Journal Logo

Article

Drug-Induced Nephrotoxicity Caused by Amphotericin B Lipid Complex and Liposomal Amphotericin B

A Review and Meta-Analysis

Safdar, Amar MD; Ma, Jonathan PhD; Saliba, Fouzi MD; Dupont, Bertrand MD; Wingard, John R. MD; Hachem, Ray Y. MD; Mattiuzzi, Gloria N. MD; Chandrasekar, Pranatharthi H. MD; Kontoyiannis, Dimitrios P. MD; Rolston, Kenneth V. MD; Walsh, Thomas J. MD; Champlin, Richard E. MD; Raad, Issam I. MD

Author Information
doi: 10.1097/MD.0b013e3181e9441b
  • Free

Abstract

INTRODUCTION

The spectrum of fungal infection continues to change as newer drugs against Aspergillus and Candida species are used commonly in immunosuppressed cancer and transplant patients.2,19,22,25,27 A rise in systemic zygomycosis among patients who have received Aspergillus active triazoles has redirected the focus on polyene-based antifungal therapy in neutropenic cancer patients and stem cell transplant recipients with graft-versus-host disease.13,26 The oldest drug in the polyene class of antifungal agents is amphotericin B-deoxycholate, which has been a mainstay of therapy for systemic fungal infections since its release in the late 1950s7,23 but is associated with many adverse effects that limit its clinical utility, including acute renal failure. During the 1990s, newer lipid preparations of amphotericin B were developed to alleviate drug toxicity.

The reported frequency of acute renal failure associated with amphotericin B-deoxycholate varies widely with the definition used and increases with cumulative dose delivered.12,24 A measurable decline in renal function has been reported to occur in up to 80% of patients who receive the drug.24 Acute renal failure associated with amphotericin B-deoxycholate may be dose-limiting and can prolong length of stay and increase treatment costs.3 An analysis of 707 adult patients at a tertiary-care center, conducted by Bates and colleagues,3 found that when renal failure develops, patients' stay in the hospital increases by 8.2 days, and total cost per patient increases by $29,823. In addition, the mortality rate was higher when renal failure occurred: 54% versus 16% in patients without renal complications.3 Patients receiving low cumulative doses of amphotericin B-deoxycholate (<1.0 g) have been reported to develop renal failure less often.24

Therapeutic advantages of using amphotericin B in a lipid delivery system have allowed increased delivery of amphotericin B to sites of infection with less toxicity. These systems have significantly improved tolerability of this drug, as patients experience fewer infusion-associated adverse reactions and a considerably improved nephrotoxicity profile. The 2 commonly used lipid formulations of amphotericin B, amphotericin B lipid complex (ABLC) and liposomal amphotericin B (L-AmB), are both associated with reduced renal toxicity compared with conventional amphotericin B.8,23 However, in a randomized trial by Wingard et al,34 in which 244 patients received empiric antifungal therapy for febrile neutropenia, 166 patients who received L-AmB had significantly lower rates of renal dysfunction compared with 78 patients who received ABLC. Since that time, new information, much of it from animal models, has become available concerning the differing pharmacokinetics of amphotericin B lipid formulations and the interaction with the immune system.1,10,14-16,30 For instance, 2 recent studies by Lewis et al15,16 demonstrate early fungal clearance following administration of ABLC compared with L-AmB in murine models of invasive pulmonary aspergillosis and zygomycosis. The clinical significance of these findings is yet to be determined. Whether there are differences in renal safety among the 2 commonly used lipid formulations of amphotericin B remains controversial. We sought to analyze the existing data for nephrotoxicity associated with ABLC and L-AmB in adults who received these agents for the treatment and prophylaxis of fungal infection.

LITERATURE SEARCH METHODS AND CONSIDERATIONS FOR STUDY INCLUSION

We searched the PubMed MEDLINE database (National Library of Medicine, Bethesda, MD) for studies comparing nephrotoxicity between ABLC and L-AmB published between 1995 and 2008. Searches with terms "Amphotericin B and Nephrotoxicity" yielded more than 280 abstracts, which were further refined by including the search terms "Abelcet, amphotericin B lipid complex, ABLC, AmBisome, liposomal amphotericin B or L-AmB." Only primary reports, in the English language, of patient studies that directly compared nephrotoxicity associated with ABLC and L-AmB treatment were considered for inclusion in the meta-analysis. Reviews were not included in the meta-analysis; however primary reports cited in reviews were retrieved and evaluated for possible inclusion.

No other databases were searched in addition to PubMed for journal articles, however, abstracts presented at key scientific meetings were researched. Meetings with an emphasis on topics including infectious diseases, microbiology, fungal infections, and pharmacy/pharmacology were prioritized for the likelihood of containing studies comparing ABLC and L-AmB. We searched more than 10 annual meeting programs for relevant study data using similar search terms as in the PubMed searches. Meeting programs searched include the Focus on Fungal Infections (FOFI), Advances Against Aspergillosis (AAA), Infectious Diseases Society of America (IDSA) Annual Meeting, Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), American College of Clinical Pharmacology, the American College of Clinical Pharmacy, and the American Society of Health-System Pharmacists meetings.

