Faculty Bibliography

Extrapulmonary infection with Pneumocystis carinii in AIDS patients on aerosolized pentamidine is occurring more frequently. We report five patients diagnosed with gastrointestinal pneumocystosis while on aerosolized pentamidine prophylaxis and have identified infections involving the peritoneum, liver, and transverse colon, as well as stomach and duodenum. Physicians should have a high index of suspicion for extrapulmonary pneumocystosis, especially involving the gastrointestinal system, in HIV-infected patients, and early diagnosis must be pursued aggressively. The use of aerosolized pentamidine as prophylaxis for P. carinii pneumonia is not protective against gastrointestinal pneumocystosis because of inadequate systemic distribution of the drug. To our knowledge, this is the first report in a clinical journal documenting and photographing P. carinii organisms in ascitic fluid

Kaposi's sarcoma (KS) is the most common tumor among HIV-infected individuals, but its involvement in the gastrointestinal tract was reported long before the AIDS epidemic. Although most cases of gastrointestinal KS are asymptomatic, advanced lesions may occasionally result in a severe and life-threatening hemorrhage that requires immediate treatment. At the NYU Medical Center, we have seen three AIDS patients present with severe upper tract bleeding (> 8 U/48 h) from KS lesions of the antrum, fundus, and duodenum. The last patient was also bleeding from an ulcerated rectal KS lesion. Because all three patients had a coexisting thrombocytopenia (platelets < 50,000/mm3) and were poor operative risks, injection sclerotherapy was performed. All four KS lesions stopped bleeding, and three out of the four lesions decreased in size. To our knowledge, this is the first report of successfully using sclerotherapy to treat severe hemorrhage due to gastrointestinal KS

OBJECTIVE: Primarily to determine whether an intestinal microsporidian recently identified in AIDS patients disseminates from the bowel to infect other organs. DESIGN: Disseminated microsporidiosis has been reported in immunocompromised humans, but never due to Enterocytozoon bieneusi, the most common species in AIDS patients and one that evidently infects only enterocytes. In animals, dissemination follows ingestion of Encephalitozoon cuniculi spores, apparently via macrophages, and pathology occurs in, for example, kidneys and brain. A second, un-named Encephalitozoon-like intestinal microsporidia has been identified in five AIDS patients with chronic diarrhea; because it infects lamina propria macrophages, it was logical to investigate its dissemination. METHODS: Light and transmission electron microscopy were used to study urine sediment from four out of five patients with biopsy-documented small intestinal infection due to the second intestinal microsporidian. The gall bladder from one patient and autopsy specimens from an E. bieneusi-infected patient were similarly studied. RESULTS: Systemic dissemination was documented by detecting abundant spores, both free and within renal tubular and transitional cells, in the urine of two patients. Many of the lamina propria macrophages in these two patients' intestinal biopsies contained microsporidia, while those of the two negative patients either contained only Mycobacterium avium complex or only occasional parasites. The gall bladder was co-infected with this microspordian and with cytomegalovirus. At autopsy, the patient with documented enteritis due to E. bieneusi 2 years before death had disseminated microsporidiosis, not of E. bieneusi, but apparently of the second intestinal species. The microsporidian had caused severe tubulointerstitial nephritis. Parasites were also observed in non-parenchymal cells of the liver and bronchial epithelium. CONCLUSION: A newly described Encephalitozoon-like intestinal microsporidian, which causes chronic diarrhea in AIDS patients, can disseminate and cause renal pathology

OBJECTIVE: To detect and quantify HIV-1 in the liver in vivo. DESIGN: Fourteen liver biopsy samples and corresponding blood lymphocytes and monocytes from patients with AIDS were studied for HIV-1 using quantitative polymerase chain reaction (PCR). In addition, expression of HIV-1 antigen and messenger (m) RNA in 10 autopsy liver specimens was examined by immunohistochemistry and in situ hybridization. RESULTS: The amount of HIV-1 DNA in nine liver samples ranged from 850 to 27,000 copies per 10(6) cells, with mean and median values of 8150 and 3500 copies per 10(6) cells, respectively. Five other samples had no detectable HIV-1 DNA by PCR. Intracellular expression of HIV-1 antigen and mRNA was also detected in both Kupffer cells and hepatocytes by in situ studies. CONCLUSION: These findings strongly indicate that HIV-1 could replicate in the liver of a majority of patients with AIDS

