key: cord-0824404-xdblodz5
authors: Wilcox, C. M.; Mönkemüller, K. E.
title: The therapy of gastrointestinal infections associated with the acquired immunodeficiency syndrome
date: 2003-11-14
journal: Aliment Pharmacol Ther
DOI: 10.1046/j.1365-2036.1997.00159.x
sha: 042df4c0b567d767b2f926ba41de5f26c8d7cde6
doc_id: 824404
cord_uid: xdblodz5

Although there have been dramatic strides in the therapy of human immunodeficiency virus infection over the last few years, the number of infected people world‐wide is tremendous and, at least in developing countries, continues to expand. Complications which involve the gastrointestinal tract are common in these patients, because the gut is a major site for involvement by opportunistic infections and neoplasms in patients with the acquired immunodeficiency syndrome. It is important to recognize the clinical spectrum of gastrointestinal diseases, as well as the appropriate and most cost‐effective diagnostic strategies, as therapies for a number of these disorders are both widely available and highly effective. This review summarizes the major gastrointestinal infections which are seen in patients with the acquired immunodeficiency syndrome, and their treatment.

Human immunode®ciency virus (HIV) infection remains a common world-wide infectious disease. Diseases of the gastrointestinal tract are among the most frequent complications of the acquired immunode®ciency syndrome (AIDS). Oesophageal disease occurs in 30± 40% of patients, and the incidence of diarrhoea may be as high as 90%, particularly for patients with severe immunode®ciency. 1±4 The incidence of some disorders has been unfavourably altered by the widespread use of prophylaxis for Pneumocystis carinii pneumonia. 5±6 Over the last decade, greater experience with gastrointestinal disorders in these patients has resulted in a better appreciation of the spectrum of potential aetiologies, and an improved approach to evaluation and therapy. Fortunately, our therapeutic options have expanded due to the application of old drugs to new problems as well as through the development of newer agents. Nevertheless, truly effective therapy is still lacking for some opportunistic infections. This review will focus on the therapy and prophylaxis for the most common infections involving the gastrointestinal tract in HIV-infected patients (see Table 1 ).

Oropharyngeal and oesophageal candidiasis are common opportunistic infections in HIV-infected patients, and are frequently the initial manifestation of infection. Oropharyngeal candidiasis (thrush) is readily diagnosed by the characteristic white plaques which coat the buccal mucosa. Oropharyngeal candidiasis may occasionally present as diffuse erythaema without plaques or angular cheilitis. 7 It is important to differentiate oral hairy leukoplakia from thrush, given the different aetiology and therapy. Although Candida albicans is by far the most common cause of candidiasis, other nonalbicans species cause the same disease including C. krusei, C. tropicalis, C. parapsilosis and C. glabrata. 8 In SUMMARY Although there have been dramatic strides in the therapy of human immunode®ciency virus infection over the last few years, the number of infected people world-wide is tremendous and, at least in developing countries, continues to expand. Complications which involve the gastrointestinal tract are common in these patients, because the gut is a major site for involvement by opportunistic infections and neoplasms in patients with the acquired immunode®ciency syndrome. It is important to recognize the clinical spectrum of gastrointestinal diseases, as well as the appropriate and most cost-effective diagnostic strategies, as therapies for a number of these disorders are both widely available and highly effective. This review summarizes the major gastrointestinal infections which are seen in patients with the acquired immunode®ciency syndrome, and their treatment. general, determining the aetiologic species is unnecessary since therapy is the same. Clinically, oesophageal candidiasis may be strongly suspected in the patient with thrush, moderately severe immunode®ciency (CD4 lymphocyte count < 200/mm 3 ), and oesophageal symptoms. 9 However, thrush may be absent in one-third of patients with oesophageal candidiasis. 9 A de®nitive diagnosis rests on the identi®cation of typical yeast forms in endoscopic mucosal biopsies or oesophageal brushings. The detection of Candida does not exclude other disorders, as Candida may coexist with other oesophageal processes in 14±25% of symptomatic patients. 1 

Therapy Topical non-systemic therapy has been the mainstay of treatment for oropharyngeal candidiasis. Although nystatin is effective, clotrimazole troches have now largely replaced nystatin as the principal non-systemic agent used due to its ease of administration, palatability, negligible side-effects and drug interactions, and effectiveness. Clinical cure is seen in 65±94% of patients following 14 days of clotrimazole therapy. 13, 14 The ®rst oral systemic antifungal agent available for widespread clinical use was ketoconazole. This agent has a broad spectrum of antifungal activity which includes Coccidioides immitis, Histoplasma capsulatum and Blastomyces, has minimal ef®cacy for Cryptococcus and lacks activity against Mucormycosis and Sporothrix. 15 It requires an acid milieu for absorption. Fluconazole, released in 1990, has a broader spectrum of antifungal activity than ketoconazole, including Cryptococcus. Excellent absorption, regardless of gastric pH, and long half-life (%30 h), make this an ideal oral antifungal agent. Importantly, the pharmacokinetics of¯uconazole are similar in HIV-infected patients as compared to other immunocompromised hosts. 16 Fluconazole is excreted solely in the urine; therefore, dose adjustments may be required in patients with renal insuf®ciency. Itraconazole is the newest oral systemic antifungal agent. Like¯uconazole, it has a broad spectrum of antifungal activity, prolonged half-life and is well tolerated. Similar to ketoconazole, elevations in gastric pH retard absorption. 17 Contrasts and comparisons between these agents are listed in Table 2 The drugs marked with * have been used for prophylaxis. candidiasis in HIV-infected patients. Fluconazole has been shown to be superior to clotrimazole troches for oropharyngeal candidiasis. 13, 14 The largest comparative trial 14 suggested overall equivalency, although uconazole-treated patients were more likely to have mycologic cure (65% vs. 48%) and had a lower rate of relapse at 2 weeks (82% vs. 50%); this relapse difference was equal at 1 month follow-up. Single-dose therapy with¯uconazole 150 mg is as effective as a 7-day treatment course. 18 Unlike oropharyngeal disease, non-systemic therapy (e.g. nystatin) is largely ineffective for oesophageal candidiasis. At most centres,¯uconazole has become the drug of choice for oesophageal candidiasis in AIDS. Randomized trials have shown¯uconazole to be superior to ketoconazole for both oropharyngeal 19, 20 and oesophageal candidiasis. 21 In a prospective randomized trial, Laine and colleagues 21 compared ketoconazole 200 mg/day to¯uconazole 100 mg/day in 143 patients with AIDS and oesophageal candidiasis. Endoscopic cure and symptom resolution were found in 91% and 85%, respectively, of¯uconazole-treated patients, compared to 52% and 65%, respectively, for patients randomized to ketoconazole. Clinical cure paralleled endoscopic cure. Both drugs were well tolerated with very few side-effects. The response rate of oesophageal candidiasis with¯uconazole tends to be very rapid, with most patients experiencing signi®cant clinical improvement by 3±5 days. 12, 22 Itraconazole 200 mg/day was found to be equivalent to ketoconazole 200 mg b.d. for oropharyngeal and oesophageal candidiasis. 23 Another study 24 found greater endoscopic cure at 3 weeks (75% vs. 38%) with¯uconazole 100 mg b.d. compared to itraconazole 100 mg b.d., although no differences in short-term clinical remission (78% vs. 73%) were demonstrated. In the largest study reported to date, Barbaro et al. 25 randomized 2213 AIDS patients with a ®rst episode of Candida oesophagitis to¯uconazole or itraconazole. Clinical cure was achieved in 81% of¯uconazole-treated patients compared to 75% of itraconazole-treated patients (P < 0.001), although there was no difference at the end of the follow-up period (96%). Approximately 25% of patients in both groups required an increase in dosage at 2 weeks.

