Amphotericin B deoxycholate versus liposomal amphotericin B: effects on kidney function

Abstract

Background

The incidence of invasive fungal infections has increased globally as a result of several factors. Conventional amphotericin B (sodium deoxycholate) has been used as standard therapy for the treatment of invasive fungal infections; however, it is associated with adverse drug reactions, including acute kidney injury (AKI). New formulations of amphotericin B have aimed to improve the safety profile of the conventional formulation.

Objectives

This review aimed to assess the effects of amphotericin B deoxycholate versus liposomal amphotericin B on kidney function.

Search methods

We searched Cochrane Kidney and Transplant's Specialised Register to 10 March 2015 through contact with the Trials' Search Co‐ordinator using search terms relevant to this review.

Selection criteria

We included randomised controlled trials (RCTs) that compared amphotericin B sodium deoxycholate with liposomal amphotericin B.

Data collection and analysis

Two authors independently assessed studies for eligibility and conducted risk of bias evaluation.

Main results

We included 12 studies (2298 participants) in this review. Of these, 10 were meta‐analysed (2172 participants). Liposomal amphotericin B was found to be significantly safer than conventional amphotericin B in terms of serum creatinine increase (RR 0.49, 95% CI 0.40 to 0.59). There was significant decrease in all infusion‐related reactions in the liposomal group compared with the conventional group: fever (4 studies, 1092 participants): RR 0.39, 95% CI 0.28 to 0.55; I² = 32%); chills and/or rigours (5 studies, 1081 participants): RR 0.27, 95% CI 0.15 to 0.48; I² = 75%); fever and/or rigours (2 studies, 720 participants): RR 0.68, 95% CI 0.52 to 0.90; I² = 58%); nausea (6 studies, 1187 participants): RR 0.50, 95% CI 0.35 to 0.72; I² = 0%); and vomiting (3 studies, 1019 participants): RR 0.51, 95% CI 0.27 to 0.95; I² = 61%). Overall, risk of bias in included studies was low or unclear for most domains. However, blinding of participants and personnel, blinding of outcome assessment and other bias (funding) tended to have a high risk of bias. The sensitivity analysis performed did not change the significance of difference in favour of the liposomal formulation. Assessment for publication bias found that review results were robust.

Authors' conclusions

Current evidence suggests that liposomal amphotericin B is less nephrotoxic than conventional amphotericin B (when the effect on kidney function is measured as an increase in serum creatinine level equal to or greater than two‐fold from the baseline level). We also found that there were fewer infusion‐related reactions associated with the liposomal formulation.

Plain language summary

Which amphotericin B formulation better preserves kidney function among people with fungal infection?

Fungal infections can cause ill health, and in some cases, death. Conventional amphotericin B has been used for many years to treat fungal infection. Although effective, this drug can cause kidney damage in about eight of every ten patients treated. People with kidney damage have longer stays in hospital, increased healthcare costs, and higher numbers of deaths.

To avoid the problems known to affect people who need treatment for fungal infection, new formulations of amphotericin B (called lipid‐associated) have been developed. Liposomal amphotericin B is a new formulation and is most often used in clinical practice to treat fungal infection.

We searched the literature published up to 10 March 2015, and identified 12 studies (2298 participants) that compared conventional amphotericin B with liposomal amphotericin B. We were able to meta‐analyse 10 studies (2172 participants). Study quality was assessed as moderate overall.

Liposomal amphotericin B was found to be associated with fewer kidney and infusion‐related reactions than conventional amphotericin B.

Summary of findings

Summary of findings for the main comparison. Amphotericin B deoxycholate versus liposomal amphotericin B: effects on kidney function.

Increase in serum creatinine (SCr) level ≥ two‐fold from baseline level
Patient or population: patients in treatment with amphotericin B Settings: inpatients Intervention: liposomal amphotericin B Comparison: conventional amphotericin B
Outcomes	Illustrative comparative risks* (95% CI)		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Conventional amphotericin B	Liposomal amphotericin B
Increase in SCr level: ≥ two‐fold from baseline level SCr level (mg/dL)	Study population		RR 0.49 (0.40 to 0.59)	2172 (10)	⊕⊕⊕ Moderate1
	261 per 1000	128 per 1000 (105 to 154)
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.

¹ The results obtained are consistent; sensitivity analysis was performed to see the influence of several factors in the effect estimate, however, the magnitude, the direction and the significance didn't change.