PubMed results and abstracts were reviewed by 2 individuals. In total, 11 studies were identified that compared nephrotoxicity between ABLC and L-AmB; 3 were excluded because of the lack of a comparison group,31 incomplete data,5 and a very young neonatal population compared with other studies (average age ≤1 mo)17 (Figure 1).

FIGURE 1
FIGURE 1:
Study search strategy flow chart.

There is no harmonized definition for nephrotoxicity resulting from amphotericin B and the lipid formulations of amphotericin B. Serum creatinine (sCr) increases ≥1.5 mg/dL or >20% higher than baseline values represent clinically relevant decreases in kidney function.29 Studies included in this review reported established standard definitions of nephrotoxicity that are well accepted in peer-reviewed journals. We also reviewed information regarding concomitant nephrotoxic drugs or agents, prehydration (saline loading), comorbidities (for example, diabetes, hypertension, peripheral vascular disease), age of patient populations, and average daily dose between the studies. In all 8 studies analyzed, ABLC and L-AmB were administered at recommended doses (5 mg/kg and 3-5 mg/kg, respectively), with the exception of the study by Fleming et al,9 in which the daily dose of ABLC was lower than that of L-AmB; in the report by Mattiuzzi et al,20 in which the daily dose of ABLC was also lower but similar to the L-AmB arm; and in the study by Hachem et al,11 where patients received doses of ABLC or L-AmB up to 10 mg/kg. See Table 1 for a summary of the characteristics of the 8 studies included in the meta-analysis.4,9,11,18,20,21,28,34

TABLE 1
TABLE 1:
Summary of Studies Analyzed in the Meta-Analysis

REVIEW OF STUDIES COMPARING NEPHROTOXICITY ASSOCIATED WITH ABLC AND L-AMB

Wingard et al34 conducted a double-blind, randomized study involving 18 centers in the United States between October 1997 and August 1998. The study directly compared ABLC and L-AmB, administered empirically to febrile neutropenic patients aged 2 years or older, who had persistent fever lasting at least 72 hours during broad-spectrum antibiotic treatment. Comparisons of the drugs were made for a number of safety parameters, including the frequency of chills/rigors, the frequency of infusion-related reactions, and the frequency of nephrotoxicity. Two hundred forty-four patients were randomized (1:1:1) to receive ABLC at a dose of 5 mg/kg per day (n = 78), L-AmB at a dose of 3 mg/kg per day (n = 85), or L-AmB at a dose of 5 mg/kg per day (n = 81). Nephrotoxicity in this study was defined as an increase in sCr greater than 100% above baseline value (2 × baseline); for patients aged more than 16 years, the post-baseline peak sCr value also had to be greater than 1.2 mg/dL. The study also evaluated increases in sCr by greater than 50% and 200% above baseline value (1.5 × and 3 × baseline, respectively).

Significantly less nephrotoxicity was observed among patients treated with L-AmB (3 mg/kg per day and 5 mg/kg/d) compared with ABLC (5 mg/kg/d), 14.1% and 14.8% versus 42.3%, respectively (p = 0.01). This lower frequency of nephrotoxicity was evident regardless of age, receipt of a bone marrow transplant, transplant type, or use of immunosuppressive drugs. Administration of concomitant nephrotoxic medications was similar among the 3 treatment groups. A peak sCR value >3 mg/dL was observed in 7.1% of the 3-mg/kg L-AmB group, 1.2% of the 5-mg/kg L-AmB group, and 12.8% of the ABLC group; the difference was statistically significant between the 5-mg/kg per day L-AmB group and the ABLC group (p < 0.01) but not between the 3 mg/kg per day L-AmB group and the ABLC group. The increase in sCr level was significantly greater in the group of patients treated with ABLC (mean sCr change [mg/dL ± SD]: 0.5 ± 0.8 for 3 mg/kg L-AmB; 0.4 ± 0.4 for 5 mg/kg L-AmB; 1.0 ± 1.0 for 5 mg/kg ABLC; p ≤ 0.001, L-AmB vs. ABLC). The peak sCr level was also higher with ABLC (peak sCr value [mg/dL ± SD]: 1.3 ± 1.0 for 3 mg/kg L-AmB; 1.2 ± 0.6 for 5 mg/kg L-AmB; 1.8 ± 1.2 for 5 mg/kg ABLC; p ≤ 0.001, L-AmB vs. ABLC). There were no statistical differences in nephrotoxicity between high- and low-risk patients in the L-AmB group (15.4% vs. 13.9%) or the ABLC group (45.5% vs. 41.8%).