Oral suspension forms of both¯uconazole and itraconazole have recently been developed, and their ef®cacy appears similar to pill forms. 26 Preliminary results of comparative trials between¯uconazole pills and itraconazole suspension suggest equivalency. 27 The suspension form is an attractive alternative to pills for patients with severe oesophageal symptoms, as well as for use in children. The additional cost of¯uconazole and itraconazole over ketoconazole (%2±3 times more expensive) may be a consideration, especially when the disease is mild and non-life-threatening (e.g. mild thrush). Monotherapy with¯ucytosine at 100 mg/kg/ day orally was found to be inferior to¯uconazole for the treatment of oesophageal candidiasis. 28 Amphotericin B is highly effective against all Candida species. Because of its toxicity, this drug is used almost exclusively in patients with thrush and/or oesophageal candidiasis with clinical and/or microbiologically resistant Candida. Although large studies evaluating the ef®cacy of amphotericin B for oesophageal candidasis in AIDS have not been conducted, clinical experience suggests this agent is highly effective. The use of intralipid with amphotericin B infusion may reduce some of the side-effects. 29, 30 Low doses of amphotericin B (0.5 mg/kg/day for 7±10 days) are usually adequate therapy for oropharyngeal and oesophageal candidiasis. The availability of a new lipid formulation (liposomal amphotericin B) reduces the side-effects and toxicity and maintains ef®cacy, although at a markedly increased cost. 31 A suspension form of amphotericin B for oral use is now commercially available, and may ®nd its greatest utility in patients with resistant thrush. 32

An emerging problem of signi®cant concern is the development of azole resistance. A number of studies of both oropharyngeal and oesophageal candidiasis have now documented clinical and microbiological resistance to these azole agents, most notably¯uconazole. 33, 34 Testing of Candida isolates from patients with AIDS have found resistance in up to 33%. 35±37 The mechanism(s) for drug resistance are incompletely understood. Risk factors for resistance include prior use of azole drugs including total dosage, severe immunode®ciency (CD4 lymphocyte count < 100/mm 3 ), and recurrent episodes of candidiasis. 27±39 Some patients may harbour resistant strains without any prior azole therapy. 35 In vitro resistance correlates with clinical resistance. 38, 40±43 With prolonged azole use, there may be selection of a single resistant strain 44±46 as well as an increase in nonalbicans strains, 8 which are inherently more resistant to these drugs. 46 Despite an increase in MIC, clinical response to¯uconazole is usually maintained regardless of high-level resistance. 37 With¯uconazole resistance, the MIC to other azoles, such as itraconazole, may also increase; 36 resistance to one azole therefore suggests cross-resistance to others. 47 Although most¯uconazoleresistant strains remain susceptible to itraconazole, there is an upward shift of the MIC to itraconazole, 48 which may result in a poor clinical response. Return of sensitivity to¯uconazole has been noted following drug discontinuation. 34 It is unclear whether intermittent azole use leads to greater resistance than continued drug administration. 49 Mild drug resistance can often be overcome with increases in the dosage of the antifungal agent; up to 800 mg/day of¯uconazole have been used. If higher doses are clinically ineffective or cause side-effects, switching to another azole can be tried; if there is no response, amphotericin B is effective. Microbiological and clinical resistance to amphotericin B is very rare. Nystatin or clotrimazole may occasionally be effective in patients with oropharyngeal candidiasis and clinical resistance to systemic azole agents. Combination therapy of¯ucytosine with azoles for drug resistance has not been well studied. 50 Until further studies are available which clarify the mechanisms of drug resistance, it appears prudent to limit the use of azoles where possible.

Despite the frequency of oropharyngeal and oesophageal candidiasis in HIV-infected patients, primary prophylaxis is not widely administered, because these disorders are non-life-threatening, therapy is very effective, and there is concern that widespread use of primary prophylaxis will exacerbate the problem of drug resistance. 51 Both non-systemic 52 and systemic therapies provide effective prophylaxis. 53, 54 Whereas primary prophylaxis for Candida is rarely provided, secondary prophylaxis is commonly given, especially for patients with multiple recurrences. Fluconazole 50± 100 mg/day or 150 mg once weekly are effective prophylaxis against recurrent oropharyngeal and oesophageal candidiasis. 55, 56 The use of¯uconazole as prophylaxis has not shown a survival bene®t. An additional bene®t of chronic azole therapy is a reduction in the incidence of systemic cryptococcosis. 6 , 57

Cytomegalovirus (CMV) is one of the most common opportunistic infections in patients with AIDS. Clinical and/or autopsy evidence of CMV disease may be observed in up to 90% of these patients, 58 and antibody positivity to CMV is present in 90% of homosexual men and 70% of intravenous drug users. 59 In fact, recent studies have shown an increasing incidence of CMV disease paralleling the widespread use of prophylaxis for Pneumocystis carinii pneumonia. 5, 6 CMV disease typically occurs when immunode®ciency is severe (CD4 lymphocyte count < 100/mm 3 ); in these patients, the incidence of disease may approach 21% at 2 years. 60 Although the retina is the most common target for CMV, gastrointestinal involvement remains important because of its frequency and morbidity. The diagnosis of CMV disease is best established by identi®cation of viral cytopathic effects (inclusions) in gastrointestinal muco-sal biopsies; viral culture of biopsy specimens is less sensitive and speci®c. 61 Therapy. Treatment for gastrointestinal CMV disease is limited to intravenous therapy with ganciclovir, foscarnet, and more recently, cidofovir (Table 3) . These drugs have similar mechanisms of antiviral action, including inhibition of viral DNA polymerases. 62 In addition to the anti-CMV effects, foscarnet, and possibly ganciclovir, also have an inhibitory effect on HIV. 63 69, 70 The only placebo-controlled trial of ganciclovir, which evaluated colitis, found no clinically signi®cant differences, probably because the treatment period was only 2 weeks. 71 A randomized trial comparing ganciclovir to foscarnet in 48 AIDS patients with gastrointestinal CMV disease found similar clinical ef®cacy (73%) regardless of the location of disease (oesophagus vs. colon), with endoscopic improvement documented in over 80%. 72 Time to progression of disease was similar (13±16 weeks) regardless of the use of maintenance therapy. Side-effects occurred in half the patients in each group.