Background

The cause of increasing global incidence of invasive fungal infections has been attributed to a number of factors including invasive medical interventions, increased use of central venous catheters, long‐term hospital admissions and acquired or induced immunodeficiency (Pfaller 2010). The rise in incidence is notorious among immunocompromised and hospitalised patients with serious underlying diseases, although in developing countries, fungal infections also frequently affect healthy people (Brown 2012).

The most common fungal pathogens in humans are Candida spp, Cryptococcus neoformans, Pneumocystis jirovecii, Histoplasma capsulatum, Cryptoccocus gatii, Coccidoides posadasii and Aspergillus spp. (Brown 2012; Del Poeta 2012; Husain 2001; Pfaller 2010). In some European countries, annual incidence of fungaemia has been reported at between 2.2 and 11 episodes per 100,000 people (Arendrup 2008; Poikonen 2009). In population‐based studies from the United States, reported rates ranged from 6 to 8.7 per 100,000 people (Hajjeh 2004).

Among people admitted to intensive care units or who are immunocompromised, invasive fungal infections are responsible for significant morbidity and mortality. The AURORA project found that 16.5 cases of invasive fungal infections per 1000 admissions were caused by yeasts and 2.3 per 1000 admissions were caused by filamentous fungi (Montagna 2013). Although reported mortality rates for the most common pathogen ‐ Candida spp. ‐ varies, many studies report infection rates of up to 40% (Paramythiotou 2014). More than 1 million AIDS‐related deaths annually are due to invasive fungal infections and account for half of all deaths among this patient population (Armstrong‐James 2014).

The magnitude of challenges associated with management of invasive fungal infections means that identification of novel antifungal agents and approaches to improve effectiveness and safety of currently available agents has become imperative.

Description of the condition

Conventional amphotericin B (sodium deoxycholate) is a broad‐spectrum antifungal that has been used as standard therapy for treatment of many invasive fungal infections since its introduction to clinical practice in the 1950s (Bassetti 2011). Conventional amphotericin B has also been used to treat non‐fungal infections including visceral leishmaniasis and Naegleria fowleri primary amoebic meningoencephalitis.

Although other classes of antifungal drugs, such as azoles and echinocandins, have been developed in recent years, amphotericin B sodium deoxycholate continues to be widely used as the primary agent for treatment of people with serious invasive fungal infections (Laniado‐Laborín 2009; Ostrosky‐Zeichner 2003). A significant limitation of amphotericin B sodium deoxycholate is its safety profile; it is associated with adverse effects, including acute kidney injury (AKI) (Horwitz 2012; Mistro 2012) mediated by vasoconstriction and direct tubular toxicity. The reported incidence of nephrotoxicity varies widely; however, significant decrease in kidney function has been reported in over 80% of patients and is associated with higher mortality rates, length of hospital stay and costs (Bates 2001).

Description of the intervention

New lipid‐associated formulations have been developed to improve the safety profile of conventional amphotericin B including liposomal amphotericin B, amphotericin B lipid complex and amphotericin B colloidal dispersion (Graybill 1996; Kleinberg 2006). These formulations appear to offer advantages over the conventional formulation, notably a marked decrease in nephrotoxicity (Bassetti 2011; Noguchi 2013). An earlier Cochrane review reported increased nephrotoxicity associated with amphotericin B deoxycholate compared with lipid soluble formulations (Johansen 2014); however, the patient population evaluated was limited to those with cancer and neutropenia and did not account for other patient populations at high risk of nephrotoxicity.

How the intervention might work

The liposomal formulation of amphotericin B incorporates the antifungal drug into a liposomal bilayer, composed for phospholipids and cholesterol. When the liposome reaches the fungal cell, it is disrupted and the drug is released into the fungal cell membrane where it binds to the ergosterol. The liposome keeps its integrity in the presence of mammalian cells resulting in minimal toxicity (Adler‐Moore 2002).

Why it is important to do this review

The clinical relevance of nephrotoxicity is important when pharmacoeconomic analyses of lipid‐associated products are performed that attempt to assign secondary costs to amphotericin B nephrotoxicity. For a 70 kg patient, the estimated acquisition cost of liposomal amphotericin B at a dose of 5 mg/kg/day is approximately USD 1318.8/day; while the cost for amphotericin B deoxycholate at a dose 1 mg/kg/day is approximately USD 63.8/day, according to the average wholesale price (RED BOOK 2014).

