screenee access to a general practitioner, nor affordable for the Australian Government — hence our call for a self-directed decision-support system as an adjunct to a doctor-guided system. This is one way of applying the principle that patients should be given unbiased information on the benefits and harms of screening that enables them to make an informed choice about their own participation in screening.4

Glenn P Salkeld · Jane M Young · Michael J Solomon

Objective: To determine the usefulness of abdominal ultrasonography for diagnosing colorectal cancer in patients presenting with abdominal distension.Design, setting and participants: A prospective case series of consecutive adult patients with abdominal distension admitted to the National Taiwan University Hospital between January 2001 and July 2004. All participants were examined by abdominal ultrasonography. Those with suspected colorectal tumours on ultrasonography had follow-up colonoscopy, while all other patients had computed tomography scans.Main outcome measures: Accuracy of abdominal ultrasonography for diagnosing colorectal cancer in patients with abdominal distension; incidence of colorectal cancer.Results: Of 511 patients eligible for inclusion in our study, 97 (19.0%) were confirmed to have colorectal cancer. For diagnosis of colorectal cancer, ultrasonography had a sensitivity of 92.8% (95% CI, 85.2%–96.8%); a specificity of 98.8% (95% CI, 97.0%–99.6%); a positive predictive value of 94.7% (95% CI, 87.6%–98.0%); a negative predictive value of 98.3% (95% CI 96.4%–99.3%); and an accuracy of 97.7%.Conclusion: Ultrasonography is a sensitive tool for diagnosing colorectal cancer in patients presenting with abdominal distension.

Shyr-Chyr Chen MD, MBA · Zui-Shen Yen MD, MPH · Hsiu-Po Wang MD · Chien-Chang Lee MD, MPH · Chiung-Yuan Hsu MD · Wen-Jone Chen MD, PhD · Chien-Yao Hsu MD · Hong-Shiee Lai MD, PhD · Fang-Yue Lin MD, PhD · Wei-Jao Chen MD, PhD

Objective: International guidelines and local practices for colorectal cancer screening suggest an important role for several different screening tests, and for consumer choice. We aimed to determine whether choice of test improved participation in screening.Design: A randomised comparative study offering one of six screening strategies: faecal occult blood testing (FOBT), FOBT and flexible sigmoidoscopy (FS), computed tomography colonography (CTC), colonoscopy, or one of two groups offered a choice of these strategies (one of which was sent an FOBT kit with the letter of invitation, while the other was required to request an FOBT kit by telephone if that was the test chosen).Setting and participants: 1679 people aged 50–54 or 65–69 years, randomly selected from the electoral roll in metropolitan Perth, Adelaide and Melbourne.Main outcome measures: Participation, yield of advanced colorectal neoplasia (CRN), acceptability and safety.Results: 346 (20.6%) were excluded from screening, mostly for a recent examination (165), symptoms (72) or personal or family history of colorectal neoplasia or cancer (83). 278 of the 1333 eligible (20.9%; 95% CI, 18.7%–23.1%) participated in screening. Participation was similar by age and sex, but lower in Perth than Adelaide (17.1% v 24.2%; P = 0.01). Participation by screening group was: FOBT, 27.4%; FOBT/FS, 13.7% (P < 0.001 compared with FOBT); CTC, 16.3% (P = 0.005); colonoscopy, 17.8% (P = 0.02); or a choice of test 18.6% (“with FOBT kit”; P = 0.03) or 22.7% (“without FOBT kit”; P = 0.3). Yield of advanced CRN was higher in participants screened by colonoscopy than FOBT (7.9% v 0.8%; P = 0.02). All tests were well accepted and no serious complications arose from screening.Conclusion: A choice of screening test did not improve participation. Participation by FOBT was higher than by other tests. Yield of advanced colorectal neoplasia on an intention-to-screen basis, determined by test sensitivity and participation, is likely to be a critical determinant of the effectiveness of screening strategies.

The Multicentre Australian Colorectal-neoplasia Screening (MACS) Group

A 90-year-old woman was admitted with acute vomiting for the previous 24 hours and acute, diffuse, intense abdominal pain. Her medical history included atrial fibrillation and hysterectomy. Clinical examination revealed a generally distended, tender, tympanic and silent abdomen. Blood pressure was 98/54 mmHg; her pulse was irregular at 67 beats per minute. A plain x-ray of the abdomen revealed bowel distension without air–fluid levels. A computed tomography scan of the abdomen showed the presence of gas in the portal venous system (Box 1) and the intestinal wall (Box 2). At laparotomy, 142 cm of necrotic ileum was removed. The postoperative course was uneventful, and the patient was discharged 1 month later. The precise mechanism for formation of gas in the portal venous system is uncertain.1 The primary factors that favour this development are intestinal wall alterations, bowel distension and intra-abdominal sepsis.1 In many cases, two or three of these circumstances may coexist. In 15% of cases, the cause of air in the portal venous system remains unknown.1 Several conditions, such as perforated ulcer, interventional procedures, trauma, and infectious or inflammatory abdominal diseases, but most commonly intestinal ischaemia (as in our patient), can cause alterations of the gastric and bowel wall, permitting the passage of intraluminal gas into the portal venous system.1-3 Gas from the intestinal lumen passes through the intestinal wall and travels via the small mesenteric veins and the superior or inferior mesenteric vein to the portal venous system. Further, the presence of gas simultaneously in the portal venous system and intestinal wall seems to be specific to intestinal ischemia.4 Because hypotension was moderate and usual in our patient, we believe that the intestinal ischaemia was secondary to embolic disease related to atrial fibrillation. This case also demonstrates that portal venous system gas formation does not necessarily imply a worse prognosis and that surgery may be beneficial in the presence of this sign.1,5 Abdominal computed tomography scan with contrast showing air in the portal venous system (1) and intestinal wall (2).