Fleming et al9 conducted a randomized study at the University of Texas, M.D. Anderson Cancer Center between August and December 1997. Seventy-five consecutive patients with leukemia developed 82 infectious episodes and were assigned to receive ABLC (n = 43 episodes) or L-AmB (n = 39 episodes). Patients with fever of unknown origin received 3 mg/kg study drug daily, and those with pneumonia, sinusitis, and cellulitis received a daily dose of 4-5 mg/kg. The daily drug dose was escalated to 5 mg/kg in patients with documented fungal infections. Infusion-related adverse events were evaluated following the first administration of ABLC or L-AmB. Renal and liver function were monitored twice a week by monitoring sCr and bilirubin levels, respectively, beginning on the first day of therapy until the end of treatment. Nephrotoxicity was defined as an increase in sCr greater than 50% above baseline value (1.5 × baseline) that persisted until the end of treatment or discontinuation of study drug. The authors reported a higher rate of nephrotoxicity in patients who received ABLC (16/40, 40%) compared with L-AmB (10/36, 28%), but this difference was not statistically significant (p = 0.26). In contrast, the incidence of hepatotoxicity was lower in the ABLC treatment arm (38%) than in the L-AmB treatment group (59%; p = 0.05). Almost all patients in both groups received concomitant nephrotoxic drugs (98% ABLC, 92% L-AmB).

Cannon et al4 carried out a single-center prospective and retrospective observational study between 1996 and 1999, which compared ABLC and L-AmB at a 350-bed teaching hospital. Of the 67 patients enrolled in the study, 46 received ABLC, and 21 received L-AmB. Twenty-six of the 46 (56%) patients receiving ABLC and 15 of the 21 (71%) patients receiving L-AmB were refractory to a prior course of antifungal therapy. Dosage and duration of ABLC and L-AmB, and underlying conditions, were similar in both treatment groups. Oncology patients accounted for most of those treated with ABLC (70%) and L-AmB (90%). In this study, nephrotoxicity was defined as >100% increase in baseline sCr. A higher incidence of nephrotoxicity was observed in patients receiving L-AmB (4 of 19 patients) than ABLC (2 of 46 patients), but this was not statistically significant. Analysis of nephrotoxicity may have been affected by a significant difference in baseline sCr levels (1.77 mg/dL ABLC vs. 1.0 mg/dL L-AmB; p = 0.003). All these patients were treated simultaneously with known nephrotoxic drugs such as vancomycin, aminoglycosides, acyclovir, and tacrolimus. However, nephrotoxicity was not statistically different, regardless of baseline sCr levels. One patient receiving L-AmB developed pulmonary Zygomycetes after 15 days of therapy. This patient died after receiving approximately 13 g of L-AmB.

The McKechnie et al study21 was a prospective and retrospective observational multicenter study at 11 tertiary care centers and 1 community hospital in Canada that evaluated differences in nephrotoxicity caused by ABLC (n = 150) and L-AmB (n = 104). Patients aged 2 years or older who received a minimum of 4 doses of ABLC or L-AmB and were not receiving dialysis were enrolled. Nephrotoxicity was defined as at least a 1.5 × increase in baseline sCr levels. No significant differences were detected between ABLC and L-AmB in any of the renal parameters analyzed in the study. Patients in the 2 groups were comparable with respect to baseline sCr levels, prior therapy with amphotericin B-deoxycholate, saline loading, and number of concurrent nephrotoxins. An average net change in sCr level (baseline to peak) of 46.7 μmol/L was observed in patients treated with ABLC, and 42.8 μmol/L for patients treated with L-AmB (p = not significant [NS]). Patients treated with ABLC and L-AmB experienced the same amount of change in sCr from baseline to end of therapy-9 μmol/L. An increase of at least 50% in sCr was experienced by 30.7% of ABLC-treated patients and 28.4% of patients in the L-AmB group; 13.6% of ABLC and 12.7% of L-AmB patients had a doubling in sCr, and 2.1% of ABLC and 3.9% of L-AmB patients had a tripling of baseline sCr (p = NS).

Mattiuzzi et al20 conducted an open-label, prospective, single-institution study with historical controls between April 1998 and May 1999. The study compared ABLC (n = 131) with L-AmB (n = 70) as antifungal prophylaxis in patients with acute myelogenous leukemia and high-risk myelodysplastic syndrome who were undergoing induction chemotherapy. Patients who received at least 4 days of antifungal prophylaxis were eligible for enrollment. National Cancer Institute (NCI) toxicity grades 0 to 4 were used to define nephrotoxicity in this study. The study did not find significant differences in nephrotoxicity between patients receiving ABLC and L-AmB for antifungal prophylaxis. Eighteen percent of patients receiving ABLC therapy and 15% of L-AmB-treated patients withdrew from the study because of probable drug-related adverse effects. Sixteen patients (12.2%) receiving ABLC and 14 patients (20%) treated with L-AmB experienced grade 1 or 2 nephrotoxicity, categorized as probable or possibly related to therapy, but no patients in either of the 2 treatment groups experienced grade 3 or 4 nephrotoxicity, and no deaths were related to study drug administration.

