Snapshots of the Literature
Liver Biopsy Findings in 501 Patients Infected with HIV
(Poles MA, et al: JAIDS 1996; 11:170): The authors from NYU performed a retrospective study of liver biopsy findings in 501 patients with HIV infection. The usual indications were abnormal liver function tests (90%), fever over two weeks (72%) and hepatomegaly (52%). Results were categorized as follows: granulomatous hepatitis--191 (37%), viral hepatitis and cirrhosis--91 (18%), neoplasia--19 (4%), infections other than Mycobacteria--14 (3%) or nonspecific findings--165 (33%). The most common diagnosis was M. avium in 17% of biopsies and M. tuberculosis in 14 (3%). Other infectious agents were rare. The most common neoplasm was lymphoma in 12 (2%). Overall, 64% of liver biopsies yielded a ahistophatological diagnosis, and 46% of these were potentially treatable conditions. The authors were careful to conclude that the study was not designed to determine when liver biopsies should be done; the utility of liver biopsy in most of these patients continues to be controversial.
Utility of Chest X-ray Screening
(Schneider RF, et al: Arch Intern Med 1996; 156:191): This is one of many studies by the Pulmonary Complications of HIV Infection Study Group. The study format was a longitudinal study of 1,065 patients in various stages of HIV infection who received chest x-rays at 0, 3, 6 and 12 months. The results showed an abnormality in 123 of 5,263 (2%) x-rays. The frequency of abnormal findings was not different among gay men and injection drug users. As expected, abnormalities were found more frequently in those with CD4 cell counts <200/mm 3 (3%), compared to those with CD4 cell counts above 200/mm 3 (1%). The primary finding was that the results of the screening x-rays were relatively useless. For tuberculosis, active disease was not detected in any of the 123 patients with an abnormal screening x-ray. X-rays in 751 with anergy failed to identify any evidence of tuberculosis. Even in the 82 patients with a positive PPD, there was only one (1%) with a deductible pulmonary lesion. During the course of the study there were 55 patients who developed a new pulmonary diagnosis. This could be detected by the prior routine screening x-ray in only 11 (20%). The authors concluded that routine chest x-rays in patients with HIV infection are not warranted. The obvious exception are those patients with a positive PPD.
Trimethoprim-sulfamethoxazole for PCP Prophylaxis
(Ioannidis JPA, et al: Arch Intern Med 1996; 156:177): The authors performed a meta-analysis of 35 clinical trials of PCP prophylaxis using TMP-SMX in a total of 6,583 patients. As expected, this drug proved superior in efficacy compared to dapsone and aerosolized pentamidine. However, there was no statistically significant survival benefit for any of these three drugs. The major reasons for discontinuation of TMP-SMX, among those with adverse reactions were: rash and/or fever--57%, nausea and/or vomiting--17%, bone marrow suppression--21% or hepatitis--3%. Among patients who were considered TMP-SMX failures, the drug had been discontinued in 51 of 62 such cases making only 11 failures in the "as treated" analysis. This observation points out the extraordinary success of this drug in preventing PCP, the deceptive results often encountered with "intent-to-treat analysis" and the potential importance and relevance of "as-treated analysis." Finally, there was no apparent dose impact on efficacy, but there was a notable impact on the frequency of side effects. These results are shown in the following Table:
Helicobacter pylori in Patients with HIV infection
(Vaira D, et al: Dig Dis Sci 1995; 40:1622): This is a study from Italy designed to evaluate the role of H. pylori in patients with HIV infection. Group 1 consisted of 210 patients at high risk for HIV infection who had typical symptoms of H. pylori (dyshpagia heartburn, dyspepsia, nausea or vomiting); 111 (53%) had positive HIV serology. Group 2 included 219 patients who were not at risk for HIV infection, but had similar symptoms. Group 3 consisted of 259 asymptomatic controls. H. pylori was found in 109 (52%) of Group 1 patients and 120 (55%) of Group 2 patients. In Group 1, this pathogen was less common in those with HIV infection: 40% vs 66%, p<001. Histologic features associated with H. pylori were similar in all patients. The conclusion is that H. pylori is somewhat less common in patients with AIDS, but disease manifestations based on clinical symptoms, histology and serologic response were essentially the same with and without HIV infection.