Ali Mofredj MD · Harry Toledano MD · Richard Boutboul MD

Clinical record Patient 1* Histopathology of colonic biopsies from the patients described† A: Biopsy from Patient 1 shows mild diffuse increase of lymphocytes within the crypt and surface epithelium (arrows). Inflammatory cell infiltration of the lamina propria is minimal and there is no collagen deposition.‡ A 78-year-old woman who normally had one to two bowel motions a day presented with a 3-week history of gradual onset of more frequent and looser motions up to six times a day and occasionally at night. She had urgency and episodes of faecal incontinence. Her motions were watery. There was no improvement after she was treated with tinidazole. Her regular medications were alendronate, hydrochlorothiazide, irbesartan, raloxifene, thyroxine, temazepam, aspirin, thiamine, vitamin C, glucosamine, and evening primrose oil. She had also taken lansoprazole (30 mg capsules) daily for 2 months to treat possible reflux. Sigmoidoscopy showed yellowish, watery stool with mucus and a few tiny scattered patches of intramucosal haemorrhage. Biopsy (Figure A) showed mild lymphocytic colitis (defined by the presence of more than one lymphocyte per 20 epithelial cells). Lansoprazole therapy was ceased and the diarrhoea settled within 24 hours. Patient 2* B: Biopsy from Patient 2 shows mild increase in plasma cells in the lamina propria and an obvious subepithelial band of collagen (arrows). There is patchy detachment of the surface epithelium.§ A 53-year-old woman presented to hospital with a 2-month history of diarrhoea and abdominal pain. She was admitted with suspected diverticulitis. Nine months before presentation, she had started taking lansoprazole (30 mg capsules) daily for heartburn. She described having bowel motions seven or eight times per day, and occasionally at night. The diarrhoea had not responded to tinidazole, metronidazole or norfloxacin. Her regular medications were gemfibrozil, tibolone, indapamide and perindopril. Colonoscopy did not detect significant abnormalities. Biopsy showed mild collagenous colitis (Figure B). Lansoprazole therapy was ceased. She was treated with loperamide, then cholestyramine and budesonide. Within 6 months, she was symptom-free without antidiarrhoeal medication. Patient 3* C: Patient 3, there is a mild increase in lymphocytes within the epithelium, and of plasma cells in the lamina propria. The surface epithelium is mildly degenerate, and there is a suggestion of abnormal subepithelial collagen deposition (arrow).‡ A 79-year-old woman had been seen in hospital with a sudden onset of diarrhoea 5 weeks before admission, associated with 10 kg weight loss. She had taken lansoprazole (30 mg capsules) daily for 2 weeks before the diarrhoea began. She had had up to 12 watery motions per day. The diarrhoea had not responded to norfloxacin. Her regular medications were alendronate, perindopril, potassium chloride, chlorthalidone and celecoxib. Appearance on sigmoidoscopy was normal, but biopsies (Figure C) showed mild lymphocytic colitis with occasional subepithelial collagen suggesting transition to collagenous colitis. Lansoprazole therapy was ceased, and she was treated with codeine. Her condition was much improved without antidiarrhoeal medication within a month. When reviewed 6 months later, she was having two loose motions daily, and the celecoxib (200 mg daily) was withdrawn. * All three patients had negative results on serological tests for coeliac disease (transglutaminase and endomysial antibody) and normal serum IgA levels. † Follow-up biopsies were not taken. ‡ Stained with haematoxylin and eosin. § Stained with Masson Trichrome, which shows collagen in green. Microscopic colitis is increasingly recognised as a major cause of persistent diarrhoea.1 It is an idiopathic clinicopathological syndrome of chronic watery non-bloody diarrhoea associated with a normal appearance on colonoscopy and specific histopathological changes of lymphocytic and/or collagenous colitis. There is a high rate of spontaneous resolution and relapse in microscopic colitis, and effective treatment is limited. The pathogenesis of microscopic colitis is poorly understood and is thought to be related to a poorly regulated epithelial immune response to luminal or epithelial antigens including bile acids, toxins, or infectious agents.2 Microscopic colitis has been associated with autoimmune diseases and with exposure to medications, predominantly non-steroidal anti-inflammatory drugs, and, rarely, salicylates, simvastatin, ticlopidine, ranitidine, carbamazepine, Cyclo 3 Fort (a combination of Ruscus aculeatus extract, hesperidin methylchalcone, and ascorbic acid), flutamide, gold salts, and, recently, lansoprazole.2 Lessons from practice Microscopic colitis should be considered as a cause of persistent watery non-bloody diarrhoea; the two major histological forms of microscopic colitis — lymphocytic and collagenous — are probably variations of the same disorder. Even if the bowel appears macroscopically normal on colonoscopy, multiple biopsies should be routinely taken for histopathology, and the pathologist informed of the possibility of microscopic colitis. The possibility of medications as the cause of diarrhoea and microscopic colitis should always be considered; drugs that may cause the symptoms should be investigated, withdrawn, and the effects observed. Not all adverse drug events occur immediately and not all are class effects. Suspected adverse drug reactions should be reported to the Adverse Drug Reaction Advisory Committee for evaluation. Lymphocytic and collagenous colitis are probably aetiologically related, and may be a spectrum of the same disease.3 This is supported by our report of three lansoprazole-associated cases, in which one showed lymphocytic colitis, one showed collagenous colitis and one showed transitional features of both on histopathological examination. The inflammatory infiltrate and collagen deposition in these cases was less severe than that seen in typical cases of lymphocytic and collagenous colitis, probably because of the limited periods of exposure to the inciting agent. Interestingly, Patient 2 had the greatest collagen deposition and the longest exposure to lansoprazole. The rapid resolution of the symptoms on cessation of lanzoprazole therapy in Patient 1 may be related to the very mild lymphocytic infiltrate and lack of collagen deposition, or perhaps lansoprazole caused diarrhoea through another mechanism in this case. Adverse drug reactions can only be recognised if doctors maintain a high index of suspicion. It is not possible to know every possible reaction to the medications that patients are exposed to, and some may not have been recognised before. The absence of an immediate temporal relationship between a drug and the associated disorder (as was seen in the cases presented here) may contribute to failure to diagnose a drug-induced disease. Drugs and their metabolites may affect the colon directly through their pharmacological actions, or through hypersensitivity reactions. Drugs also act indirectly on the colon by altering colonisation by gastrointestinal organisms.4 The association between lansoprazole and microscopic colitis may be related to its action on the colonic proton pump, affecting colonic secretions and pH, which may affect colonic flora and bile salt solubility.5 Alternatively, there may be an idiosyncratic direct hypersensitivity reaction by the colonic mucosa. The association between lansoprazole and microscopic colitis has been reported previously,5-7 and before our patients presented, the Adverse Drug Reaction Advisory Committee had received a single report of a probable association between rabeprazole and unspecified colitis. In a previous report of lansoprazole-associated microscopic colitis,5 substituting omeprazole for lansoprazole did not lead to recurrent diarrhoea. One explanation for the absence of a class effect is that while all proton-pump inhibitors bind the parietal cell proton pump covalently at cysteine 813 or 822, only lansoprazole and rabeprazole bind to cysteine 321.8 Binding of cysteine 321 also inhibits the colonic proton pumps.9 This may affect colonic secretion and pH, predisposing to diarrhoea and microscopic colitis. However, the rarity of the association between lansoprazole and microscopic colitis favours an idiosyncratic immune reaction. Minor variations in the structures of the proton-pump inhibitors may result in different immunological activation, although allergic cross-sensitivity has occasionally been reported among the proton-pump inhibitors.10 Our cases illustrate the importance of considering exposure to medication as a cause of chronic diarrhoea and particularly microscopic colitis, and that adverse drug reactions are not always class effects. Stopping the causative medication may reverse the abnormality in this difficult-to-treat condition. In patients with persistent watery diarrhoea, biopsies should be performed even if colonoscopic appearances are normal, and pathologists should be informed of the clinical possibility of microscopic colitis, as the histological changes may be subtle.

Sarah N Hilmer PhD, FRACP · Timothy R Heap MB BS, FRACP · Robert P Eckstein MB BS, FRCPA · Gillian M Shenfield DM, FRACP · Christopher S Lauer MB BS, FRCPA

During routine colonoscopy we were surprised to find this sign of peace produced by colonic mucosal folds (see Figure). The image was produced with a Pentax EC-3870FK2 video colonoscope.

Richard H Riley FANZCA · Richard Sallie MD, FRACP

A 30-year-old woman with a history of bipolar disorder presented after ingestion of 4 g of carbamazepine and 40 mg of alprazolam in a suicide attempt. She became increasingly drowsy and eventually required intubation and mechanical ventilation. The patient received a total of 350 g of activated charcoal via a nasogastric tube over 25 hours. She experienced some charcoal-stained vomiting, but made an otherwise unremarkable recovery and was discharged home. The patient was hospitalised again 3 days later complaining of constant sharp abdominal pain, diagnosed on culture as a urinary tract infection. Over the next 4 days, she experienced multiple bouts of nausea, charcoal vomits and colicky central abdominal pain. A Gastrografin meal (Schering Pty Ltd, Sydney) and follow-through showed complete small bowel obstruction (Box 1). Subsequent laparotomy revealed a 2.5 × 5 cm bezoar in the distal small bowel, about 60 cm from the ileocaecal valve (Box 2). The bezoar was manually fragmented and passed through the ileocaecal valve into the caecum. The patient recovered well after the operation, and was discharged with psychiatric follow-up. Charcoal bezoars are a rare complication of activated charcoal administration. They have been associated with treatments for intoxication with carbamazepine,1 amitriptyline,2 theophylline,3 benzodiazepines and barbiturates.4,5 The parasympatholytic effects of the drugs can precipitate or contribute to paralytic ileus, allowing charcoal to accumulate (potentially with remnants of undigested tablets) and form bezoars. Additional factors that influence bowel obstruction secondary to charcoal administration include the dose and timing of activated charcoal therapy, patient age and comorbidities, and previous intra-abdominal surgery. Gastrointestinal complications should be considered whenever activated charcoal is administered. Prompt recognition and treatment at the first sign of ileus or obstruction may prevent bowel necrosis and subsequent perforation/peritonitis. Charcoal-stained vomiting, abdominal distension and ongoing pain should raise the suspicion of mechanical obstruction. We would advocate Gastrografin follow-through as both a diagnostic and potentially therapeutic tool in incomplete obstruction. However, complete obstruction may signal the need for early laparotomy. 1 Abdominal x-ray after Gastrografin meal, showing complete small bowel obstruction 2 Operative photograph of bezoar in distal small bowel