Given high acquisition costs, a review of pivotal kidney health endpoints may inform resource allocation decisions for liposomal amphotericin B use, particularly in resource‐limited settings. Other adverse effects of amphotericin B products, such as infusion–related reactions, should also be considered when assessing the relative cost–effectiveness of these agents. Data limitations meant that we evaluated the effects of amphotericin B sodium deoxycholate versus liposomal amphotericin B on kidney function according to a number of kidney function parameters. Liposomal amphotericin B was selected as the comparator because it is a widely used lipid‐associated formulation.

Objectives

This review aimed to assess the effects of amphotericin B deoxycholate versus liposomal amphotericin B on kidney function.

Methods

Criteria for considering studies for this review

Types of studies

All RCTs and cluster‐RCTs comparing amphotericin B deoxycholate sodium versus liposomal amphotericin B were eligible for inclusion regardless of publication language.

Types of participants

Inclusion criteria

Patients diagnosed with proven, probable or possible invasive fungal infection were included, as well as those with documented or suspected neutropenia (absolute neutrophil count < 500 cells/mm³), those considered at high risk for developing invasive fungal infection by investigators, and those with other infectious diseases where amphotericin B is used as primary treatment.

Exclusion criteria

Patients receiving any form of parenteral amphotericin B within 10 days prior to study drug administration were excluded.

Types of interventions

Studies reporting patients receiving amphotericin B sodium deoxycholate alone or in combination with other licensed antifungal agents (control group) versus liposomal amphotericin B alone or in combination with other licensed antifungal agents (treatment group) irrespective of dose and/or duration of treatment were included. Amphotericin B formulations were administered intravenously.

Types of outcome measures

Primary outcomes

• Increase in serum creatinine (SCr) level ≥ than two‐fold from baseline.

Secondary outcomes

1. 50% increase in SCr occurring at any time during the study period

2. Discontinuation of amphotericin B therapy due to nephrotoxicity as determined by the investigators

3. Increase in SCr > 2 mg/dL at any time during the study period

4. Change in creatinine clearance (CrCl) from beginning to end of the study

5. Infusion‐related reactions as determined by the investigators.

Search methods for identification of studies

Electronic searches

We searched Cochrane Kidney and Transplant's Specialised Register to 10 March 2015 through contact with the Trials Search Co‐ordinator using search terms relevant to this review. The Specialised Register contains studies identified from the following sources.

1. Monthly searches of the Cochrane Central Register of Controlled Trials (CENTRAL)

2. Weekly searches of MEDLINE OVID SP

3. Handsearching of renal‐related journals and the proceedings of major renal conferences

4. Searching of the current year of EMBASE OVID SP

5. Weekly current awareness alerts for selected renal journals

6. Searches of the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov.

Studies contained in the Specialised register, are identified through search strategies for CENTRAL, MEDLINE, and EMBASE based on the scope of Cochrane Kidney and Transplant. Details of these strategies, as well as a list of handsearched journals, conference proceedings and current awareness alerts, are available in the Specialised Register section of information about the Cochrane Kidney and Transplant.

See Appendix 1 for search terms used in strategies for this review.

Searching other resources

1. Reference lists of review articles, relevant studies and clinical practice guidelines.

2. Letters seeking information about unpublished or incomplete studies to investigators known to be involved in previous studies.

Data collection and analysis

Selection of studies

The search strategy described was used to obtain titles and abstracts of studies relevant to the review. Titles and abstracts were screened independently by two authors, who discarded studies that were not applicable. However, studies and reviews thought to potentially include relevant data or information on studies were retained initially. Two authors independently assessed retrieved abstracts, and where necessary the full text of studies to determine which satisfied the inclusion criteria

Data extraction and management

Data extraction was carried out independently by two authors using a standardised data extraction form. Studies reported in languages other than English and Spanish were translated before assessment. Where more than one publication of one study existed, reports were grouped together and the publication with the most complete data was used in the analyses. Where relevant outcomes were only published in earlier versions, these data were used. Any discrepancies between published versions were to be highlighted.

Assessment of risk of bias in included studies

The following items were assessed independently by two authors using the risk of bias assessment tool (Higgins 2011) (Appendix 2).

• Was there an adequate sequence generation (selection bias)?

• Was allocation adequately concealed (selection bias)?

• Was knowledge of the allocated interventions adequately prevented during the study?

  • Participants and personnel (performance bias)

  • Outcome assessors (detection bias)

• Were incomplete outcome data adequately addressed (attrition bias)?

