This review compared the effectiveness and safety of lipid-based formulations of amphotericin B (AmB) with conventional AmB for the treatment of systematic fungal infections. The authors concluded that lipid-based AmB reduced mortality and renal damage compared with conventional AmB. However, apparent differences between the included studies limit the strength of the review.

To compare the effectiveness and safety of lipid-based formulations of amphotericin B (AmB) with conventional AmB for the treatment of systemic fungal infections.

MEDLINE (up to November 2001), EMBASE, Biological Abstracts, AIDSLINE, Cancerlit, the Centre for Reviews and Dissemination's databases and the Cochrane Controlled Trials Register (up to August 2001) were searched for studies published in any language; the search terms were stated. Unpublished studies were identified by searching BIOSIS Biological Abstracts, Interscience Conference of Antimicrobial Agents and Chemotherapy Abstracts (1993 to 2001), American Society of Hematology abstracts (1993 to 2001), American Society for Microbiology abstracts (1993 to 2001) and Infectious Diseases Society of America abstracts (1997 to 2001).

Study designs of evaluations included in the review

Randomised controlled trials (RCTs), clinical trials, open-label trials, cohort studies, case-control studies, qualitative studies, systematic and other comprehensive reviews were eligible for inclusion. Case series, case reports, brief reviews and comments were excluded

Specific interventions included in the review

Studies that compared lipid-based formulations with conventional AmB for first-line or salvage therapy were eligible for inclusion. Studies of prophylactic treatments were excluded. The lipid-based formulations used in the included studies were AmB lipid complex, liposomal AmB and AmB colloidal dispersion. The cointerventions in some studies included potassium or sodium supplementation, and potassium sparing diuretics.

Participants included in the review

Studies of adults or children at risk of, or diagnosed with, systemic fungal infections were eligible for inclusion. Studies of patients with superficial fungal infection were excluded.

In the included studies, the characteristics of the patients varied widely over a range of proven or suspected fungal infections and underlying medical conditions (further details were reported).

Outcomes assessed in the review

Studies that assessed treatment efficacy (response rate, mortality) or safety (changes in serum creatinine, infusion-related reactions) were eligible for inclusion. The outcomes included were:

all-cause mortality;

empirical treatment efficacy, defined as the successful resolution of persistent febrile neutropenia;

antifungal treatment efficacy, defined as proven or probable clinical response;

change in serum creatinine (doubling from baseline, more than 3 times the upper limit of normal, or the number of patients with doubling, 1 mg/dL increase, or more than 50% reduction in clearance); and

infusion-related reaction (measured before or after premedication).

How were decisions on the relevance of primary studies made?

One reviewer screened abstracts and a second reviewer checked a sample for reliability.

Studies were assessed and scored for the quality of reporting, study power, internal validity (bias and confounding) and external validity (generalisability) using criteria proposed by Downs and Black. The maximum possible score was 28 points. One reviewer assessed validity and this was checked by a second reviewer.

The authors did not state how the data were extracted for the review, or how many reviewers performed the data extraction. Data on the occurrence of each outcome of interest were extracted from individual studies and were used to calculate odds ratios (ORs). For studies with more than two treatment groups, data on different doses of the same formulation of AmB were pooled.

How were the studies combined?

The studies were combined using a random-effects meta-analysis. The pooled ORs and 95% confidence intervals (CIs) were calculated separately for all-cause mortality, treatment efficacy (empirical plus antifungal combined), changes in serum creatinine, and infusion-related reactions. The number-needed-to-treat (NNT) was calculated for each group. Publication bias was assessed by examining a funnel plot for each outcome measure.

How were differences between studies investigated?

The results for combined studies were displayed graphically using forest plots. Some differences between the studies were discussed, but no formal assessment of statistical heterogeneity was presented. The authors considered that there were too few studies to investigate heterogeneity.

Eight RCTs (n=1,933) were included.

The validity scores ranged from 15 to 26 (median 22). All of the RCTs scored more than 14 points, while 5 RCTs scored more than 20 points.

All-cause mortality: lipid-based AmB significantly reduced the likelihood of all-cause mortality compared with conventional AmB (OR 0.72, 95% CI: 0.54, 0.97). The NNT to prevent one death with lipid-based AmB compared with conventional AmB was 31.

Efficacy: based on 1,620 patients in 6 studies, there was no significant difference in efficacy between lipid-based formulations and conventional AmB (OR 1.21, 95% CI: 0.98, 1.49).

There were insufficient data to analyse the difference between treatments for clinical response rates for proven infection.

Changes in serum creatinine: based on 3 studies, AmB lipid complex and liposomal AmB significantly reduced the likelihood of doubling serum creatinine (OR 0.42 (95% CI: 0.33, 0.54). The NNT to prevent a doubling of serum creatinine for AmB lipid complex or liposomal AmB compared with conventional AmB was 6.

Infusion-related reactions: there was no significant difference between lipid-based and conventional AmB for infusion-related reactions before premedication (OR 0.61, 95% CI: 0.31, 1.20) or after premedication (OR 0.79, 95% CI: 0.58, 1.08), based on 191 patients in 3 studies and 782 patients in 3 studies, respectively.

Funnel plots showed no evidence of publication bias, but the number of studies was small.

The authors stated that economic evaluations were eligible for inclusion in the review but none were reported in the paper.

Lipid-based AmB reduced mortality and renal damage compared with conventional AmB.

The review question was clear in terms of the study design, intervention, participants and outcomes. Several relevant sources were searched, the search terms were stated, and attempts were made to minimise language and publication bias. The selection of studies and the validity assessment were checked by a second reviewer, but the lack of independence may lead to errors and bias. The methods used to extract the data were not described, so it is not known whether any efforts were made to reduce errors and bias. Validity was assessed using specified established criteria, but only the total scores were reported. There were insufficient details to assess whether the follow-up was sufficient, both in completeness and duration, to adequately assess mortality.

Statistical heterogeneity was not formally assessed, although the authors stated that there was a high degree of heterogeneity; the decision to statistically pool was not, therefore, appropriate. The authors discussed the results in terms of risk when this is not what the analysis was based on. It may have been more appropriate to have used relative risks, rather than ORs, given the outcomes of the review. Given the limitations in the reporting of the review process and the apparent heterogeneity across studies and possible inappropriate statistical analysis, the authors' conclusions may not be reliable. Furthermore, the results were generally based on short-term follow-up.

Practice: The authors did not state any implications for practice.

Research: The authors stated that future trials in patients with proven fungal infection should control for factors such as premedication, infusion rates, fluid preloading, sodium or potassium supplementation, and concomitant medication.

Elan Pharma Ltd.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.