Justin C Y Chan MB BS BMedSci · Chaminda Saranasuriya MBBS · Bruce P Waxman MB BS, FRACS, BMedSci

Graham D Hocking Child Psychiatrist, 25 High St South, Kew, VIC 3101. ghockinATiprimus.com.au To the Editor: Catto-Smith gives a very good account of the medical management of constipation and soiling in children,1 but fails to mention psychological, interpersonal and social factors in the main part of his article. He does mention “behavioural abnormalities” towards the end, in the section “When to refer”. I think it is generally accepted among paediatricians and child psychiatrists that the problem of constipation and soiling, or encopresis, often has multiple determinants and varied psychological effects on the child and the family. Certainly, the older the child is, the more likely these effects will be present.2 If this condition is to be managed in general practice over a period of “6–12 months”, general practitioners need to be aware of these factors so they can be addressed. Twelve months is a long time in the life of a 5 year old, and in that time pathological patterns can become well established and hard to shift. Most children over 5 years with soiling have developed secondary psychological problems as a result of the soiling.3 At this stage, assessment by a child psychiatrist will often reveal that the child has developed a pathological fantasy world around what they believe is happening inside them. The physical management of constipation and soiling is an essential part of the management no matter what the aetiology, but addressing the psychological interpersonal and social factors is equally important. If these factors are obvious to the GP and are not responding to intervention, the family should be referred to a child psychiatrist. With children over 5 years, the secondary effects have almost always become significant, and I believe that all these families should be referred for assessment. Catto-Smith quotes a 30%–50% relapse rate,1 and “there is evidence that they do not improve on reaching puberty”. This is not my experience with families that have the benefit of a multidisciplinary approach to the disorder.

Graham D Hocking

Anthony G Catto-Smith Director, Gastroenterology and Clinical Nutrition, Royal Children's Hospital, Flemington Road, Parkville, VIC 3052. tony.cattosmithATrch.org.au In reply: Hocking emphasises secondary behavioural and emotional effects that occur in some children with longstanding faecal soiling. Fortunately, there is good evidence that these tend to resolve with effective multimodal treatment of the constipation.1 The relatively high long-term relapse rate of soiling among children who have been treated in tertiary centres has only recently been recognised,2 but the psychological features of this relapsing group are not well defined. My review was directed toward general practitioners and was as much as possible evidence-based. I am unaware of any good quality evidence to support Hocking’s assertion of the benefits of automatic referral to a child psychiatrist of all children over the age of 5 years with ongoing faecal soiling. Given the beliefs of both myself and Hocking, that constipation and soiling are likely to have multiple determinants and varying psychological effects, it would seem to be appropriate to triage “problem” patients through a general paediatrician, with referral for psychological assistance if deemed appropriate. This is best summed up in my article in the section that Hocking mentions, “When to refer”.3

Anthony G Catto-Smith

Obesity is shaping up to be the major health care problem and one of the most frequent causes of preventable death in Western countries in the 21st century. Bariatric surgery is the only current treatment that has been shown to achieve major and durable weight loss. Major weight loss in the severely obese leads to total or partial control of a wide range of common and serious diseases, such as diabetes, heart disease and hypertension. Laparoscopic adjustable gastric banding is the most common type of obesity surgery performed in Australia. It is effective, relatively safe and minimally invasive. The blocks to broader application of bariatric surgery should be identified and resolved.

Paul E O’Brien MD, FRACS · Wendy A Brown MB BS, PhD, FRACS · John B Dixon MB BS, PhD, FRACGP