The study by Malani et al18 was a retrospective review of medical records from an outpatient clinic at a tertiary care center performed between January 1997 and July 2002. The authors evaluated the types and frequencies of adverse events associated with home infusion of amphotericin B-D, and the 3 amphotericin B lipid formulations (ABLC, L-AmB, amphotericin B colloidal dispersion). Most if not all of the patients had fungal infections, and nephrotoxicity was defined as >1 mg/dL or 2 × increase in baseline sCr. Significant rates of 4 types of adverse events occurred: nephrotoxicity was associated with 46/113 courses of therapy (41%), electrolyte abnormalities with 40/113 courses (35%), infusion reactions with 13/113 courses (12%), and venous access device complications with 13/113 courses (12%). Nephrotoxicity was noted in 45% of courses of ABLC, compared with 32% courses of L-AmB, but the difference was not significant (p = 0.36). In this study, the time interval between administration of the antifungal agents and development of nephrotoxicity was also evaluated; for ABLC-treated patients, it was 19 ± 19 days, and for L-AmB, it was 23 ± 19 days (p = 0.43). In general, nephrotoxicity was significantly lower in children aged younger than 13 years (12%) compared with patients aged 60 years or older (67%) (p = 0.002).

Saliba et al28 reported a prospective multicenter survey conducted in France between April 2003 and December 2004 to assess renal safety of ABLC and L-AmB. Eighty-eight patients with a mean age of 49.6 ± 14 years were enrolled; 60 patients were treated with ABLC and 28 patients with L-AmB. Sixty-eight percent of the patients received 2 or more nephrotoxic drugs (72% for ABLC, 61% for L-AmB). The study also analyzed a subset of patients who received treatment for at least 10 days with ABLC (n = 37) or L-AmB (n = 19). Nephrotoxicity was defined as a doubling in sCr levels from baseline to end of study; changes in creatinine clearance were also measured. No significant difference between ABLC and L-AmB was reported in the change in sCr from baseline to end of therapy (ABLC: baseline 81.3 μmol/L, end of therapy 115.0 μmol/L; L-AmB: baseline 88.8 μmol/L, end of therapy 106.0 μmol/L). In addition, there was no significant difference in the change in creatinine clearance from baseline between the 2 lipid formulations (ABLC: baseline 94.6 mL/min, end of therapy 89.5 mL/min; L-AmB: baseline 85.2 mL/min, end of therapy 71.6 mL/min). Fourteen patients on ABLC and 2 patients on L-AmB developed elevations of sCr 2 × normal levels. In the subset of patients who received treatment for at least 10 days, 4 patients on ABLC and 1 patient on L-AmB developed elevations of sCr 2 × normal levels. Nephrotoxicity was reported in a larger proportion of patients in the ABLC treatment group than in the L-AmB group, but this was not statistically significant (p = 0.067 for both the entire population and the subgroup of patients that received treatment for 10 days). Forty percent of patients in the ABLC treatment group received 3 or more concomitant nephrotoxic drugs or agents, compared with 21% of patients receiving L-AmB.

In 2008, Hachem et al11 published results of a retrospective single-center study performed between June 1993 and December 2005 in adults with advanced hematologic malignancy and proven or probable invasive aspergillosis. The study compared the safety of ABLC and L-AmB, used as either primary or salvage therapy, in 381 consecutive patients. Of these patients, 158 received primary antifungal therapy with either ABLC (n = 52) or L-AmB (n = 106), and 30 ABLC and 51 L-AmB salvage regimens were administered. Patients evaluated in this study had been assessed weekly for hepatotoxicity, defined as a 2 × increase from baseline in sCr levels. The results of this study are similar to those reported by Wingard et al,34 in that the use of ABLC as primary antifungal therapy was associated with significantly more adverse events and nephrotoxicity than L-AmB (p < 0.001). However, adverse events and nephrotoxicity were not significantly different between the 2 amphotericin B lipid formulations in patients receiving salvage therapy (p = 0.67).11 Of note, this report evaluated a group of patients that was not typical of the hematologic cancer population with invasive aspergillosis; individuals in the study had advanced-stage hematologic malignancy or were severely ill, probably resulting in reduced renal reserves.

META-ANALYSIS OF DRUG-INDUCED NEPHROTOXICITY

We used the Cochran-Mantel-Haenszel test to compare and determine the odds ratio (OR) and relative risk (RR) of nephrotoxicity between ABLC and L-AmB. Data from the 8 studies were pooled into the meta-analysis with study as the stratification variable. Common OR and RR were generated to measure the association between the rate of nephrotoxicity and amphotericin B lipid formulations. An OR and/or RR closer to unity indicates a similarity between the 2 formulations. The further away the OR and RR are from unity, the stronger the evidence that the 2 formulations are different in terms of nephrotoxicity. The Cochran-Mantel-Haenszel method assumes that the studies are homogeneous. We used the Breslow-Day test for homogeneity of the ORs to test the hypothesis that these studies were homogeneous and had similar ORs. A significant p value (p < 0.05) indicates that the assumption of homogeneity of the ORs may have been violated and that there was statistical evidence that at least 1 of the ORs from the studies was very different from the rest. In such a case, it is important to look into the possible underlying reasons, revalidate which studies should be included in the meta-analysis, and carefully interpret the results accordingly. Statistical analyses were performed using SAS version 9.1.3 (SAS Institute, Inc., Cary, NC).