Ganciclovir has a signi®cant inhibitory effect on haematopoiesis; leukopenia occurs in up to 40% with long-term use 62 and is more severe when concomitant drugs which affect the bone marrow are given, such as azidothymidine (AZT). 73 The use of growth-stimulating factors will usually restore cell counts toward normal, permitting continued drug administration. In contrast to ganciclovir, foscarnet has no myelosuppressive effects, but causes renal insuf®ciency and disturbances in calcium, phosphorus and magnesium homeostasis. 74 Foscarnet is an anionic compound that binds divalent cations, such as calcium and phosphorus, which accounts for the metabolic effects on these cations. Adjustments in dosage based on creatinine clearance, and the intravenous administration of isotonic saline prior to and during foscarnet infusion help prevent renal insuf®ciency. 74 When symptomatic, electrolyte disturbances may be treated by oral supplementation. Coadministration of foscarnet with intravenous pentamidine should be avoided given the similar effects on calcium homeostasis. Similarly, drugs which cause renal insuf®ciency such as amphotericin B and aminoglycosides should be used cautiously when administering foscarnet. Genital ulceration in both males and females has been reported, and appears to be due to exposure of these tissues to high urinary concentrations of foscarnet. 74 The newest agent released for the treatment of CMV disease is cidofovir. This drug is a nucleotide analogue with activity against all herpes viruses; it has a very long half-life and can be given once weekly. It must be given with probenecid to prevent renal insuf®ciency. 75 To date this drug has been tested exclusively in patients with retinal disease, but future studies assessing ef®cacy for gastrointestinal CMV disease are anticipated. The ef®cacy, tolerability and cost of ganciclovir have established it as ®rst line therapy for gastrointestinal CMV disease in AIDS. Our current policy for the therapy of gastrointestinal CMV disease is to administer intravenous ganciclovir, assuming there are no major contraindications to this agent such as pancytopenia. We treat with induction doses for 2±4 weeks, depending on the clinical and endoscopic response. In our experience, oesophageal disease tends to respond more rapidly than does colonic disease. Endoscopic re-examination following therapy is important for those patients with persistent symptoms. Ophthalmological examination is mandatory at the time of diagnosis in all patients to exclude retinal disease; long-term drug administration will be necessary in these patients. Failure to respond to ganciclovir may be the result of low serum levels 76 or drug resistance. 62 For patients with major contraindications to ganciclovir or in whom bone marrow stimulating factors are ineffective, foscarnet should be given. If a patient does not respond to ganciclovir (and the diagnosis is well established), foscarnet is usually effective. 68, 77 Combination therapy of foscarnet and ganciclovir appears to be effective for ganciclovir failures and has been used as primary therapy. 78, 79 (See Table 3 for comparisons between ganciclovir and foscarnet.)

If retinal disease is absent and a complete symptomatic and endoscopic response is documented following induction therapy, we stop therapy and look for recurrent symptoms. The relapse rate for oesophageal and colonic disease is similar (30±50%). 68, 72 For those patients with frequent relapses of gastrointestinal disease, long-term once-daily maintenance intravenous administration is appropriate. There are no data regarding the ef®cacy of oral ganciclovir for either maintenance therapy or treatment of acute gastrointestinal disease.

Prophylaxis. Oral ganciclovir has been recommended for primary prophylaxis when immunode®ciency is severe (CD4 count < 100/mm 3 ), given the incidence of disease in this setting. The oral absorption of ganciclovir is poor, with a bioavailabilty of 6±9%. 62 The half-life of oral ganciclovir (3±7.3 h) is similar to intravenous administration (5 mg/kg); however, serum drug levels are much less (0.5 l g/mL) than those achieved with intravenous therapy (4.5±10 mg/mL). 62 Oral ganciclovir occasionally causes bone marrow suppression. Randomized placebo-controlled studies of oral ganciclovir for primary prophylaxis have demonstrated a reduction in the incidence of retinal and gastrointestinal involvement. In one study, 80 the incidence of CMV disease at 1 year was 14% in the treated group compared to 26% in the placebo group. The number of patients developing gastrointestinal CMV was low in both groups. The development of resistance is a concern with long-term use of oral ganciclovir. Nevertheless, studies to date have shown resistance to be rare with either oral or intravenous administration. 62 Resistance to foscarnet has not been extensively studied.

Herpes simplex virus (HSV) is an uncommon gastrointestinal pathogen in HIV-infected patients, in contrast to other immunocompromised patients. Since HSV primarily infects squamous mucosa, oropharyngeal, oesophageal and perianal involvement are the most common sites of disease. Oropharyngeal disease may be isolated, or may occur in association with oesophageal disease. In a large prospective study of 100 HIV-infected patients with ulcerative oesophagitis, HSV oesophagitis was identi®ed in only 5%, whereas the prevalence of CMV disease was almost 50%. 81 Like CMV, the incidence of HSV disease increases as immunode®ciency worsens, with the greatest frequency occurring when the CD4 count is < 100/mm 3 . 82 Mucosal biopsy is the most speci®c diagnostic method; cytology and culture also appear to be reliable techniques.

For the patient with mild to moderate disease who is able to tolerate pills, oral administration of acyclovir 15±30 mg/kg/day is effective. 81, 83 Absorption of oral acyclovir is inconsistent and may be <30%; 84 thus for patients with more severe disease, a higher dose may be required. Intravenous administration should be used when severe odynophagia limits oral intake or when the patient has not responded to high-dose oral therapy. In general, resistance is de®ned clinically as progression of disease despite acyclovir therapy. Although uncommon, acyclovir resistance has been documented, and is usually caused by a mutation in the thymidine kinase gene. 85 The incidence of acyclovir resistance during long-term therapy is unknown. In patients who are not immunocompromised, in vitro resistance can usually be overcome by increasing doses, whereas immunocompromised patients usually require an alternative therapy. 86 Ganciclovir, foscarnet and famciclovir are also highly effective against HSV; foscarnet is the preferred therapy for acyclovir resistance. 87 Foscarnet resistance has also been reported following long-term therapy for HSV. 88 The relapse rate of gastrointestinal disease is not well de®ned, but is probably similar to CMV. Primary prophylaxis is not currently recommended; secondary prophylaxis is usually provided for patients with genital disease or those with frequent relapses of oropharyngeal or oesophageal disease.

A number of other viral pathogens have been reported to involve the gastrointestinal tract in patients with AIDS. Epstein±Barr virus has been described as a cause of oesophageal ulcer. 89 Rotavirus has been linked to both acute and chronic diarrhoea. 90 Adenovirus has been reported to cause diarrhoea and colitis. 90±92 Several unusual viruses have been identi®ed in HIVinfected patients with chronic diarrhoea, including astrovirus and picobirna virus 93 and coronavirus. 94 Although the true incidence of these viruses as gastrointestinal pathogens is unknown, it is probably low, and therapy is not currently available.