A 25-year-old man presented with macroscopic haematuria associated with a body mass index of 20 kg/m2 and a severe coagulopathy consistent with vitamin K deficiency. The diagnosis of a profound malabsorption syndrome secondary to coeliac disease was confirmed by small bowel histology and positive coeliac serology. Clinical recordA 25-year-old man presented to our hospital’s emergency department with frank haematuria and bilateral loin pain. He had been unwell for 10 days with abdominal pain and vomiting. Two days before presentation, he sought medical attention and was prescribed naproxen for the abdominal pain (he only took two of these tablets); he was on no other medications. Two years previously, a diagnosis of irritable bowel syndrome had been made by his doctor, based on a history of loose bowel motions over 2 years and a normal colonoscopy. 1 Angular cheilitis This image is similar to but is not of the patient described in this article. At the time of his hospital presentation, our patient had a body mass index (BMI) of 20 kg/m2 and appeared pale, but had no sign of bruising. He was apyrexial and well hydrated, with unremarkable cardiovascular and respiratory findings. Gastrointestinal examination revealed a red swollen tongue and angular cheilitis (Box 1). There was some mild left renal angle tenderness. Urinalysis revealed > 500 ×106 non-glomerular red blood cells per litre (normal range < 13 ×106/L) but no pyuria. Abdominal ultrasound performed on the day of review was unremarkable. Peripheral blood tests revealed normal electrolyte levels and renal function, with a coagulopathy reflected by an international normalised ratio (INR) of > 10 (reference range, < 1.4) and activated partial thromboplastin time (APTT) of 115 seconds (reference range, 25–38 seconds) (Box 2). These results were verified with repeat testing. Clotting factor studies revealed severe deficiencies in factors II, VII, IX and X. There were also deficiencies in iron, vitamin B12, and serum folate, vitamin A and vitamin E concentrations (Box 2). Vitamin D levels were normal, but the serum alkaline phosphatase level was twice the upper limit of normal. Results of other liver function tests were normal. Endomysial antibody test results were positive, and tissue transglutaminase antibody (IgA) levels were more than five times the upper limit of normal. A preliminary diagnosis of vitamin K deficiency, leading to a profound coagulopathy, secondary to a malabsorption syndrome from coeliac disease was made. Duodenal biopsies confirmed the diagnosis of coeliac disease (Box 3). Gastroscopy showed no gastritis or peptic ulcer disease. The duodenal mucosa appeared abnormal and was consistent with villous atrophy. No conclusion could be reached regarding the nonspecific abdominal pain the patient experienced on presentation. Inflammatory bowel disease was a differential diagnosis, but inflammatory markers were only marginally elevated (Box 2). Parenteral vitamin K was administered, and the prothrombin time reduced to 15 seconds (INR, 1.3) and APTT became normal within 24 hours. The macroscopic haematuria resolved. The patient was also given parenteral iron and vitamin B12 and oral multivitamin replacement. He was reviewed by the dietetic services and educated about coeliac disease. He also started a gluten-free diet. On clinical review 2 months later, he was feeling well and had put on 15 kg in weight (BMI, 26 kg/m2). Duodenal biopsies 6 months after the initial presentation revealed variable villous abnormality, consistent with partially treated coeliac disease. The vitamin deficiencies had all resolved, and clotting factor studies were normal. Endomysial antibodies had become negative, and tissue transglutaminase antibody levels (IgA) were within the normal reference range. Bone mineral density at this time was normal. Genotyping revealed the presence of the HLA DQ2 allele. DiscussionThe most common presentation of coeliac disease is diarrhoea (43%); other reasons for presentation include anaemia (8%), bone disease (6%), weight loss (6%), and abdominal pain (5%).1 The remaining 32% of patients are asymptomatic or present with vague symptoms. Delay in the diagnosis of coeliac disease is common (mean, 11 years).2 About 47% of patients will have been misdiagnosed, and of those with classical symptoms of coeliac disease, 59% have been misdiagnosed as having irritable bowel syndrome.3 The clinical diversity of coeliac disease is recognised, and disorders involving nearly every organ system have been described with this condition.4 Haematuria is an unusual presenting symptom of coeliac disease.5 The haematuria in our patient indicated a systemic bleeding diathesis due to malabsorption of vitamin K, with subsequent prolongation of the prothrombin time and APTT.6 In patients with untreated coeliac disease, the prothrombin time is prolonged (INR, ≥ 1.4) in about 20%,7 and the coagulopathy is usually asymptomatic,8,9 but corrects quickly with administration of vitamin K.10 In the long-term treatment of coeliac disease, a gluten-free diet is paramount and leads to resolution of the underlying small bowel villous abnormality. A partial histological response to a gluten-free diet may reflect poor compliance, accidental exposure to gluten-containing food products, or slow recovery. Our patient claimed close adherence to the diet and was provided with adequate education and dietetic review. Histological recovery can be delayed, with 35% of patients still showing features of coeliac disease up to 2 years after commencing a gluten-free diet.11,12 Failure to achieve histological recovery can occur in the absence of gluten.13 The use of serology tests as surrogate markers for histological resolution is controversial.14-16 Some coeliac patients with negative results for endomysial and tissue transglutaminase antibodies have persistent villous abnormalities, necessitating the use of biopsies to monitor the response to a gluten-free diet (as in our patient). On HLA genotyping, our patient possessed the HLA DQ2 allele, which is present in more than 90% of patients with coeliac disease, but in only 20% of the general population.17 Endomysial IgA antibodies have a sensitivity of 90% and specificity of 100%, tissue transglutaminase IgA has a sensitivity and specificity of 98%. Sensitivities for the IgG class endomysial and tissue transglutaminase antibodies are around 40%.18 Deficiency of IgA occurs in 1.7%–2.6% of coeliac patients, and these patients have negative results on IgA antibody testing. Total serum IgA levels should be tested together with IgG (endomysial, tissue transglutaminase) to improve the overall sensitivity of antibody testing.19 Our patient demonstrates the need to be aware of the various presenting features of coeliac disease, and reminds us that coeliac disease is often misdiagnosed. 2 Results of laboratory tests Investigation Result Reference range Haemoglobin 123 130–180 g/L White cell count 12.4 4.0–11.0 × 109/L Platelets 727 150–400 × 109/L Mean corpuscular volume 73 82–95 fL Mean corpuscular haemoglobin 23.3 24–32 pg Reticulocytes 74 10–90 × 109/L Differential white cell count Neutrophils 9.52 2.0–7.5 × 109/L Lymphocytes 1.48 1.0–4.0 × 109/L Monocytes 1.25 0.1–0.8 × 109/L Eosinophils 0.12 < 0.4 × 109/L Basophils 0.02 < 0.2 × 109/L Blood film Microcytosis with moderate hypochromasia, and small numbers of elongated cells. Neutrophilia with no left shift and thrombocytosis. CRP 22.4 1.6–8.7 mg/L ESR 14 5–12 mm/hour Alkaline phosphatase 249 40–129 U/L Tissue transglutaminase antibody 101 < 20 U/mL Serum folate 4.7 > 6.8 nmol/L Red cell folate Not measured Vitamin B12 93 132–857 pmol/L Vitamin A 0.1 0.8–3.1 μmol/L Vitamin D 38 30–110 nmol/L Vitamin E 4.9 > 18.6 μmol/L Ferritin 10 24–336 μg/L International normalised ratio > 10 < 1.4 Activated partial thromboplastin time 115 25–38 seconds Factor II 15% 50%–120% Factor VII 3% 50%–120% Factor IX 17% 50%–120% Factor X 5% 50%–120% 3 Duodenal biopsies A At time of diagnosis. There is severe villous abnormality associated with marked enterocyte damage (black arrows), in keeping with untreated coeliac disease. B After 6 months of gluten-free diet. There is partial improvement of the previous villous abnormality, but a moderate villous abnormality with associated crypt hyperplasia remains.

John S Lubel BDS, MB BS, MRCP · Louise M Burrell MBChB, MD, FRACP · Vicki Levidiotis MB BS, FRACP, PhD.

A 44-year-old woman presented with a 3-day history of worsening right upper quadrant pain associated with nausea, anorexia and fever. She had localised tenderness and guarding in the epigastrium. Initial ultrasound imaging showed a distended gallbladder containing calculi, without wall thickening. A computed tomography scan (performed because of increasing pain) showed a whorled structure in the anterior abdomen (Box 1). At laparotomy this was seen to be torsion of a segment of the greater omentum. Histology of the resected specimen showed congestion of the vessels, haemorrhagic infarction and focal fat necrosis (Box 2). She made an uncomplicated and rapid recovery. Primary omental torsion is a rare cause of acute abdomen. It may affect children and adults, and is commonly misdiagnosed preoperatively as appendicitis. Some cases have presented as acute cholecystitis.1,2 Kimber et al identified 13 cases of omental torsion or infarction in about 8000 cases of suspected appendicitis over a 20-year period.3 Large meals, sudden postural change, and abdominal trauma may be precipitating factors in primary torsion,4 while adhesions, hernia, tumour or focus of inflammation3 occur with secondary torsion. Resection is the preferred treatment,3,4 although some clinicians suggest conservative management2 when the diagnosis is apparent on computed tomography. Laparoscopic resection has also been advocated.5 1 Computed tomography scan of abdomen A whorled structure (arrow) is seen in the anterior abdomen adjacent to the transverse colon. 2 Histological section of omentum There is marked congestion and haemorrhagic infarction, with areas of fat necrosis and an acute inflammatory reaction. (Inset: higher mag-nification shows fat necrosis and a polymorph neutrophil response.)

Wen-Chan Yeow MB BS · Mohan V Jayasundera FRACS, MB ChB · Graham Hool FRACS · Rajalingam Sinniah DSc, FRCPath, FRCPA

Objective: To evaluate the role of zinc and vitamin A supplementation in the recovery of Indigenous children hospitalised for acute diarrhoea.Design: A randomised controlled 2 by 2 factorial trial of supplementation with zinc and vitamin A.Setting and participants: Aboriginal children (aged < 11 years) hospitalised for acute diarrhoea at Alice Springs Hospital, Northern Territory, April 2001–July 2002.Main outcome measures: Duration of diarrhoeal illness; re-admission for diarrhoeal illness within 120 days.Results: Our study involved 392 Aboriginal children with 436 episodes of diarrhoea. Supplementation with zinc, vitamin A, or combined zinc and vitamin A had no significant effect on duration of diarrhoea or rate of re-admission compared with placebo. Median diarrhoea duration after starting supplementation was 3.0 days for the vitamin A and zinc supplemented and placebo groups (P values 0.25 and 0.69, respectively). The number of re-admissions did not differ significantly between those receiving vitamin A or zinc and the relevant placebo groups (relative risk [95% CI], 1.2 [0.7–2.1] and 1.3 [0.8–2.1], respectively).Conclusion: Vitamin A and zinc supplementation may not be indicated for in-hospital management of acute diarrhoeal disease in Aboriginal children living in remote areas. This finding may not apply to children with malnutrition, for whom other studies suggest a benefit. Larger trials incorporating more comprehensive data on the vitamin A and zinc status as well as nutritional status of study populations might help to explain the different results in different populations.