The 8 studies identified were used in different combinations to generate 8 different analyses (Figure 2), and homogeneity of the ORs across the studies was assessed. The meta-analysis of all 8 studies (n = 1160) evaluating nephrotoxicity showed an increased probability of nephrotoxicity in patients treated with ABLC (n = 588) compared with those treated with L-AmB (n = 572). The Cochran-Mantel-Haenszel OR and RR were 1.75 and 1.55, respectively, and were slightly lower when the analysis was performed with the salvage patient population reported in the Hachem11 study (OR, 1.59; RR, 1.44; n = 1083). However, there was a significant lack of homogeneity across these studies (p < 0.001 and p = 0.003, respectively, by Breslow-Day test). The same analysis conducted with a subset of patients in the Saliba study28 who received treatment for 10 days (n = 1128) yielded similar results. Patients receiving ABLC had an increased probability of developing nephrotoxicity compared with those receiving L-AmB (OR, 1.68; RR, 1.50); analysis with the salvage patient population reported in the Hachem study11 yielded lower OR and RR (OR, 1.52; RR, 1.39; n = 1051). Breslow-Day analysis continued to show a lack of homogeneity in these studies (p < 0.0001 and p = 0.0005, respectively). The lack of homogeneity indicates that the increased OR and risk of developing nephrotoxicity associated with ABLC should be interpreted cautiously.

FIGURE 2
FIGURE 2:
Meta-analysis comparing nephrotoxicity between ABLC and L-AmB. Results indicate that there is no difference in nephrotoxicity between ABLC and L-AmB when the Wingard study34 is excluded from the meta-analysis, particularly when the analysis was performed with the salvage patient population reported in the study by Hachem et al.11 Cochran-Mantel-Haenszel odds ratios and relative risks with their associated 95% confidence intervals are given.

We tested various permutations of the 8 studies by re-evaluating the statistical analysis while excluding 1 study at a time. We found that when the trial by Wingard et al34 was omitted (n = 916), the probability of nephrotoxicity was similar between patients treated with ABLC (n = 510) versus those treated with L-AmB (n = 406) (OR, 1.31; RR, 1.24); the 7 studies included in this subanalysis were not homogenous by Breslow-Day test (p = 0.0023). No differences in nephrotoxicity were found between the 2 lipid formulations when the Wingard study34 was excluded from the analysis with the salvage patient population from the Hachem study11 (OR, 1.12; RR, 1.09; n = 839), and the subset of studies was more homogenous (p = 0.054).

Additionally, the probability of experiencing nephrotoxicity was similar between ABLC and L-AmB in analyses excluding the Wingard study,34 using the subpopulation of patients treated for 10 days in the Saliba study28 (OR, 1.22; RR, 1.17; n = 884), although these studies were not homogeneous based on results of the Breslow-Day test (p = 0.0055). The same analysis performed with patients treated for 10 days in the Saliba study and the salvage population of patients in the Hachem study11 also showed no differences in the probability of nephrotoxicity between the 2 lipid formulations (OR, 1.02; RR, 1.02; n = 807). Breslow-Day analysis indicated that this subset of studies was homogeneous (p = 0.1430). Results from these analyses suggest that there is no difference in nephrotoxicity between ABLC and L-AmB when the Wingard study is excluded from the meta-analysis, especially when data from the salvage patient population reported by Hachem et al were analyzed (Figure 2).

We conducted additional subanalyses using various factors that were most commonly reported in the studies to define subgroups summarized in Table 2.4,9,11,18,20,21,28,34 The 8 studies were grouped by study design (randomized vs. non-randomized), mean/median age of patients (≥50 yr), percentage of bone marrow transfusion recipients (>25% of patients), and the proportion of patients receiving concomitant nephrotoxic drugs (≥80% of patients). Similar to the overall meta-analysis, the probability of patients experiencing nephrotoxicity was not related to the subanalyses. Higher rates of renal toxicity were observed in patients treated with ABLC than L-AmB only in subgroup analyses that include the Wingard study.34 Results from the subgroup analyses are represented in Figure 3.

FIGURE 3
FIGURE 3:
Subgroup meta-analyses evaluating nephrotoxicity between ABLC and L-AmB. Studies were grouped by study design, mean/median age of patients, percentage of bone marrow transplantation recipients, and the proportion of patients receiving concomitant nephrotoxic drugs. Results demonstrate that the probability of nephrotoxicity was similar in patients treated with ABLC compared with L-AmB, except in the subgroups that include the Wingard study.34 Cochran-Mantel-Haenszel odds ratios and relative risks with their associated 95% confidence intervals are given.
TABLE 2
TABLE 2:
Factors Reported in the Studies Used to Define Meta-Analysis Subgroups

DISCUSSION

Differences in drug-induced nephrotoxicity between lipid formulations of amphotericin B have been a subject of debate. We analyzed 8 studies by the Cochran-Mantel-Haenszel method that directly compared nephrotoxicity between ABLC and L-AmB; only 2 of the studies evaluated, those by Wingard et al34 and Fleming et al,9 were randomized. With the exception of meta-analyses that included the Wingard study, our results demonstrate that the rate of nephrotoxicity is not significantly different in patients treated with ABLC versus L-AmB. Additional subanalyses revealed that, except for subgroups that included the Wingard study, no other factors, including randomization, age, percentage of bone marrow transfusion recipients, or the proportion of patients receiving concomitant nephrotoxic drugs, are associated with higher rates of nephrotoxicity in patients treated with ABLC compared with L-AmB.