• Are reports of the study free of suggestion of selective outcome reporting (reporting bias)

• Was the study apparently free of other problems that could put it at a risk of bias?

Measures of treatment effect

We reported estimates for dichotomous outcomes (e.g. numbers of patients with increased SCr ≥ two‐fold from baseline level) as risk ratio (RR) with associated 95% confidence interval (CI). We calculated measures of effect by applying the random‐effects model; the Mantel‐Haenszel method (Mantel 1959) was used to pool results.

Unit of analysis issues

No cluster‐RCTs were found in the search. In studies with multiple arms (specifically those with two liposomal amphotericin B arms), data were pooled for both arms; where there were multiple arms, where one was not liposomal amphotericin B, conventional and liposomal formulation arms were included in the analysis.

Dealing with missing data

Any further information required from the original author was requested by written correspondence (e.g. e‐mailing and/or writing to corresponding author/s) and any relevant information obtained in this manner was included in the review. For all outcomes, we presented data for all randomised participants where reported; otherwise, estimates were based on complete case data.

Assessment of heterogeneity

We considered clinical heterogeneity (the degree to which RCTs varied in terms of participants, interventions, doses, duration and frequency) and statistical heterogeneity. Statistical heterogeneity was assessed using the Chi² test (P < 0.10 was considered to indicate statistically significant heterogeneity) in conjunction with the I² statistic (Higgins 2003). I² values of 25%, 50% and 75% correspond to low, medium and high levels of heterogeneity.

Assessment of reporting biases

We planned to investigate potential for small study bias by constructing and interpreting funnel plots (Higgins 2011). Our analysis found no substantial risk for small study bias.

Data synthesis

Pooled estimates of treatment effect were derived using the random‐effects model. subgroups and sensitivity analyses were conducted to prove the robustness of the model. All outcomes were dichotomous and the summary estimate as a risk ratio (RR) was presented with 95% CI.

Subgroup analysis and investigation of heterogeneity

Subgroup analyses were conducted to explore possible sources of heterogeneity. Heterogeneity in treatments could be related to prior drug(s) used, dose, duration, indication for use of amphotericin B and concomitant nephrotoxic drug use.

Sensitivity analysis

Sensitivity analyses were performed to explore the influence of the following factors on effect size:

• Repeating the analysis taking risk of bias into account

• Repeating the analysis excluding any very long or large studies to establish how much they dominate the results

• Repeating the analysis excluding each study in turn

• Excluding studies with dose or duration of liposomal amphotericin B off‐label.

Summary of findings table

Results relating to the primary outcome are presented in Table 1, which also includes an overall assessment of the evidence related to each of the main outcomes using the GRADE approach (Guyatt 2011).

Results

Description of studies

See: Characteristics of included studies and Characteristics of excluded studies

Results of the search

Our search identified 42 records of 33 studies that were assessed as full text articles. Following assessment, 26 records were excluded: 21 investigated interventions that were not relevant to this review and one study was a pharmacoeconomic. We included 12 studies that reported data for 2298 participants in the qualitative synthesis and 10 studies (2172 participants) in the meta‐analyses (Figure 1).

1.

Study flow diagram

Included studies

We included 12 studies (2298 participants) in this review (Bodhe 2002; Hamill 2010; Jadhav 2012; Johnson 2002a; Leenders 1997; Leenders 1998; Prentice 1997; Shah 2005; Sundar 2010; Sundar 2004; Thakur 2001; Walsh 1999). Ten studies (2172 participants) were meta‐analysed (see Characteristics of included studies). Two studies could not be included in the meta‐analysis: Shah 2005 was the only study which assessed CrCl in mL/min and Thakur 2001 only reported rise in SCr (is not possible to know the magnitude of the change respect baseline values).

Participants

Participants included suspected or proven systemic fungal infection (Bodhe 2002); HIV or AIDS and cryptococcal meningitis (Hamill 2010; Leenders 1997); AIDS and moderate to severe histoplasmosis (Johnson 2002a); neutropenia (Leenders 1998); neutropenia and suspected invasive fungal infection (Jadhav 2012); neutropenia and fever of unknown origin (Prentice 1997; Walsh 1999); liver transplantation (Shah 2005); and signs or symptoms of visceral leishmaniasis (Sundar 2004; Sundar 2010; Thakur 2001).