Patricia C Valery MD, MPH, PhD · David M Purdie BSc(Hons), PhD · Paul J Torzillo MB BS, FRACP, FFICM · Peter A Stewart MB BS, FRCPA · Naomi C Boyce B Nursing · Anne B Chang MPHTM, FRACP, PhD · Andrew V White MB BS, FRACP · Gavin R Wheaton MB BS, FRACP · John Wakerman MB BS, MPH

Clinical records Patient 1 A 31-year-old Indigenous man from a remote Cape York community was referred for investigation of weight loss from 57 kg to 36 kg over the previous 3 years. He drank alcohol heavily and had had seizures which were probably related to this alcohol use. He had diffuse crusted scabies and there was obvious wasting. His haemoglobin level was 90 g/L (normal range [NR], 130–180 g/L) with a mean cell volume (MCV) of 97 fL (NR, 80–100 fL). His white cell count was 11.4 × 109/L (NR, 4–11 × 109/L) with an eosinophilia of 4.34 × 109/L (NR, 0.04–0.4 × 109/L). The platelet count was normal. His ferritin level was 430 μg/L (NR, 30–300 μg/L) and levels of serum vitamin B12 and red cell folate were normal. The prothrombin time was 18 seconds (NR, 11–15 seconds). The albumin concentration was 16 g/L (NR, 35–45 g/L). Upper abdominal ultrasound showed diffuse increased echogenicity of the liver. Stools had no white cells, red cells, bacterial pathogens, ova, cysts or parasites. Skin scrapings confirmed the clinical diagnosis of scabies. Results of serological testing for HIV, strongyloides and coeliac disease were negative. He was deficient in vitamin A (0.1 μmol/L; NR, 1.6–2.3 μmol/L), 25-hydroxyvitamin D (< 12 nmol/L; NR, 25–150 nmol/L) and vitamin E (6 μmol/L; NR, 11–45 μmol/L). Despite a high-protein and high-calorie diet, on Day 13 of his admission his weight was 34.4 kg. Small bowel biopsies taken at upper gastrointestinal endoscopy showed partial villous atrophy and a significant increase in inflammatory cells within the lamina propria. The endoscopic findings suggested tropical sprue, and prompted commencement of doxycycline and folate therapy. His weight improved immediately, and on discharge 10 days later he weighed 38.5 kg. Ten weeks after discharge, despite continuing to misuse alcohol, he had gained 7.6 kg. His albumin level had increased to 33 g/L and the prothrombin time was 11 seconds. His haemoglobin level was 136 g/L, and white cell and eosinophil counts were normal. Patient 2 1 Upper gastrointestinal biopsy on Day 13 of admission 2 Upper gastrointestinal biopsy on Day 24 of doxycycline and folic acid therapy A 62-year-old Indigenous man from a community near Cairns presented for investigation of weight loss, diarrhoea and macrocytic anaemia. He drank alcohol heavily and weighed only 39.5 kg. Investigations showed a haemoglobin level of 73 g/L, MCV of 97 fL, a white cell count of 3.9 × 109/L and a platelet count of 97 × 109/L (NR, 150–400 × 109/L). The blood film showed hypersegmented neutrophils and macrocytes. His serum vitamin B12 level was normal and red cell folate level was < 73 nmol/L (NR, 295–1800 nmol/L). His ferritin level was 244 μg/L. His potassium level was 2.3 mmol/L (NR, 3.4–4.5 mmol/L) and creatinine clearance was normal. His albumin level was 30 g/L, but other liver function test results were normal. Thyroid function was normal. Upper gastrointestinal endoscopy and biopsy performed on Day 13 of admission showed partial villous atrophy and increased inflammatory cells in the lamina propria (Figure 1). Therapy with doxycycline and folic acid was begun on Day 17 of admission. Follow-up endoscopy and biopsy 24 days after commencing this therapy showed a persistent inflammatory exudate, but an improvement in the villous appearance (Figure 2). By discharge, his haematological and biochemical test results were normal, his diarrhoea had stopped and his weight had improved to 47.3 kg. Patient 3 A 66-year-old Indigenous man from a community near Cairns was referred for investigation of diarrhoea and weight loss. He had previously had a hemicolectomy for diverticular disease. He had lost 15 kg over 3 months and was having significant diarrhoea. In the week before referral he had shortness of breath on minimal exertion. On examination, he weighed 42 kg and was emaciated. Investigations showed a haemoglobin level of 64 g/L with an MCV of 108 fL, a white cell count of 2.6 × 109/L and a platelet count of 40 × 109/L. A blood film showed hypersegmented neutrophils and macrocytes. His serum folate level was 3.2 nmol/L (NR, 6.6–35.3 nmol/L), vitamin B12 level was 137 pmol/L (NR, > 210 pmol/L) and ferritin level was 408 μg/L. The prothrombin time was 14 seconds. His sodium level was 129 mmol/L (NR, 135–145 mmol/L), potassium level was 1.6 mmol/L, urea level was 8.2 mmol/L (NR, 2.5–8.0 mmol/L) and the creatinine level was 0.16 mmol/L (NR, 0.05–0.12 mmol/L). Liver function test results were otherwise normal. Stool microscopy showed no white or red cells, ova, cysts or parasites, and no bacterial pathogens. Upper gastrointestinal biopsy performed on Day 4 of admission showed atrophic villi and an inflammatory infiltrate in the lamina propria. Vitamin B12 and folate supplementation were commenced on the day of admission and doxycycline therapy on Day 7. In the first week of antibiotic therapy, he gained 3 kg and his diarrhoea stopped. His white cell and platelet count became normal and haemoglobin level improved to 79 g/L. His biochemical test results normalised entirely. He was discharged home on Day 12 and booked for repeat endoscopy, but did not return. The clinical presentation, histological findings and response to tetracycline and folate therapy in our patients, seen at Cairns Base Hospital between 1998 and 2004, support a diagnosis of tropical sprue. To our knowledge, these are the first reported cases in Australia. All of our patients were Indigenous patients living in remote Indigenous communities. The alcohol use by the first patient clouds the picture, as high alcohol intake is associated with nutritional deficiencies. However, the partial villous atrophy on biopsy is more suggestive of tropical sprue. Additionally, despite being in hospital and not drinking alcohol for 2 weeks, there was no response in his weight or biochemical test results until therapy for tropical sprue commenced. Finally, his weight gain and improvement in biochemical test parameters continued despite ongoing alcohol misuse on discharge. Seeing resolution of the histological changes in all the patients as they responded to treatment would have been desirable, but their remote location and the fact that two did not attend for follow-up made this difficult. It would also have been interesting to examine the small bowel flora of the patients, but this was not prospectively considered. The absence of an agreed definition for tropical sprue has created difficulties. It has been noted that many asymptomatic people living in the tropics will have subclinical malabsorption and even abnormal small bowel biopsy results when rigorously assessed. Uncertainty in the diagnosis of tropical sprue is compounded by the condition’s incompletely understood aetiology. It is proposed that gastrointestinal infections, which occur at a higher rate in the tropics, cause mucosal injury and initiate the process. The mucosal injury leads to a disturbance in intestinal motility, allowing the overgrowth of coliform bacteria in the small bowel.1,2 Enterotoxins from these bacteria potentiate the mucosal damage which leads to malabsorption and protein loss. The resulting nutritional deficiencies hinder epithelial recovery, and a vicious circle develops. Alteration in the intestinal microflora may also lead to a change in toll-like receptor signalling and further retardation of intestinal healing.3 It seems likely that tropical sprue is not a single disease, but rather a pathophysiological process with heterogeneous manifestations resulting from subtle differences in the interplay of the initiating infection and the patient’s diet, living standards and genetics.4 A milder form of the process may lead to asymptomatic abnormalities — sometimes referred to as tropical enteropathy — while more severe expression may lead to the classical debilitating tropical sprue syndrome. The nature and the extent of the nutritional deficiencies are likely to be related to the duration of the disease and the extent of bowel involved. Admission rates for gastroenteritis in Australia are nearly seven times higher in Indigenous communities than in non-Indigenous communities. Rates are also higher in regional and remote areas than in urban settings.5 Our own analysis of hospital separation data from Queensland’s northern zone for 1996–2001 shows an age-standardised hospital separation rate for gastrointestinal infection 2.32 times higher for Indigenous patients than non-Indigenous patients (95% CI, 2.17–2.48). A number of studies of chronic diarrhoea in Indigenous Australian children have documented partial villous atrophy on small bowel biopsy.6,7 Higher colony counts and more frequent isolation of gram-negative organisms are found in duodenal aspirates from Indigenous children with chronic diarrhoea.8 Some of these studies noted a similarity with contemporary Indian studies of tropical sprue, but therapy with antibiotics and folate does not seem to have been considered in the Indigenous children. Our anatomical pathology service believes that small-bowel biopsies from Indigenous patients from remote communities do show a subtle increase in inflammatory cells (Dr M Jagusch, Anatomical Pathology Department, Cairns Base Hospital, personal communication), perhaps representing the subclinical tropical enteropathy described above. The relatively high incidence of gastrointestinal infections in these Indigenous communities may explain these histological findings and, by initiating the pathological process, may predispose these populations to tropical sprue. Dramatic presentations like those we describe are uncommon, but doctors working with remote Indigenous populations frequently encounter patients with evidence of milder nutritional deficiency. While there are many factors to consider in the assessment of poor nutrition in this population, and tropical sprue is a diagnosis of exclusion, the availability of a cheap, safe and rapidly effective treatment mandates that the diagnosis be considered in the appropriate clinical situation. Lessons from practice Tropical sprue should be considered in patients presenting with chronic weight loss, diarrhoea and nutritional deficiency. While the aetiology of the disease is by no means clear, the high rates of gastrointestinal infection in remote Indigenous communities may predispose this group to the condition. Tetracycline and folic acid therapy can be rapidly and dramatically effective, although the tetracycline course should continue for 3–6 months.