The emergence of the AIDS epidemic has greatly expanded the spectrum of gastrointestinal protozoal infections. Unlike fungi and viruses, these pathogens primarily involve the small intestine. As such, they play the greatest role as causative agents of diarrhoea. These pathogens may also infect biliary epithelium, resulting in the AIDS cholangiopathy syndrome. 95, 96 Cryptosporidia In most series, Cryptosporidium parvum is the most common protozoal infection causing diarrhoea, identi-®ed in up to 11% of symptomatic patients. 97 Although a cause of acute diarrhoea, cryptosporidiosis is most commonly found in HIV-infected patients with chronic diarrhoea. Outbreaks of cryptosporidiosis are well described in both immunocompetent and immunode®cient hosts and result from contamination of public water sources. 98, 99 In contrast to immunocompetent patients with cryptosporidiosis, where spontaneous cure is uniform, the natural history is much more variable in HIV-infected patients. 100 This variablity is due to the effect of immunode®ciency, as patients with CD4 lymphocyte counts > 180/mm 3 usually have a selflimited illness, 101 whereas in patients with a CD4 count < 50/mm 3 , the disease is often devastating, resulting in severe malabsorption, electrolyte disturbances, dehydration and weight loss, with a median survival of <12 weeks. 100 The pathogenesis of mucosal injury is poorly understood, although the degree of architectural distortion (villous atrophy) and in¯ammation are related to the parasite burden. 102 Therapy. Over 60 therapies have been used for the treatment of intestinal cryptosporidiosis, most without success. 103 Several case reports suggest that immune reconstitution, either through potent antiretroviral therapy 104 or improvements in nutritional status, 105 may result in a clinical remission. A novel therapy includes the use of bovine colostrum. This consists of a concentrate of immunoglobulin prepared from bovine colostrum following immunization with cryptosporidial antigens; 106, 107 case reports have demonstrated clinical improvement in 50% of patients following the use of bovine colostrum. The results of an open-label trial 108 found a reduction in stool frequency and weight in patients receiving the powder (but not pill) formulation of bovine immunoglobulin, although less than half of the patients had a 50% reduction in stool weight or clearance of the pathogen from the stool. 108 Letrazuril, a drug with activity against coccidia, was found to have some ef®cacy in two studies. 109, 110 The most effective agent currently available for the treatment of cryptosporidiosis is paromomycin. This non-absorbable 111 oral aminoglycoside agent has previously been used for the treatment of other parasitic diseases, but its mechanism of action for cryptosporidiosis is unknown. In vitro models using a human enterocyte cell line suggest an antimicrobial effect, 112 whereas a study using cryptosporidia in culture did not demonstrate sensitivity with this agent. 113 Case reports and small open-label trials of paromomycin have documented response rates up to 100%. 103, 114±116 In a study of 35 patients, a complete response was seen in 20% of patients, with a partial response observed in an additional 43%; 116 responders had higher preserved immune function as assessed by CD4 count. In a prospective open-label trial of 24 patients, 22 (92%) had a clinical response, with a complete remission observed in 18. 115 In the 22 responders, clearance of the organisms was noted on follow-up stool studies and/or small bowel biopsy. Other studies, however, have found persistent oocyst excretion, despite clinical improvement. 114 The ef®cacy of paromomycin was best shown in a randomized double-blind placebo-controlled crossover trial of 10 patients, where both stool frequency and oocyst excretion were signi®cantly reduced with paromomycin as compared to placebo. 117 Although the available literature supports the use of this agent, in our experience and others, 99 those patients with the most severe disease (and most severe immunode®ciency) are the least likely to respond. If the CD4 count is >200/mm 3 , drug discontinuation after clinical cure is appropriate with close follow-up. For those patients with a CD4 count < 100/mm 3 in whom therapy is clinically effective, long-term administration is required to prevent relapse. Nevertheless, despite continued therapy, relapse may still occur. 115 

Over the last decade, microsporidia have gained global attention as gastrointestinal pathogens in both immu-node®cient and immunocompetent patients. Intense investigation has demonstrated that these parasites are common intestinal and biliary pathogens in patients with AIDS. 118±120 Microsporidia are a heterogeneous group of obligate intracellular spore-forming (coccidian) protozoa which may involve a variety of organ systems causing either localized or disseminated disease. 121, 122 The environmental source and mode of transmission of these pathogens are unknown. Six microsporidial genera have been linked to human disease; gastrointestinal disease has been reported from only two of these, Enterocytozoon bienusi and Encephalitozoon intestinalis; E. bienusi is the cause of most cases of gastrointestinal disease. 121 Coinfection with these two microsporidia or with other pathogens has been reported. 123 In some studies of HIV-infected patients, microsporidia are the most commonly identi®ed pathogen. Kotler and Orenstein 119 found microsporidia in 39% of AIDS patients undergoing gastrointestinal evaluation for diarrhoea. This high prevalence is probably not related to an increasing incidence of disease, but rather to greater recognition and improved diagnostic testing. Although electron microscopy of small bowel biopsies is considered the gold standard for diagnosis, recent studies have shown haematoxylin and eosin, brown-brenn, Giemsa, or modi®ed trichrome staining of small bowel biopsies to have sensitivities of 77±83% with speci®cities approaching 100%; 121, 124 large comparative trials of stool testing with small bowel biopsy are lacking. 125 Immunouorescent stains are being developed and are likely to provide additional sensitivity over current stool testing methods.

Therapy. Treatments for microsporidia are variably effective. Initial studies of metronidazole demonstrated some ef®cacy, 126 although our experience, as well as others, 127 has shown this agent to be largely ineffective. Atovaquone showed some ef®cacy in a small open-label trial. 128 Albendazole, an anti-helminthic drug, has shown promise in open-label trials, 129±131 with response rates of %50%. Despite clinical improvement, the organisms persist in the stool and on small bowel biopsy. 130 In contrast, studies of patients with E. intestinalis show response rates to albendazole of 66±100%, with some patients having clearance of the organism, 127, 130 and no relapse. 130 With the recognition that two microsporidial species involve the bowel, it has become clear that albendazole is highly effective for E. intestinalis but largely ineffective for E. bienusi. This response difference emphasizes the importance of a species-speci®c diagnosis of intestinal microsporidiosis. Currently, albendazole is only available on a compassionate basis from SmithKline. Prophylactic studies for microsporidia have not been performed and are unlikely to be initiated given the variable geographical prevalence and lack of a widely available effective therapy for E. bienusi.

Isospora belli is a rare gastrointestinal pathogen in HIVinfected patients in the US, whereas it is endemic in many developing countries such as Haiti, 122 and is a major cause of chronic diarrhoea. As with other protozoa, it is primarily a small bowel pathogen. The diagnosis is best established by modi®ed acid-fast stool staining; 122 small bowel biopsy may also be diagnostic. In contrast to cryptosporidia and microsporidia, effective therapy is available. Trimethoprim±sulphamethoxazole results in a cure in most patients. The relapse rate is unknown. The widespread use of trimethoprim± sulphamethoxazole prophylaxis for Pneumocystis carinii may be one explanation for the low incidence of this infection in developed countries.

Cyclospora, another coccidian protozoa, have recently been recognized throughout the world as gastrointestinal pathogens both in immunocompetent patients and patients with AIDS. 131, 132 The prevalence of cyclospora in both developed and developing countries is unknown. A number of similarities exist in the microbiology, epidemiology and clinical expression of cyclospora and cryptosporidia. Cyclospora have a similar morphological appearance to cryptosporidia, although larger in size (8± 10 lm vs. 4±6 lm, respectively). 131 These pathogens are dif®cult to appreciate on routine microscopy of small bowel biopsies, although electron microscopy is often diagnostic. Since trimethoprim±sulphamethoxazole is a highly effective therapy, 132 the frequency of cyclospora in developed countries is likely to be low.