We have reviewed 28 esophageal biopsies from 28 patients with the acquired immune deficiency syndrome (AIDS), over a 1-yr period. Indications for esophageal biopsy were dysphagia persisting after antifungal therapy and/or radiologic evidence of esophageal ulcer. We compared the frequency of detecting cytomegalovirus (CMV) infection on hematoxylin and eosin (H&E) stain with immunoperoxidase staining for CMV antigens. Five biopsies were positive for CMV by H&E stain and immunoperoxidase. Infected cells could often be identified in the granulation tissue and, in one severe case, in stromal papillae of the intact mucosa. Squamous cells were never positive. Thirteen biopsies consisted only of squamous epithelium, and all of these were negative by both techniques. Among the remaining 10 cases, no CMV inclusions were identified by H&E. Three of these biopsies displayed staining for viral antigens. In all cases positive by immunoperoxidase, numerous cells positive for viral antigens did not display any of the CMV-specific morphologic diagnostic criteria. Other coexisting diagnoses included candidiasis, Kaposi's sarcoma, and malignant lymphoma. We conclude 1) CMV infection of the esophagus is common in AIDS patients with esophageal ulcer or esophagitis resistant to antifungal therapy; 2) multiple infections or neoplasms may coexist; 3) since CMV apparently does not infect squamous epithelium and only rarely endothelium in stromal papillae, deep biopsies are necessary for diagnosis; and 4) immunoperoxidase staining is required for maximum diagnostic yield

We report a homosexual patient with the acquired immune deficiency syndrome (AIDS) and histopathologic evidence for cytomegalovirus (CMV) appendicitis in a patient with no prior history of CMV infection. The patient presented with right lower quadrant pain and intermittent fevers. The diagnosis of appendicitis was difficult to make in this patient because of the presumptive diagnosis of tuberculosis ileitis and the frequent presentation of abdominal pain with fever in AIDS patients. Although CMV colitis is frequently seen in AIDS patients, the prevalence of CMV appendicitis is exceedingly rare. The problems related to making a diagnosis of CMV appendicitis and the therapeutic management of CMV infections are reviewed

As part of a double-blind, placebo-controlled study of ganciclovir in cytomegalovirus (CMV) colitis, the clinical characteristics of 44 patients enrolled at one center were analyzed in detail. All were homosexual men who had CMV on colonic biopsy. CMV colitis was the index diagnosis for acquired immune deficiency syndrome (AIDS) in 11 (25%) of the 44 patients. All had diarrhea, but it was intermittent in 13 patients (30%). Bleeding was uncommon, but 35 patients (80%) were febrile (median temperature of 38.9 degrees C). Weight loss was reported by 39 patients (89%), among whom the median loss was 6.8 kg. Endoscopy revealed normal colonic mucosa but CMV on biopsy in 11 patients (25%). Colonoscopic biopsies positive for CMV were found only in the cecum in 7 (39%) of 18 patients. Most patients (54%) had received zidovudine before the diagnosis of CMV colitis. The median time to the development of CMV colitis after the diagnosis of AIDS was 16 months in those patients who had received zidovudine and 3 months in those who had not (p less than 0.02). We conclude that CMV colitis can present early in AIDS and often with such nonspecific signs as fever, intermittent diarrhea, weight loss, and hematochezia. Importantly, it can appear normal on colonoscopy and occurs frequently only in the right colon, necessitating full colonoscopy and multiple biopsies for accurate diagnosis

OBJECTIVE: To assess the toxicity, efficacy, and pharmacology of combined zidovudine and ganciclovir therapy in patients with the acquired immunodeficiency syndrome (AIDS) and serious cytomegalovirus (CMV) disease. DESIGN: Prospective, phase I multicenter trial (ACTG 004) with patients grouped by previous study drug history. SETTING: Three university-based AIDS Clinical Trials Units sponsored by the National Institute of Allergy and Infectious Diseases (NIAID). PATIENTS: Forty-one patients with AIDS-related CMV disease. Previous therapy with either zidovudine or ganciclovir was allowed. INTERVENTIONS: Patients were treated with zidovudine, 600 to 1200 mg/d; or, if on ganciclovir maintenance, ganciclovir, 5 mg/kg body weight; blood was sampled for pharmacokinetic studies. The other drug was then administered to the patient with blood sampling and, finally, the two drugs in combination were given. Patients were continued on both drug therapies with dose reduction of zidovudine only for grade 3 or 4 toxicity. MEASUREMENTS AND MAIN RESULTS: Forty patients were eligible. Hematologic toxicity was frequent, with 9 of the 10 patients requiring dose reductions for grade 3 or 4 toxicity at zidovudine doses of 1200 mg/d. With zidovudine doses of 600 mg/d, 82% experienced such hematologic toxicity. Median survival was 6 months; 10 patients developed intercurrent infection and 19, progressive CMV disease. Pharmacokinetic variables (alpha and beta half-lives, volume of distribution, clearance) were not affected in combination therapy. CONCLUSION: The combination of zidovudine and ganciclovir is poorly tolerated in patients with AIDS and serious-CMV disease, with 82% developing severe to life-threatening hematologic toxicity. Such toxicity is not a result of pharmacologic interactions, drug metabolism, or excretion