Joshua P Hanson MB BS, DTM

To the Editor: We read with interest the recent cross-sectional survey of Helicobacter pylori infection in remote and urban Aboriginal populations in Western Australia.1 The prevalence of H. pylori was shown to be consistent with that in developing countries. This is not surprising, given the high prevalence of diseases such as chronic suppurative otitis media, rheumatic fever, scabies, and tuberculosis affecting Aboriginal peoples — all of which relate to poverty and overcrowding.2 Aboriginal and Torres Strait Islander people have a massively disproportionate share of the overcrowded households in Australia. In the 2001 Census, Aboriginal people in WA accounted for 53% and 93% of the two-bedroom and three-bedroom households that accommodated seven to nine and ten or more people, respectively (from 3% of the population).3 Windsor et al speculate that high H. pylori infection rates may be the result of children not wearing nappies and of poor personal hygiene — even though these matters (and housing standards, the presence of functional washing facilities and the degree of overcrowding) were not investigated. We are concerned that such conclusions reflect negatively on the Aboriginal population who took part in the survey in good faith with good will. Moreover, it is misleading to suggest that H. pylori is a cause of poor growth among Aboriginal children. There is insufficient evidence to support screening for H. pylori infection in children, as no studies have demonstrated that treating H. pylori infections improves their growth. It is expected that “some of the participants who tested positive for H. pylori have asked to be treated with antibiotics”. We assign considerable importance to research protocols that adhere to the criterion “no research proceeds without service”.4 Did those who were H. pylori positive with dyspepsia, a history of peptic ulcer complications, or a family history of gastric cancer5 receive treatment? Windsor et al do not describe what follow-up their survey participants received. Without clarity on this point (especially appropriate advice to those who were asymptomatic), we wonder what negative impacts a positive H. pylori finding had on participants’ social and emotional well-being. We are concerned by the authors’ anthropological musings: “Indigenous people may have their own H. pylori strains”. Given Australia’s heterogeneous Indigenous population, this potential research question is of no strategic relevance.4 The promotion of expanded testing for H. pylori is not supported by the evidence. Talley’s accompanying editorial prioritises a “randomised controlled trial to test the health benefits (and risks) of population-based screening and antibiotic treatment [for H. pylori] in Indigenous Australians”.5 In the absence of a clinical endpoint for an as-yet undefined health problem, there is no convincing argument for such a trial. Narrow medical answers to health problems that ignore economic and environmental solutions are not evidence-based. Both articles should have argued strongly for political commitment to these solutions to address a wide range of existing poverty-related diseases which currently affect Abori-ginal and Torres Strait Islander people on a massive scale.

Naomi R Mayers · Sophie Couzos · Richard Murray · John Daniels

In reply: We agree with Mayers and colleagues that the vexing issue of Indigenous health is a political one. We undertook this study because we thought it very strange that the prevalence of Helicobacter pylori was known in most populations on the globe, but not in Australian Indigenous people. Results from each participant in the study were forwarded to the clinicians at the test sites. These results were discussed with the participants and those who needed, or asked for, treatment received antibiotic therapy. In a previous editorial in the Journal, Mayers and Couzos state that “preventive health assessments are obviously needed earlier, given the occurrence of preventable chronic disease at younger ages and higher rates than in other Australians”.1 We agree with this, and it is to be hoped that our data will encourage further assessment and awareness of H. pylori infection in people of all ages in the Australian Indigenous community.

Barry J Marshall · Helen M Windsor

Adam P Morton,* H David McIntyre† * Endocrinologist, † Director, Endocrinology and Obstetric Medicine, Mater Hospital, Raymond Terrace, Sth Brisbane, QLD 4101. AmortonATmater.org.au To the Editor: We read with interest the article by Hussein and colleagues on their experience with thiazolidinediones (TZDs).1 These agents are only approved by the Pharmaceutical Benefits Scheme as part of dual therapy. We wish to present our experience of adding TZDs to metformin and sulfonylureas — hence, triple therapy — in patients with suboptimally controlled type 2 diabetes mellitus. The records of 28 patients (15 men, 13 women) with type 2 diabetes, for whom pioglitazone was added to maximal doses of metformin and sulfonylurea because of suboptimal control, were reviewed. Baseline patient characteristics are shown in . Mean follow-up was 9.6 months (range, 3–24 months); the average pioglitazone dose was 31.3 mg. The mean fall in the level of glycohaemoglobin (HbA1c) was 1.26% — 10 patients achieving an HbA1c level of less than 7% at last review. Four patients did not respond to therapy; none withdrew because of side effects. Eight patients whose HbA1c fell less than 0.5% after 3 months continued taking pioglitazone, achieving an average fall in HbA1c of 1.25% after a mean of 12 months follow-up. Mean weight gain was 3.35 kg (– 3.2 kg to 11.6 kg). Mean changes in HbA1c level and weight compared with baseline over 24 months are shown in . There was no correlation between these outcomes, and no baseline characteristic predicted glycaemic response. Six studies have reported the efficacy of TZDs in triple therapy (), and show a consistent fall in HbA1c level at the expense of weight gain, with a low rate of withdrawals because of adverse effects. Our findings were similar to those of these previous reports in terms of glycaemic response and low rate of side effects. The much higher rate of side effects reported by Hussein et al1 is likely to be the result of the coprescription of TZDs with insulin in 64% of patients in their study. While fluid retention has been reported in up to 5% of patients taking TZDs as monotherapy or in combination with oral hypoglycaemics, 15% of patients using TZDs with insulin may develop significant oedema. Most reports describing precipitation of cardiac failure with TZDs have been in patients using combination therapy with insulin. It would be interesting to know what proportion of the patients who developed peripheral and pulmonary oedema in the study by Hussein et al1 were also receiving insulin. One prospective randomised trial comparing the addition of pioglitazone and bedtime insulin to maximal metformin and sulfonylurea found similar efficacy in improving glucose control, but less hypoglycaemia and improved high density lipoprotein cholesterol levels with pioglitazone.3 A study of the long-term efficacy of triple therapy found 26 of 35 patients (74%) had good control after a mean follow-up of 37 months, their HbA1c level having fallen from 8.7% to 6.9%.8 In conclusion, the experience of our unit and the published literature is that TZDs are efficacious in improving suboptimal diabetic control in patients on maximal doses of metformin and sulfonylurea. Eight individuals in our group had a significant improvement in control subsequent to minimal response after the initial 3 months of treatment, suggesting a longer trial of TZDs should be employed before classifying patients as non-responders. It is to be hoped that the regulatory authorities will allow the use of TZDs in triple therapy. 1 Characteristics of our 28 patients at baseline Variable Mean (range) Age (years) 57.4 (31–74) Weight (kg) 96.3 (56–137) Body mass index (kg/m2) 34.6 (24–50.3) Duration of diabetes (years) 11 years (1–48) Glycohaemoglobin (HbA1c) level (%) 9.0 (7.1–10.4) 2 Changes in glycohaemoglobin (HbA1c) level and weight compared with baseline values 3 Studies of thiazolidinediones added to maximal dose metformin and sulfonylurea Variable Roy et al2 Aljabri et al3 Dailey et al4 Kiayias et al5 Kiayias et al5 Byrne et al6 Yale et al7 Thiazolidinedione Rosiglitazone Pioglitazone Rosiglitazone Rosiglitazone* Rosiglitazone† Rosiglitazone Troglitazone Duration (weeks) 16 16 24 20 20 nr 24 No. of patients 48 30 181 19 19 24 101 Baseline body mass index (kg/m2) nr 26 32 31 31 nr 30.1 Baseline HbA1c level (%) 9.3 9.7 8.1 8.9 9 9.6 9.6 Fall in HbA1c level (%) 1.8 1.9 0.9 1.1 1.4 1.2 1.3 Weight gain (kg) nr 2.6 3 4.2 4.6 0.7 0.9 % Patients withdrawn 4.2 0 5.5 0 0 0 2 % Patients with satisfactory control (HbA1c level) 65 (< 7.5) 23 (< 7) 42 (< 7) nr nr nr 43 (< 8) HbA1c = glycohaemoglobin. nr = not reported. * 4 mg/day; † 8 mg/day.