Giardia species, including Giardia lamblia, have no increased prevalence in HIV-infected patients, and the clinical presentation and diagnostic methods are also similar to HIV-seronegative patients. It is well recognized that multiple stool tests obtained on different days may be required for diagnosis as intestinal shedding is sporadic. 133 Light microscopic detection of giardia cysts and, less frequently, trophozoites, continues to be the mainstay of diagnosis. Fresh stool specimens should be examined or ®xed with polyvinyl alcohol formalin and then stained with trichrome or iron haematoxylin. Cyst detection can be improved by the use of immuno¯uores-cent antibody to cyst protein. Although still not widely used in routine diagnostic laboratories, faecal immunouorescent tests have shown promising results. Small bowel aspiration and biopsy may be diagnostic when stool testing is negative. Therapy with metronidazole (500 mg b.d. for 5±7 days) is highly effective, resulting in clinical cure. In the patient with historical features and clinical ®ndings compatible with giardiasis such as dyspepsia, crampy abdominal pain, borborygmi and watery diarrhoea, an empirical trial of metronidazole is appropriate particularly if initial stool testing is negative.

Like giardia, HIV-infected patients do not appear to have an increased suspectibility to amoeba. 134 Amoeba are frequently found in routine stool studies from asymptomatic and symptomatic homosexual men; 135±138 however, amoebic colitis is distinctly uncommon. In these patients, non-pathogenic amoeba (non-pathogenic zymodemes) including Entamoeba dispar, are likely, as they are indistingushable by light microscopy from pathogenic amoeba. 139 Other non-pathogenic amoeba such as Entamoeba hartmanii and Entamoeba coli have also been commonly identi®ed on stool testing of homosexual HIV-infected men with diarrhoea. 136 It may be expected that colonization, even with nonpathogenic strains, might cause signi®cant disease in immunocompromised patients. Instead, a benign clinical course has been found. A number of symptomatic patients in whom Entamoeba were identi®ed had other potential pathogens, suggesting that a search for other causes is always appropriate in a symptomatic HIVinfected patient with diarrhoea and amoebic cysts. In addition, despite clearance of these protozoa from the stool, treatment has not been shown to reliably cure diarrhoea, suggesting that in most patients these do not represent pathogens (e.g. Entamoeba dispar). It is interesting to note that in the developing world, a higher percentage of asymptomatic infection may be due to pathogenic organisms. Metronidazole (750 mg t.d.s. for 10±14 days) is highly effective for E. histolytica, and relapse is rare. Because metronidazole is not an effective agent for cysts, use of a luminal acting agent to eradicate intestinal colonization is recommended for those with invasive amoebiasis. Three major luminal agents are available: iodoquinol, diloxanide furoate and paromomycin. All have ef®cacy rates of 85±95% for the eradication of cyst passage.

With the widespread use of Pneumocystis carinii prophylaxis, mycobacteria, have emerged as increasingly important pathogens in AIDS. 5, 6 Patients with a prior AIDS-de®ning illness have an incidence of Mycobacterium avium complex (MAC) of 23% at one year, 140 but 39% in those with a CD4 count < 10/mm 3 . 141, 142 Mycobacterium tuberculosis may complicate HIV disease at any stage of immunode®ciency, whereas MAC (formally termed Mycobacterium avium intracellulare) is only seen in patients with severe immunode®ciency. Indeed, the mean CD4 lymphocyte count in patients with MAC is 60/mm 2 . 140, 143 Although Mycobacterium tuberculosis can present in an atypical fashion in HIV-infected patients, gastrointestinal involvement remains rare, especially in developed countries. 144 MAC has rarely been reported to involve the oesophagus, biliary tree and colon; small intestinal disease is the most common site of luminal gastrointestinal involvement. 145 The pathogenesis of MAC is believed to be ingestion of mycobacteria with subsequent small intestinal infection, followed by widespread dissemination. 146 Nevertheless, autopsy studies of AIDS patients with MAC bacteremia may fail to identify any foci of disease in up to 30% of patients. 147 The liver and spleen are the most common sites for dissemination. 148 Small bowel involvement is often diffuse. Massive in®ltration of the small bowel, mimicking Whipple's disease, has been described, and may account for the severe malabsorption seen in some patients. 149 Although diarrhoea is common, systemic symptoms and signs of fever and wasting often dominate the clinical presentation. Positive blood cultures establish the diagnosis of disseminated MAC, but does not prove active gastrointestinal disease. In those with suspected disease, blood cultures may be negative, necessitating repetitive cultures, bone marrow biopsy or empirical therapy. The presence of a positive stool culture suggests, but does not prove, gastrointestinal involvement; stool culture positivity is a marker for subsequent disseminated disease. 150, 151 Therapy. Therapy for MAC has improved substantially over the last decade. Previously used multi-drug regimens were poorly tolerated, associated with signi-®cant side-effects, and had low ef®cacy. 152 More recently, dual therapy with clarithromycin and ethambutol has been shown to reduce bacterial load and provide clinical bene®t. 153 In general, single-agent therapy is insuf®cient and frequently leads to resistance. 154 Clarithromycin appears to be the most effective single agent; ethambutol is also effective. 154, 155 A combination of rifabutin, ethambutol and clarithromycin was shown to be superior to rifampin, ethambutol, clofazamine and cipro¯oxacin; 155 bacteremia was cleared at 1 month in 78% of those receiving three drugs as compared to 40% in the fourdrug regimen; survival was also signi®cantly greater with the three-drug regimen (8.6 vs. 5.2 months, respectively). Relapse at 16 weeks was not seen in the three-drug group, suggesting the absence of resistance. Lifelong therapy is often given if a patient responds, because clearance of bacteria does not prove cure, as complete eradication is probably never achieved. Depending on the clinical setting, there may be initial concern over the possibility of Mycobacterium tuberculosis in some patients. Multidrug regimens are effective for M. tuberculosis with microbiological and clinical cure observed at 9 months, provided drug resistance is not present; 156 long-term therapy is thus unnecessary.

Prophylaxis. Rifabutin was the ®rst agent studied for MAC prophylaxis. Placebo-controlled trials of this agent found a reduction in incidence of bacteremia and improvement in some clinical parameters, but no differences in survival. 157 Based on this evidence, rifabutin was recommended for AIDS patients with CD4 count < 50/mm 3 . 158 More recently, prophylaxis trials have compared clarithromycin to placebo, 159 and azithromycin to rifabutin vs. the combination of these two agents. 160 These studies show clarithromycin to be highly effective as monotherapy with a 6 month incidence of infection of 6% as compared to 16% for placebo. 159 Of the 19 patients developing infection, 11 (58%) had clarithromycin-resistant strains. In the other study, 160 azithromycin was found to be superior to rifabutin, with the combination regimen most effective. However, side-effects were signi®cantly more common with the multidrug regimen (15.3% vs. 6% vs. 2.3%). As with the previous study, 159 patients developing clinical disease while on azithromycin had developed in vitro resistance. These studies in combination suggest that the preferred therapy for prophylaxis should consist of a macrolide antibiotic, probably clarithromycin 500 mg b.d. Because of side-effects and drug interactions with dual therapy including rifabutin, monotherapy is appropriate for prophylaxis. Nevertheless, resistance commonly develops.