Ever since its introduction, amphotericin B has been a mainstay in the treatment of invasive fungal infections because of its activity against a broad spectrum of fungal pathogens and rare cases of resistance. The use of conventional amphotericin B is associated with high rates of infusion-related reactions and adverse events, especially nephrotoxicity, which have hampered its clinical utility. Lipid formulations of amphotericin B have the same spectrum of activity but generally cause fewer adverse events and are associated with decreased renal toxicity compared with amphotericin B-deoxycholate, even when administered at higher doses.6,8,23 Both ABLC and L-AmB appear to be equally effective in treating fungal infections and are considered therapeutically equivalent for most fungal infections, as was the case in the studies analyzed in this meta-analysis that provided efficacy data.4,9,20,34 The 2008 study by Hachem et al11 also demonstrates that efficacy is comparable for both lipid formulations of amphotericin B in the treatment of invasive aspergillosis in patients with underlying hematologic malignancy.

Excluding the study by Wingard et al,34 the current meta-analysis suggests a lack of clinically relevant differences between ABLC and L-AmB with regard to nephrotoxicity. It is unclear why there was such a large difference in nephrotoxicity in the earlier trial by Wingard et al. One plausible explanation could be the way nephrotoxicity and infusion reactions were measured in the groups of heterogeneous patients in regard to severity of their underlying clinical conditions and the toxicity of concomitant treatments. Another explanation could be when nephrotoxicity was assessed, from day 1 through the 7-day follow-up evaluation. The average duration of therapy was shorter in that trial than the other studies. In a study by Walsh et al,33 creatinine levels were measured over a longer time course and suggested improvement with time; among 162 patients with elevated baseline sCr values at the start of ABLC therapy, the mean sCr values decreased significantly from the first week through the sixth week (p ≤ 0.0003). Improvements in renal function during therapy with ABLC were similar for all patient risk groups. Thus, nephrotoxicity during treatment with amphotericin B lipid formulations may be transient. These results suggest that the time point when nephrotoxicity is assessed during treatment is critical, the duration of therapy is important, and these issues should be taken into account when evaluating studies investigating renal toxicity of lipid forms of amphotericin B. Of course, an alternative interpretation is that there is more nephrotoxicity with ABLC, as shown by the Wingard trial, and there may have been unrecognized reasons for the lack of differences noted in the other studies. For example, the studies by McKechnie et al21 and Mattiuzzi et al20 required a minimal duration of therapy for inclusion that may have excluded those patients removed from therapy due to early nephrotoxicity; thus, early nephrotoxicity differences would not have been detected.

Results from the study by Hachem et al11 are similar to those reported by Wingard et al,34 in that the use of ABLC as primary antifungal therapy was associated with significantly more nephrotoxicity than L-AmB. However, this difference was not observed in patients receiving salvage therapy. In addition, although meta-analysis results showed an increased probability of developing nephrotoxicity when the primary treatment arm was included in the analyses, the OR and RR were considerably less than those observed with the Wingard study. As indicated earlier, the Hachem study looked at severely ill patients with advanced hematologic malignancies, who may have had reduced renal function, confounding the nephrotoxicity analysis between ABLC and L-AmB.

Despite reduced rates of adverse reactions compared with conventional amphotericin B, patients receiving antifungal therapy with ABLC and L-AmB still experience few infusion reactions and renal toxicity.8,23,29 Using a murine candidiasis model, Andes et al1 found that lipid formulations of amphotericin B distribute preferentially to the mononuclear phagocytic system. The interaction of free amphotericin B drug released from lipid vehicles with macrophages may be responsible for observed infusion reactions.1 However, macrophage nitric oxide synthase expression and tumor necrosis factor-α production is reduced by lipid formulations of amphotericin B.14 In addition, ABLC and L-AmB appear to downregulate or have no effect on genes coding for proinflammatory cytokines, which could explain the lower levels of infusion-related reactions compared with conventional amphotericin B.30 Lipid formulations of amphotericin B may have reduced nephrotoxicity because the drug is distributed to tissues of the reticulo-endothelial system, sparing the kidneys. Furthermore, in the kidneys, less amphotericin B is released from the lipid carrier, because the synthetic phospholipids have a greater affinity for amphotericin B than does cholesterol in renal epithelial cell membranes.32 Toxic effects and vasoconstriction associated with conventional amphotericin B are diminished as a result of lower amounts of amphotericin B reaching kidney cells.