Adam P Morton · H David McIntyre

Nirusha Arnold,* Mark McLean,† David R Chipps,† N Wah Cheung† * Advanced Endocrinology Trainee, † Endocrinologist, Centre for Diabetes and Endocrinology Research, Westmead Hospital, Hawkesbury Road, Westmead, NSW 2145. nirusha_arnoldATozemail.com.au To the Editor: It was with interest that we read the recent article on real-life experience with thiazolidinediones by Hussein et al.1 The authors noted the absence of severe liver toxicity with the newer agents, rosiglitazone and pioglitazone, in contrast to troglitazone, which was withdrawn because of cases of hepatic failure.2 However, the sample was too small to conclude that these thiazolidinediones (TZDs) carry no hepatic risk, and they endorsed the current Pharmaceutical Benefits Scheme recommendations that liver function tests (LFTs) be monitored every 2 months. We have collected similar clinic data showing that the development of significant abnormalities in results of LFTs with TZD therapy is uncommon, and in fact, there are often improvements in LFT findings. We reviewed the files of 166 patients with type 2 diabetes treated with TZDs between 1 August 2000 and 30 November 2002, with the aim of assessing their long-term effect on LFT results. Therapy was discontinued within 3 months in 26 patients. The reasons were non-compliance (7), therapy ineffective (8), weight gain (5), dyspnoea (1), peripheral oedema (1), malaise (1), dizziness (1), angio-oedema (1), and pre-existing LFT abnormality (1). We analysed data on the remaining 140 patients (see Box) treated for a mean of 188 ± 4 days with either pioglitazone (109 patients) or rosiglitazone (31 patients). All LFT results improved significantly (Box). At baseline, 90 patients had abnormal findings on LFTs. These findings normalised in 43 of these patients (including one with steatohepatitis proven on biopsy); improved in 29 patients; and were unchanged in nine patients. LFT findings deteriorated in nine patients, leading to cessation of therapy in two. Most patients with normal LFT results at baseline experienced improvements of these parameters within the normal range. Three patients developed new abnormalities in their LFT findings, and therapy was stopped in one patient, leading to resolution of LFT abnormalitites. Changes in glycohaemoglobin (HbA1c) levels correlated positively with changes in activity of alkaline phosphatase (correlation coefficient [r], 0.33; P < 0.01), aspartate aminotransferase (r, 0.27; P < 0.01) and alanine aminotransferase (r, 0.29; P < 0.01). Our findings support those of Hussein et al, that TZD therapy is usually stopped for reasons other than hepatic dysfunction. In contradistinction to early concerns about their hepatic safety, the improvements in LFT findings seen in our patients suggest that TZDs may even benefit hepatic function. In patients with diabetes, abnormal findings on LFTs are often attributed to fatty liver, which predisposes to steatohepatitis. TZDs may well alleviate or prevent steatohepatitis,3 thereby providing benefits beyond that of improved glycaemic control, and mild abnormalities in LFTs should not discourage their use in patients with diabetes. Changes in liver function and glycohaemoglobin (HbA1c) level Variable Baseline 6-month follow-up Change P* Weight (kg) 93 ± 2 96 ± 2 3 ± 0.4 < 0.001 HbA1c (%) 8.9 ± 0.1 7.9 ± 0.1 – 1.0 ± 0.1 < 0.001 Albumin (g/L) 41 ± 0.2 41 ± 0.2 – 0.5 ± 0.2 < 0.02 Bilirubin (μmol/L) 9.4 ± 0.4 8.5 ± 0.3 – 0.9 ± 0.3 < 0.003 ALP (U/L) 92 ± 2 79 ± 2 – 13 ± 2 < 0.001 GGT (U/L) 44 ± 3 31 ± 2 – 13 ± 2 < 0.001 AST (U/L) 26 ± 1 22 ± 1 – 3 ± 1 < 0.001 ALT (U/L) 33 ± 2 25 ± 1 – 8 ± 2 < 0.001 ALP = alkaline phosphatase. GGT = γ-glutamyl transferase. AST = aspartate aminotransferase. ALT = alanine aminotransferase. * Paired t test.

Nirusha Arnold · Mark McLean · David R Chipps · N Wah Cheung

Morbid obesity (defined as having a body mass index [BMI] > 40 kg/m2, or BMI > 35 kg/m2 with obesity-related comorbidities) is a medical disorder associated with increased morbidity and mortality. Management guidelines published by the National Health and Medical Research Council and by similar US and UK bodies have recommended surgery as the most effective treatment available for selected patients with morbid obesity. A recent meta-analysis of obesity surgery has documented its safety and effectiveness in resolving some of the major medical comorbidities that occur in obese patients. To date, no intervention other than surgery has proven either effective or cost-effective in treating severe obesity and its associated medical conditions. Targeting patients with metabolic complications of obesity (eg, type 2 diabetes) could lead to substantial cost savings for the public health system. Currently, Medicare pays for privately insured patients to undergo obesity surgery, while uninsured patients are denied access to surgery in public hospitals. This raises significant equity issues that should be addressed.