Unusual presentations of common bacterial diseases became apparent early in the AIDS epidemic where Salmonella sp. 161 or Campylobacter sp. bacteremia 162 were reported as initial manifestations of AIDS. Bacteria were frequently identi®ed in earlier studies of diarrhoea in HIV-infected patients. 163, 164 Currently, the prevalence of these infections as causes of diarrhoea are not well known, although are probably lower than in the past given the use of trimethoprim±sulphamethoxazole for prophylaxis. The spectrum of bacterial causes of diarrhoea and clinical presentation in AIDS are similar to immunocompetent patients. Blood and/or stool cultures are usually diagnostic; blood cultures may be positive when stool cultures are negative. 165 Colitis may be identi®ed by¯exible sigmoidoscopy; mucosal biopsies should be performed in severely immunocompromised patients as CMV colitis may appear endoscopically similar. The role of enteroadherent E. coli as a cause of diarrhoea is unknown. 166 Clostridium dif®cile colitis is an important cause of diarrhoea in HIV-infected patients. A high frequency of C. dif®cile colitis would be anticipated in these patients, given the prevalence of antibiotic use and frequent hospitalizationsÐboth factors which have been linked to C. dif®cile disease. 167, 168 In the appropriate clinical setting, detection of C. dif®cile toxin is diagnostic. Faecal leucocytes are usually present (60%) and are an important clue to the diagnosis. 168 Flexible sigmoidoscopy is warranted in the patient in whom the disease is suspected but stool toxin is negative. The clinical presentation and response to therapy of C. dif®cile colitis are no different in HIV-infected as compared to uninfected patients. 169 Metronidazole, which can be administered either orally or intravenously, represents ®rst line therapy. Vancomycin should be reserved for those patients with contraindication to or failure with metronidazole or when the disease is lifethreatening; this agent is only effective when administered orally. Clinical cure can be obtained in essentially all patients. The relapse rate appears to be similar in HIV-infected as compared to uninfected patients. 169 

An important entity not clearly linked to a speci®c infection is the HIV-associated idiopathic oesophageal ulcer (IEU). These lesions can present at the time of seroconversion, 170 although typically occur when im-munode®ciency is severe; the median CD4 count in these patients is <50/mm 3 . 80 Several studies have identi®ed HIV-infected in¯ammatory cells in the ulcer base of these lesions, suggesting an aetiological role for HIV. 171, 172 However, HIV has not been identi®ed in oesophageal squamous mucosa but rather in in¯ammatory cells, and has been found in HIV-infected patients with oesophageal diseases other than IEU. 173, 174 These lesions present similarly to ulcerative oesophagitis from other causes; severe odynophagia is almost uniformly present. IEU are almost as common as CMV oesophagitis in patients with AIDS, comprising %40% of oesophageal ulcers in these patients. 81 The diagnosis is one of exclusion; CMV oesophagitis and IEU are indistinguishable clinically, radiographically and endoscopically. 175 Therapy. Treatment of HIV-associated IEU is rewarding. Prospective studies have documented healing rates of over 90% with oral corticosteroids. 176 The regimen most commonly employed is prednisone 40 mg/day decreasing to 10 mg/week for a 1 month treatment course. 176 Shorter courses of therapy may be effective for smaller ulcers. Although bene®cial, intralesional injection of corticosteroids should be considered as second line therapy. 172 The side-effects of corticosteroids are well recognized; patients with AIDS may be more likely to develop CMV disease while on therapy. 177 Because oropharyngeal and/or oesophageal candidiasis may complicate steroid use and confuse the therapeutic response, we routinely use short courses of azole therapy with prednisone. The response to corticosteroids is rapid, with most patients experiencing signi®cant pain relief within days. 176 Although not as well studied, thalidomide also appears to be highly effective for IEU. 178, 179 Its mechanism of action is unknown, but it has been suggested that it is an inhibitor of tumour necrosis factor-a production. 180 In doses of 200± 300 mg/day, thalidomide has been documented to result in a clinical response rate and endoscopic cure in over 90% of treated patients. 178, 179 Thalidomide is well tolerated, with the main side-effect being somnolence; administration of the drug at bedtime tends to overcome this. Peripheral neuropathy and skin rash have also been seen. 181 The major fear with thalidomide is the inadvertent use in the ®rst trimester of pregnancy, which consistently results in severe birth defects. Thus, most would not use this agent in women of child-bearing age, unless the patient is surgically sterile. Both prednisone and thalidomide are similarly effective for oropharyngeal apthous ulcerations. The relapse rate of IEU is %40±50% regardless of therapy. 81, 176 SYMPTOMATIC THERAPY OF DIARRHOEA For patients in whom antimicrobial therapy is ineffective or no speci®c cause for diarrhoea is found, symptomatic therapy will be necessary. When the diarrhoea is mild, bulking agents, bismuth or Kaopectate may provide relief. With more severe diarrhoea, medications to reduce intestinal transit are required. We routinely use diphenoxylate (Lomotil) in doses up to 10 tablets/day. Larger doses may cause anticholinergic side-effects, because diphenoxylate is combined with atropine in Lomotil. 182 Despite the potential for abuse, narcotic agents are also highly effective. In patients with severe diarrhoea, tincture of opium (paregoric) is very effective. It is usually provided with a dropper which provides a morphine concentration of 0.4 mg/mL. The normal dose is 5±10 mL/day in divided doses, and the dose can be titrated up to 20 mL/day. Some patients may experience somnolence or abdominal cramps; these tend to dissipate over time. Octreotide, a somatostatin analogue, is an antisecretory agent with a variety of inhibitory functions throughout the gastrointestinal tract. Initial studies in patients with AIDS found this drug to provide effective control of diarrhoea in %50%. 183±185 However, a large randomized placebo-controlled trial failed to demonstrate ef®cacy of this agent in patients with and without identi®able pathogens. 186 The drug is given subcutaneously in doses of 50±100 lg t.d.s. Side-effects include decreased biliary motility (gallstones), diabetes and steatorrhea, the latter of which may potentially exacerbate diarrhoea. Although the drug has not been clearly proved to be effective, its use may be attempted in patients with severe diarrhoea requiring hospitalization (e.g. cryptosporidia). We consider a clinical response to be a reduction in stool frequency/volume of at least 50%.