Conclusions

Renal failure associated with the use of conventional amphotericin B results in longer hospital stays, increased treatment costs, and higher mortality rates.3 The development of lipid formulations of amphotericin B has improved delivery of the drug and decreased the incidence of adverse reactions and nephrotoxicity associated with conventional amphotericin B. A comprehensive review and analysis of studies comparing amphotericin B formulations demonstrate that lipid forms help preserve renal function and improve survival in patients critically ill from invasive fungal infections.29 Considering the costs associated with renal failure, the ability to use lipid formulations of amphotericin B becomes increasingly important in patients requiring broad-spectrum antifungal therapy. Our meta-analysis raises questions about the previously known relative nephrotoxicity of ABLC or L-AmB. In addition, no conclusive differences in response and outcome have been reported in patients with invasive fungal infections treated with ABLC or L-AmB. Therefore, cumulative evidence suggests that ABLC or L-AmB can be administered to immunocompromised individuals for the treatment or prophylaxis of invasive mycoses, with comparable efficacy and safety.

REFERENCES

1. Andes D, Safdar N, Marchillo K, Conklin R. Pharmacokinetic-pharmacodynamic comparison of amphotericin B (AmB) and two lipid-associated AmB preparations, liposomal AmB and AmB lipid complex, in murine candidiasis models. Antimicrob Agents Chemother. 2006;50:674-684.
2. Baddley JW, Pappas PG, Smith AC, Moser SA. Epidemiology of Aspergillus terreus at a university hospital. J Clin Microbiol. 2003;41:5525-5529.
3. Bates DW, Su L, Yu DT, Chertow GM, Seger DL, Gomes DR, Dasbach EJ, Platt R. Mortality and cost of acute renal failure associated with amphotericin B therapy. Clin Infect Dis. 2001;32:686-693.
4. Cannon JP, Garey KW, Danziger LH. A prospective and retrospective analysis of the nephrotoxicity and efficacy of lipid-based amphotericin B formulations. Pharmacotherapy. 2001;21:1107-1114.
5. Couch K, Chan S, Cincotta E, Piper JP, Forrest G. Comparative nephrotoxicity, adverse events, and clinical outcomes of lipid associated amphotericin B products. (Abstract P-047.) The 13th Focus on Fungal Infections; Maui, Hawaii; March 19-21, 2003.
6. Deray G. Amphotericin B nephrotoxicity. J Antimicrob Chemother. 2002;49(Suppl 1):37-41.
7. Dismukes WE. Introduction to antifungal drugs. Clin Infect Dis. 2000;30:653-657.
8. Dupont B. Overview of the lipid formulations of amphotericin B. J Antimicrob Chemother. 2002;49(Suppl 1):31-36.
9. Fleming RV, Kantarjian HM, Husni R, Rolston K, Lim J, Raad I, Pierce S, Cortes J, Estey E. Comparison of amphotericin B lipid complex (ABLC) vs. ambisome in the treatment of suspected or documented fungal infections in patients with leukemia. Leuk Lymphoma. 2001;40:511-520.
10. Groll AH, Lyman CA, Petraitis V, Petraitiene R, Armstrong D, Mickiene D, Alfaro RM, Schaufele RL, Sein T, Bacher J, Walsh TJ. Compartmentalized intrapulmonary pharmacokinetics of amphotericin B and its lipid formulations. Antimicrob Agents Chemother. 2006;50:3418-3423.
11. Hachem RY, Boktour MR, Hanna HA, Husni RN, Torres HA, Afif C, Kontoyiannis DP, Raad II. Amphotericin B lipid complex versus liposomal amphotericin B monotherapy for invasive Aspergillosis in patients with hematologic malignancy. Cancer. 2008;112:1282-1287.
12. Hoitsma AJ, Wetzels JF, Koene RA. Drug-induced nephrotoxicity. Aetiology, clinical features and management. Drug Saf. 1991;6:131-147.
13. Kontoyiannis DP, Lionakis MS, Lewis RE, Chamilos G, Healy M, Perego C, Safdar A, Kantarjian H, Champlin R, Walsh TJ, Raad II. Zygomycosis in a tertiary-care cancer center in the era of Aspergillus-active antifungal therapy: a case-control observational study of 27 recent cases. J Infect Dis. 2005;191:1350-1360.
14. Larabi M, Legrand P, Appel M, Gil S, Lepoivre M, Devissaguet J, Puisieux F, Barratt G. Reduction of NO synthase expression and tumor necrosis factor alpha production in macrophages by amphotericin B lipid carriers. Antimicrob Agents Chemother. 2001;45:553-562.
15. Lewis RE, Albert ND, Hou J, Kontoyiannis DP. Comparative pharmacodynamics of amphotericin B lipid complex (ABLC) and liposomal amphotericin B (L-AMB) in a murine model of acute invasive pulmonary zygomycosis. (Abstract M-2139.) 48th Annual Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC)/46th Annual Infectious Diseases Society of America (IDSA); Washington, DC; October 25-28, 2008.