Michael L Talbot MB ChB, FRACS · John O Jorgensen MB BS, FRACS · Ken W Loi MB BS, FRACS

Clinical record A 35-year-old pregnant woman (gravida 4, para 3) presented in February 2004 at 35 weeks’ gestation with a 1-day history of vomiting, abdominal pain and drowsiness. She had no significant past medical history, was not taking any prescription medication, and had had an uncomplicated pregnancy up to that time. Her level of consciousness subsequently deteriorated, and she became responsive only to painful stimuli. She had a blood pressure of 190/110 mmHg, abdominal tenderness and generalised oedema. Urinalysis was positive for protein and leukocytes. Blood biochemical levels included creatinine 0.19 mmol/L (reference range [RR], 0.05–0.09 mmol/L); urea 9.2 mmol/L (RR, 2.5–8.3 mmol/L); ionised calcium 2.64 mmol/L (RR, 1.12–1.3 mmol/L); bicarbonate 32 mmol/L (RR, 22–28 mmol/L); amylase 2397 IU/L (RR, < 160 IU/L); lipase 1364 IU/L (RR, < 60 IU/L); urate 0.65 mmol/L (RR, 0.2–0.47 mmol/L); and phosphate 0.58 mmol/L (RR, 0.8–1.5 mmol/L). Abdominal ultrasound excluded acute cholecystitis. A fetal cardiotocograph showed reduced variability in fetal heart rate. Pre-eclampsia was diagnosed, and the woman underwent an urgent caesarean section 3 hours after presentation. The pre-eclampsia was managed with magnesium sulfate infusion, intravenous hydralazine and metoprolol. A 2574 g boy was delivered (Apgar score 3 at 1 minute, 9 at 10 minutes). Initially, he had a blood pH of 7.37 and serum calcium level of 4.12 mmol/L (RR, 2.1–2.6 mmol/L). Within 1 week he had made a full recovery from respiratory distress, jaundice and sepsis. Neonatal hypocalcaemia did not occur. Subsequently, a computed tomography (CT) scan of the woman’s abdomen confirmed acute pancreatitis, with an oedematous pancreas and fluid in the paracolic gutter (Box 1). Her serum parathyroid hormone (PTH) level on Day 2 was suppressed (0.66 pmol/L; RR, 1.3–6.8 pmol/L), with PTH-related protein undetectable (< 1 pmol/L). On Day 7, her serum 25-hydroxy-vitamin D level was inappropriately high (120 nmol/L; RR, 25–108 nmol/L). (On later questioning, she said she had not taken vitamin D supplements.) The level of 1,25-dihydroxyvitamin D was not measured. The serum level of angiotensin-converting enzyme was normal. The patient’s hypercalcaemia was managed by intravenous administration of saline, followed by frusemide and pamidronate 90 mg. Over the next few days, her condition improved and her blood pressure and renal function normalised. The hypercalcaemia resolved quickly — to the extent that, by Day 8, she had developed symptomatic hypocalcaemia requiring oral supplementation with calcium carbonate and calcitriol (Box 2). When her delirium had resolved, she reported a 1-month history of severe heartburn, for which she had self-medicated with antacids, taking up to 10 Rennie tablets a day (an over-the-counter [OTC] preparation containing calcium carbonate 680 mg and magnesium carbonate 80 mg per tablet) (10 tablets contain about 3 g elemental calcium). She had also been drinking up to three glasses of milk a day. In light of this history, milk-alkali syndrome was diagnosed, and the pancreatitis was attributed to hypercalcaemia. By Day 16, the pancreatitis had resolved and the calcium level normalised. She was discharged home without medication and has remained well. Milk-alkali syndrome — the triad of hypercalcaemia, metabolic alkalosis and renal insufficiency1 — is associated with ingesting large amounts of calcium and absorbable alkali. Before the advent of modern treatment for peptic ulcer disease (with H2-receptor antagonists, proton-pump inhibitors and antimicrobial therapy), excessive calcium intake related to over-the-counter (OTC) antacid preparations commonly caused hypercalcaemia. In recent years, there has been a resurgence of milk-alkali syndrome. In a series of 100 patients admitted with hypercalcaemia between 1990 and 1993, it was reported as the third most common cause (12%) of hospital admissions for hypercalcaemia after primary hyperparathyroidism and hypercalcaemia of malignancy.2 Reasons for this include the recent emphasis on calcium therapy for osteoporosis, ready availability of OTC calcium preparations, and the use of calcium carbonate rather than aluminium (as a phosphate binder) in patients with advanced chronic kidney disease. Oral calcium carbonate is now the predominant source of calcium and alkali associated with the development of milk-alkali syndrome (with or without milk intake). The amount of calcium carbonate required to be ingested per day to cause milk-alkali syndrome is reported to vary from as low as 4 g to as high as 60 g.3,4 Thus, there appears to be no direct link between the amount of calcium ingested and development of the syndrome, although reports are few.2 After as little as 1 week of treatment with calcium carbonate, patients can present with symptoms of hypercalcaemia, severe metabolic alkalosis and acute renal failure.5 Three previous cases of milk-alkali syndrome complicating pregnancy have been reported,6-8 one with associated pancreatitis. In our case, the pancreatitis resolved quickly as the calcium level fell. The parathyroid hormone (PTH) concentration was appropriately suppressed in response to the hypercalcaemia. The subsequent hypocalcaemia resulted from cessation of the high calcium ingestion together with suppression of PTH. This was compounded by the concurrent administration of pamidronate and, possibly, by calcium sequestration and fat saponification associated with the pancreatitis. Although we have attributed the pancreatitis to hypercalcaemia, it should be noted that the causal association between hypercalcaemia and pancreatitis is disputed. In a Mayo Clinic series of 1153 patients with hyperparathyroidism, the incidence of pancreatitis (1.5%) was similar to that in the general population.9 Pregnancy itself has also been independently associated with pancreatitis.10 The diagnosis of milk-alkali syndrome depends on a history of ingestion of excess calcium and absorbable alkali with exclusion of other causes of hypercalcaemia. Only a minority of exposed patients appear to develop the syndrome. Kapsner et al reported that, of 297 cardiac transplant patients receiving large doses of calcium carbonate as prophylaxis for osteoporosis, 65 patients developed hypercalcaemia but only three had milk-alkali syndrome.11 Susceptibility appears to depend upon variables such as pre-existing renal disease and concurrent medications (eg, thiazide diuretics and vitamin D supplements). Management of the syndrome includes removal of the precipitating (dietary) cause, together with rehydration. In pregnancy, additional complications may involve the fetus (suppression of fetal parathyroid function, neonatal hypocalcaemia and tetany). Our patient had experienced a fairly typical pregnancy, heartburn being a common symptom in pregnant women. An OTC preparation that she took to relieve the heartburn caused life-threatening biochemical derangements, and other pregnant women are potentially susceptible to such a course of events. A diagnosis of milk-alkali syndrome may be missed unless a detailed history is obtained, and information about OTC preparations is essential in all patients presenting with hypercalcaemia. As H2-receptor antagonists and proton-pump inhibitors are not often prescribed in pregnancy for gastro-oesophageal reflux symptoms, many women rely on OTC heartburn relief medication. It is therefore important to educate pregnant patients about OTC calcium carbonate-containing antacids. A limit of 1.2–1.5 g per day of elemental calcium (3.0–3.75 g calcium carbonate) appears safe.2 This equates roughly to six Rennie tablets, five Quick-Eze tablets, 150 mL Mylanta Plus suspension or 230 mL Gaviscon liquid. Interestingly, Gaviscon tablets and all other preparations (liquid or tablets) of Mylanta (except Mylanta Rolltabs) contain no calcium. Appropriate warning labels on all OTC calcium-containing preparations would help prevent further cases of this type. Lessons from practice Gastro-oesophageal reflux (“heartburn”) in pregnancy is common. Excessive use of over-the-counter (OTC) calcium-containing antacids to relieve this condition can lead to milk-alkali syndrome. Pregnant women need to be educated about the potential risks of milk-alkali syndrome. All patients presenting with hypercalcaemia should be asked about their use of OTC medications. A daily dose of six Rennie or five Quick-Eze tablets appears safe; alternatively, one of the many non-calcium containing antacids can be recommended. 1 Abdominal computed tomography scan at Day 2 The scan shows an oedematous pancreas (arrow) and fluid in the left paracolic gutter (arrowhead). 2 Serum calcium and albumin levels during the patient’s hospital stay

Michelle V Gordon MB BS(Hons) · P Shane Hamblin MB BS(Hons), FRACP · Lawrence P McMahon MD, BS, FRACP