Oesophageal symptoms, their causes, treatment, and prognosis in patients with the acquired immunode®ciency syndrome

Gastrointestinal symptoms in ambulatory HIV-infected patients

for the Multicenter AIDS cohort study. Incidence of clinical AIDS conditions in a cohort of homosexual men with CD4+ cell counts < 100/mm 3

Diarrhea and quality of life in ambulatory HIV-infected patients

Clinical manifestations of AIDS in the era of pneumocystis prophylaxis

Trends in infectious diseases and cancers among persons dying of HIV infection in the United States from 1987 to

Oral manifestations of HIV infection

Oral Candida in HIV infection and AIDS: New perspectives/new approaches

Prospective evaluation of oropharyngeal ®ndings in human immunode®ciency virusinfected patients with esophageal ulcer

The causes of esophageal symptoms in human immunode®ciency virus infection: A prospective study of 110 patients

Evaluation of a technique to evaluate the underlying mucosa in patients with AIDS and severe Candida esophagitis

Fluconazole compared with endoscopy for human immunode-®ciency virus-infected patients with esophageal symptoms

Comparison of oral uconazole and clotrimazole troches as treatment of oral candidiasis in patients infected with human immunode®-ciency virus

Therapy for oropharyngeal candidiasis in HIV-infected patients: A randomized, prospective multicenter study of oral¯uconazole vs. clotrimazole troches

Oral azole drugs as systemic antifungal therapy

Pharmacokinetics and bioavailability of¯ucon-azole in patients with AIDS

Short report: the absorption of¯uconazole and itraconazole under conditions of low intragastric acidity

Single-dose vs. 7 days of uconazole treatment for oral candidiasis in human immu-node®ciency virus-infected patients: a prospective, randomized pilot study

Comparison of uconazole and ketoconazole in oropharyngeal candidiasis in AIDS

Fluconazole vs. ketoconazole in the treatment of oropharyngeal candidiasis in HIV-infected children

Fluconazole compared with ketoconazole for the treatment of Candida esophagitis in AIDS. A randomized trial

Time course of clinical response to¯uconazole for Candida oesphagitis in AIDS

Itraconazole vs. ketoconazole in the treatment of oral and oesophageal candidosis in patients infected with HIV

Comparison of therapeutic activity of¯uconazole and itraconazole in the treatment of oesophageal candidiasis in AIDS patients: a double-blind, randomized, controlled clinical study

Fluconazole vs. itraconazole for Candida esophagitis in acquired immunode-®ciency syndrome

Prospective study of¯uconazole suspension for the treatment of oesophageal candidiasis in patients with AIDS

Itraconazole oral solution (IS) compared with¯uconazole (F) for treatment of esophageal candidiasis

Fluconazole vs.¯u-cytosine in the treatment of esophageal candidiasis in AIDS patients: A double-blind, placebo-controlled study

Trial of glucose vs. fat emulsion in preparation of amphotericin for use in HIV infected patients with candidiasis

Randomized comparison of amphotericin B deoxycholate dissolved in dextrose or intralipid for the treatment of AIDS-associated cryptococcal meningitis

Lipid formulations of amphotericin B. Recent progress and future directions

Incidence and natural history of cytomegalovirus disease in patients with advanced human immunode®ciency virus disease treated with zidovudine

Frequency of positive tests for cytomegalovirus in AIDS patients: endoscopic lesions compared with normal mucosa

Antiviral effects of foscarnet and ganciclovir therapy on human immunode®ciency virus p24 antigen in patients with AIDS and cytomegalovirus retinitis

Foscarnet decreases human immunode®ciency virus RNA

Retinal and gastrointestinal disease due to cytomegalovirus in patients with the acquired immune de®ciency syndrome: Prevalence, natural history, and response to ganciclovir therapy

Ganciclovir treatment of life-or sight-threatening cytomegalovirus infection: Experience in 314 immunocompromised patients

Ganciclovir treatment of gastrointestinal infections caused by cytomegalovirus in patients with

Cytomegalovirus esophagitis in AIDS: a prospective study of clinical response to ganciclovir therapy, relapse rate, and long-term outcome

Treatment of HIV-related cytomegalovirus disease of the gastrointestinal tract with foscarnet

A reapprasial of its antiviral activity, pharmacokinetic properties and therapeutic use in immunocompromised patients with viral infections

Ganciclovir treatment of cytomegalovirus colitis in AIDS: A randomized, double-blind, placebo-controlled multicenter study

Treatment of AIDS-associated gastrointestinal cytomegalovirus infection with foscarnet and ganciclovir: a randomized comparison

Toxicity of combined ganciclovir and zidovudine for cytomegalovirus disease associated with AIDS. An AIDS Clinical Trials Group Study

Review of the toxicities of foscarnet

Anticytomegaloviral activity and safety of cidofovir in patients with human immnode®ciency virus infection and cytomegalovirus viruria

Low plasma concentrations achieved with conventional schedules of administration of ganciclovir in patients with AIDS

Foscarnet treatment of cytomegalovirus gastrointestinal infections in acquired immunode®ciency syndrome patients who have failed ganciclovir induction

Concurrent use of ganciclovir and foscarnet to treat cytomegalovirus infection in AIDS patients

Combination foscarnet and ganciclovir therapy vs. monotherapy for the treatment of relapsed cytomegalovirus retinitis in patients with AIDS. Studies of the Ocular Complications of AIDS Research Group in Collaboration with the AIDS Clinical Trials Group

Oral ganciclovir for the prevention of cytomegalovirus disease in persons with AIDS

Causes, response to therapy, and long-term outcome of esophageal ulcer in patients with human immunode®ciency virus infection

Relationship between herpes simplex virus ulceration and CD4+ cell counts in patients with HIV infection

Herpes simplex esophagitis in patients with AIDS: Report of 34 cases

A point mutation in the thymidine kinase gene is responsible for acyclovirresistance in herpes simplex virus type 2 sequential isolates

Resistance to antivirals

Management of acyclovir-resistant herpes simplex and varicella-zoster virus infections

Foscarnet-resistant herpes simplex virus infection in patients with AIDS

Epstein±Barr virus associated oesophageal ulcers in AIDS

Gastrointestinal viral infections in homosexual men who were symptomatic and seropositive for human immunode®ciency virus

Adenovirus colitis in the acquired immunode®ciency syndrome

Adenovirus infections in human immunode®ciency virus-positive patients: clinical features and molecular epidemiology

Enteric viruses and diarrhea in HIV-infected patients

Stool viruses, coinfections, and diarrhea in HIV-infected patients

Cryptosporidiosis of the biliary tract in AIDS

AIDS-related cholangitis: diagnostic features and course in 15 patients

Cryptosporidial diarrhea in AIDS and its treatment

Cryptosporidiosis: An outbreak associated with drinking water despite state-of-the-art water treatment

Biliary cryptosporidiosis in HIV-infected people after the waterborne outbreak of cryptosporidiosis in Milwaukee

The natural history of cryptosporidia diarrhea in HIV-infected patients

Cryptosporidium infection and CD4 counts

Intestinal function and injury in acquired immunode®ciency syndrome-related cryptosporidiosis

Treatment of cryptosporidiosis with paromomycin

Resolution of intestinal cryptosporidiosis after treatment of AIDS with AZT

Resolution of Cryptosporidium infection in an AIDS patient after improvement of nutritional and immune status with octreotide

Bovine colostrum immunoglobulin concentrate for cryptosporidiosis in AIDS

Bovine colostrum immunoglobulin concentrate for cryptosporidiosis in AIDS

Treatment of severe diarrhea caused by Cryptosporidium parvum with oral bovine immunoglobulin concentrate in patients with AIDS

Treatment with letrazuril of refractory cryptosporidial diarrhea complicating AIDS

A phase I study of letrazuril in AIDS-related cryptosporidiosis

Absence of systemic absorption of oral paromomycin during long-term, high-dose treatment for cryptosporidiosis in AIDS