16. Lewis RE, Liao G, Hou J, Chamilos G, Prince RA, Kontoyiannis DP. Comparative analysis of amphotericin B lipid complex and liposomal amphotericin B kinetics of lung accumulation and fungal clearance in a murine model of acute invasive pulmonary aspergillosis. Antimicrob Agents Chemother. 2007;51:1253-1258.
17. Lopez Sastre JB, Coto Cotallo GD, Fernandez Colomer B. Neonatal invasive candidiasis: a prospective multicenter study of 118 cases. Am J Perinatol. 2003;20:153-163.
18. Malani PN, Depestel DD, Riddell J, Bickley S, Klein LR, Kauffman CA. Experience with community-based amphotericin B infusion therapy. Pharmacotherapy. 2005;25:690-697.
19. Marr KA, Carter RA, Crippa F, Wald A, Corey L. Epidemiology and outcome of mould infections in hematopoietic stem cell transplant recipients. Clin Infect Dis. 2002;34:909-917.
20. Mattiuzzi GN, Kantarjian H, Faderl S, Lim J, Kontoyiannis D, Thomas D, Wierda W, Raad I, Garcia-Manero G, Zhou X, Ferrajoli A, Bekele N, Estey E. Amphotericin B lipid complex as prophylaxis of invasive fungal infections in patients with acute myelogenous leukemia and myelodysplastic syndrome undergoing induction chemotherapy. Cancer. 2004;100:581-589.
21. McKechnie M, Rotstein C, McTaggart B. A multi-center, retrospective comparison of the nephrotoxicity effects of amphotericin B lipid complex and liposomal amphotericin B. (Abstract P-048.) The 13th Focus on Fungal Infections; Maui, Hawaii; March 19-21, 2003.
22. Nucci M, Marr KA, Queitoz-Telles F, Martins CA, Trabasso P, Costa S, Voltarelli JC, Colombo AL, Imhof A, Pasquini R, Maiolino A, Souza CA, Anaissie E. Fusarium infection in hematopoietic stem cell transplant recipients. Clin Infect Dis. 2004;38:1237-1242.
23. Ostrosky-Zeichner L, Marr KA, Rex JH, Cohen SH. Amphotericin B: time for a new "gold standard". Clin Infect Dis. 2003;37:415-425.
24. Pathak A, Pien FD, Carvalho L. Amphotericin B use in a community hospital, with special emphasis on side effects. Clin Infect Dis. 1998;26:334-338.
25. Pfaller MA, Diekema DJ. Rare and emerging opportunistic fungal pathogens: concerns for resistance beyond Candida albicans and Aspergillus fumigatus. J Clin Microbiol. 2004;42:4419-4431.
26. Roden MM, Zaoutis TE, Buchanan WL, Knudsen TA, Sarkisova TA, Schaufele RL, Sein M, Sein T, Chiou CC, Chu JH, Kontoyiannis DP, Walsh TJ. Epidemiology and outcome of zygomycosis: a review of 929 reported cases. Clin Infect Dis. 2005;41:634-653.
27. Safdar A. Difficulties and healthcare utilization/cost associated with the management of fungal infections in patients with acute myelogenous leukemia. Clin Advan Hematol Oncol. 2008;6(Suppl 12):4-10.
28. Saliba F, Cheron N, Ichai Ph, Renault A, Sirvent A, Depil S, Saulquin B, Witz B, Visanca S, Bastien O, Guillaumin S. A prospective French National Survey to assess renal safety of amphotericin B lipid complex and liposomal amphotericin B. (Abstract 123.) Int J Infect Dis. 2006; 10(Suppl 1):S68-S69.
29. Saliba F, Dupont B. Renal impairment and amphotericin B formulations in patients with invasive fungal infections. Med Mycol. 2008;46:97-112.
30. Simitsopoulou M, Roilides E, Dotis J, Dalakiouridou M, Dudkova F, Andreadou E, Walsh TJ. Differential expression of cytokines and chemokines in human monocytes induced by lipid formulations of amphotericin B. Antimicrob Agents Chemother. 2005;49:1397-1403.
31. Slain D, Miller K, Khakoo R, Fisher M, Wierman T, Jozefczyk K. Infrequent occurrence of amphotericin B lipid complex-associated nephrotoxicity in various clinical settings at a University Hospital: a retrospective study. Clin Ther. 2002;24:1636-1642.
32. Walsh TJ, Anaissie EJ, Denning DW, Herbrecht R, Kontoyiannis DP, Marr KA, Morrison VA, Segal BH, Steinbach WJ, Stevens DA, van Burik JA, Wingard JR, Patterson TF; Infectious Diseases Society of America. Treatment of aspergillosis: clinical practice guidelines of the infectious diseases society of America. Clin Infect Dis. 2008;46:327-360.
33. Walsh TJ, Hiemenz JW, Seibel NL, Perfect JR, Horwith G, Lee L, Silber JL, DiNubile MJ, Reboli A, Bow E, Lister J, Anaissie EJ. Amphotericin B lipid complex for invasive fungal infections: analysis of safety and efficacy in 556 cases. Clin Infect Dis. 1998;26:1383-1396.
34. Wingard JR, White MH, Anaissie E, Raffalli J, Goodman J, Arrieta A. A randomized, double-blind comparative trial evaluating the safety of liposomal amphotericin B versus amphotericin B lipid complex in the empirical treatment of febrile neutropenia. Clin Infect Dis. 2000;31:1155-1163.
© 2010 Lippincott Williams & Wilkins, Inc.