Pyrimethamine-clindamycin vs. Pyrimethamine-sulfadiazine Treatment of Toxoplasmosis in Patients with AIDS
(Katlama C, et al: CID 1996; 22:268): This represents a multi-center, randomized, prospective study by the European Network for Treatment of AIDS (ENTA) Toxoplasmosis Study Group. There 340 patients randomized to receive pyrimethamine (50 mg/day) combined with either clindaycin (2.4 gm/day) or sulfadiazine (4 gm/day) for six weeks followed by maintenance with pyrimethamine (25 mg/day) with either clindamycin (1.2 gm/day) or sulfadiazine (2 gm/day). The results showed that the risk of progression was 1.84 times higher for patients in the clindamycin treatment group. There was no significant difference during the acute treatment period; the difference in efficacy occurred during the maintenance phase of treatment. The rate of side effects was similar, but those treated with clindamycin required discontinuation of treatment less frequently (11% vs. 30%). The most common side effects in the clindamycin group were rash = fever in 29%, and diarrhea in 19%. Major side effects in the sulfadiazine group were skin rash and fever in 39%. The authors concluded that pyrimethamine-sulfadiazine is the most effective treatment for toxoplasmosis.
Lung Abscess in Patients with AIDS
(Furman AC, et al: CID 1996; 22:81): The authors did a retrospective analysis of lung abscess in AIDS patients at Cornell for the period 1989-1994. There were 31 patients with a mean CD4 cell count of 17/mm 3. A microbiological diagnosis was made in 28. The most common were bacterial pathogens in 20 (65%) including P. aeruginosa-11, S. pneumoniae--6, Klebsiella pneumoniae--5, S. aureus--4. Other pathogens included P. carinii--5, Aspergillus--3 and Cryptococcus neoformans--1. There were no cases of tuberculosis. Follow-up of these patients showed resolution of the lung abscess in only 11 (36%); there were 6 deaths (36%) and death due to unrelated causes in 3 (9%). This is quite different from most studies of lung abscess in terms of the prevalence of aerobic bacteria, the surprising paucity of tuberculosis and the poor response to antibiotics.
Endocarditis in Injection Drug Users: Importance of HIV Serostatus
(Pulviranti JJ, et al: CID 1996; 22-40): The authors reviewed the experience at five Chicago hospitals for cases of endocarditis from 1987 through 1990. There were 102 patients who met the criteria for endocarditis and had HIV serology; 45 were seropositive and 57 were seronegative. S. aureus was the predominant organism in both groups, accounting for 64 cases (62%). The mortality rate was 9% in HIV seronegative patients and 13% in seropositive patients. The most striking observation was that mortality correlated with CD4 cell count. Mortality rates were 56% in those with a CD4 cell count <200/mm 3, 9% in those with counts of 200-500/mm 3 and no deaths in those with CD4 cell counts >500/mm 3. The authors note that the high mortality rate in patients with advanced HIV infection may be explained in part by the absence of PCP prophylaxis.
Risk Factors for Fluconazole-Resistant Candidasis in HIV-infected Patients
(Maenza JR, et al: JID 1996; 173:219): This is a Hopkins-based case-control study. Cases included patients with oral or esophageal candida infections characterized by in vitro resistance to fluconazole and lack of clinical response to fluconazole treatment. There were 25 case-patients; the great majority had infections involving C. albicans, and the median MIC to fluconazole was 64 ug/mL. Patients with infections refractory to fluconazole had more treated episodes (3.1 vs. 1.8) lower median CD4 cell counts counts (11 vs. 71/mm 3) and longer systemic azole treatment (median of 272 vs. 14 days). The authors concluded that risk factors for fluconazole resistance were advanced immunosuppression and prolonged exposure to azoles.
This article was provided by Johns Hopkins AIDS Service. It is a part of the publication Moore News for Care Providers.