Paromomycin inhibits Cryptosporidium infection of a human enterocyte cell line

Anticryptosporidial activity of paromomycin

Paromomycin for the treatment of cryptosporidial diarrhoea in AIDS patients

Paromomycin: an effective treatment for cryptosporidial diarrhea in patients with AIDS

Effectiveness of aminosidine (paromomysin) sulfate in chronic Cryptosporidium diarrhea in AIDS patients: An open, uncontrolled, prospective clinical trial

Paromomycin for cryptosporidiosis in AIDS: A prospective, double-blind trial

Intestinal microsporidiosis in human immunode®ciency virus-infected patients with chronic unexplained diarrhea: Prevalence and clinical and biologic features

Prevalence of intestinal microsporidiosis in HIV-infected patients referred for gastroenterological evaluation

Microsporidia infection in patients with the human immunode®ciency virus and unexplained cholangitis

Human microsporidial infections

Understanding intestinal spore-forming protozoa: Cryptosporidia, microsporidia, isospora, and cyclospora

Clinical features of microsporidiosis in patients with AIDS

Light microscopic diagnosis of microsporidiosis in patients with AIDS

Improved light-microscopical detection of microsporidia spores in stool and duodenal aspirates

Clinical signi®cance of small-intestinal microsporidiosis in HIV-1-infected individuals

Disseminated microsporidiosis due to Septata intestinalis in nine patients infected with the human immunode®ciency virus: Response to therapy with albendazole

Atovaquone is effective treatment for the symptoms of gastrointestinal microsporidiosis in HIV-1 infected patients

Treatment with albendazole for intestinal disease due to Enterocytozoon bieneusi in patients with AIDS

Disseminated microsporidiosis due to Septata intestinalis in patients with AIDS. Clinical features and response to albendazole therapy

Cyclospora: An overview

Cyclospora infection in adults infected with HIV: Clinical manifestations, treatment, and prophylaxis

The prevalence of invasive amebiasis is not increased in patients with AIDS

Entamoeba histolytica as a commensal intestinal parasite in homosexual men

Prevalence of enteric parasites in homosexual patients attending an outpatient clinic

Prevalence of enteric pathogens in homosexual men with and without acquired immunode®ciency syndrome

Prevalence of intestinal protozoans in French patients infected with HIV

Characterisation of 20 Entamoeba histolytica strains isolated from patients with HIV infection

Disseminated Mycobacterium avium complex infection: clinical identi®cation and epidemiologic trends

Incidence of Mycobacterium avium-intracellulare complex bacteremia in human immunode®ciency virus-positive patients

Incidence and natural history of Mycobacterium avium-complex infections in patients with advanced human immuno-de®ciency virus disease treated with zidovudine

Natural history of opportunistic disease in an HIV-infected urban clinical cohort

Major differences in the spectrum of gastrointestinal infections associated with AIDS in India vs. the west: An autopsy study

Atypical mycobacterial infection of the gastrointestinal tract in AIDS patients

Mycobacterium aviumintracellulare: a cause of disseminated life-threatening infection in homosexuals and drug abusers

Autopsy ®ndings in AIDS patients with Mycobacterium avium complex bacteremia

Pathologic ®ndings in disseminated Mycobacterium avium-intracellulare infection

Disseminated Mycobacterium avium-intracellulare complex in patients with AIDS

A prospective evaluation of Mycobacterium avium complex colonization of the respiratory and gastrointestinal tracts of persons with human immunode®ciency virus infection

Mycobacterium avium complex in the respiratory or gastrointestinal tract and the risk of M. avium complex bacteremia in patients with human immunode®ciency virus infection

Treatment of Mycobacterium avium complex bacteremia in AIDS with a four-drug oral regimen: Rifampin, ethambutol, clofazimine, and cipro¯oxacin

Disseminated Mycobacterium avium complex disease in patients with AIDS

Clarithromycin therapy for bacteremic Mycobacterium avium complex disease: A randomized, double-blind, dose-ranging study in patients with AIDS

A comparision of two regimens for the treatment of Mycobacterium avium complex bacteremia in AIDS: Rifabutin, ethambutol, and clarithroymcin vs. rifampin, ethambutol, clofazimine, and cipro¯oxacin

Tuberculosis of the gastrointestinal tract and peritoneum

Two controlled trials of rifabutin prophylaxis against Mycobacterium avium complex infection in AIDS

Recommendations on prophylaxis and therapy for disseminated Mycobacterium avium complex disease in patients infected with the human immunode®ciency virus

A randomized trial of clarithromycin as prophylaxis against disseminated Mycobacterium avium complex infection in patients with advanced acquired immunode®ciency syndrome

Prophylaxis against disseminated Mycobacterium avium complex with weekly azithromycin, daily rifabutin, or both

Intestinal infections in patients with the acquired immunode®ciency syndrome (AIDS): etiology and response to therapy

Campylobacter infections in HIV-infected patients: clinical and bacteriological features

Salmonella typhimurium enteritis and bacteremia in the acquired im-munode®ciency syndrome

Intestinal infections in patients with acquired immunode®ciency syndrome: A prospective study in 132 patients

Shigella in HIV infection

Chronic bacterial enteropathy in patients with AIDS

Risk factors for Clostridium dif®cile-associated diarrhoea in HIV-infected patients

Clostridium dif®cile colitis: An ef®cient clinical approach to diagnosis

Clostridium difficile-associated diarrhea in patients with HIV positivity and AIDS: a prospective controlled study

Acute HIV infection presenting with painful swallowing and esophageal ulcers

HIV-1 gp41 antigen demonstration in esophageal ulcers with acquired immunode®ciency syndrome

Chronic idiopathic esophageal ulceration in the acquired immunode®ciency syndrome

Esophageal disease in AIDS is associated with pathologic processes rather than mucosal human immunode®ciency virus type 1

Evaluation of idiopathic esophageal ulcer for human immu-node®ciency virus

Prospective endoscopic characterization of cytomegalovirus esophagitis in patients with AIDS

Comparison of two corticosteroid regimens for the treatment of idiopathic esophageal ulcerations associated with HIV infection

Treatment with corticosteroidsÐa risk factor for the development of clinical cytomegalovirus disease in AIDS

Thalidomide as treatment of refractory aphthous ulceration related to human immunode®ciency virus infection

A prospective trial of thalidomide for the treatment of HIV-associated idiopathic esophageal ulcers

Thalidomide selectively inhibits tumor necrosis factor production by stimulated human monocytes

Thalidomide in human immunode®ciency virus (HIV) patients: a review of safety considerations

Review article: antidiarrhoel pharmacology and therapeutics

Effect of octreotide on refractory AIDS-associated diarrhea. A prospective multicenter clinical trial

Octreotide therapy of large-volume refractory AIDS-associated diarrhea: a randomized controlled trial

Multicentre AIDS±Diarrhea Study Group. Octreotide therapy in AIDS-related, refractory diarrhea: results of a multicentre Canadian±European study

Multicenter trial of octreotide in patients with refractory aquired immunode®-ciency syndrome-associated diarrhea

No ®nancial support was provided for this study. REVIEW: THERAPY OF GI ASSOCIATED WITH AIDS