Participants' ages were reported in 10 studies (two studies specified numbers of enrolled adults and children as a categorical variable (Bodhe 2002; Prentice 1997). Two studies did not report participants' sex (Jadhav 2012; Leenders 1997) and Shah 2005reported sex by episode of treatment.

Setting

Seven studies were multicentric and five were undertaken in one centre. Ten studies were undertaken in inpatients. The countries included in studies are: India, USA, Canada, Dutch, Australia, France and UK (see Characteristics of included studies).

Design

All included studies were RCTs. Eight studies included two arms and four studies were designed around three comparisons: Hamill 2010, Jadhav 2012 and Prentice 1997 investigated two doses of liposomal amphotericin B versus conventional amphotericin B and Sundar 2004 included an amphotericin B lipid complex (ABLC) arm, a liposomal amphotericin arm and the control group with conventional amphotericin.

Interventions

The comparator in all included studies was conventional amphotericin B (sodium deoxycholate salt); four studies reported the trade name of the treatment received for the control group (Fungizone®) (Bodhe 2002; Prentice 1997; Sundar 2004; Sundar 2010); in nine studies the treatment group received liposomal amphotericin B as amBisome® (Hamill 2010; Johnson 2002a; Leenders 1997; Leenders 1998; Prentice 1997; Shah 2005; Sundar 2004; Thakur 2001; Walsh 1999); and one study (Bodhe 2002) the liposomal formulation was manufactured in‐house at the Seth GS Medical College and King Edward Memorial Hospital in Mumbai (L‐AMP‐LRC‐1).

• Three studies compared 1 mg/kg liposomal amphotericin B (L‐AMB) with 1 mg/kg conventional amphotericin B (C‐AMB) (Bodhe 2002; Jadhav 2012; Prentice 1997). Jadhav 2012 and Prentice 1997 also compared 3 mg/kg L‐AMB with 1 mg/kg C‐AMB

• Hamill 2010 compared 3 mg/kg and 6 mg/kg L‐AMB with 0.7 mg/kg C‐AMB

• Walsh 1999 compared 3 mg/kg L‐AMB with 0.6 mg/kg C‐AMB

• Johnson 2002a compared 3 mg/kg L‐AMB with 0.7 mg/kg C‐AMB

• Leenders 1997 compared 4 mg/kg L‐AMB with 0.7 mg/kg C‐AMB

• Leenders 1998 compared 5 mg/kg L‐AMB with 1 mg/kg C‐AMB

• Shah 2005 compared 50 mg/d L‐AMB with 15 mg/d C‐AMB

• Sundar 2004 compared 2 mg/kg L‐AMB with 1 mg/kg C‐AMB

• Sundar 2010 compared 10 mg/kg L‐AMB with 1 mg/kg C‐AMB

• Thakur 2001 compared 15 mg/kg L‐AMB with 1 mg/kg C‐AMB.

Premedication to prevent infusion‐related reactions was permitted in eight studies (Jadhav 2012; Johnson 2002a; Leenders 1997; Leenders 1998; Prentice 1997; Sundar 2004; Sundar 2010; Walsh 1999), however Jadhav 2012, Johnson 2002a and Walsh 1999 allowed premedication before the first dose of the study medication. Hamill 2010 used only IV saline load and Bodhe 2002 and Thakur 2001 did not report if premedication was allowed.

Excluded studies

We excluded 26 records of 21 studies. Of these, 20 studies investigated interventions or comparators outside the scope of this review, and one was a pharmacoeconomic analysis (see Characteristics of excluded studies).

Risk of bias in included studies

Overall, study quality was suboptimal. The risk of bias in included studies was low or unclear for most domains. However, blinding of participants and personnel, blinding of outcome assessment and other bias (funding) tended to have a high risk of bias. See Figure 2 and Figure 3.

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies

3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study

Allocation

Seven studies clearly described random sequence generation, and were assessed at low risk of bias (Hamill 2010; Jadhav 2012; Prentice 1997; Shah 2005; Sundar 2004; Sundar 2010; Walsh 1999). Five studies did not include reporting of sequence generation methods and were deemed at unclear risk (Bodhe 2002; Johnson 2002a; Leenders 1997; Leenders 1998; Thakur 2001).

Five studies reported using sealed or closed envelopes, but did not report if envelopes were opaque and sequentially numbered, and were assessed as unclear risk of bias for this domain (Johnson 2002a; Leenders 1997; Leenders 1998; Shah 2005; Sundar 2004). Four studies did not report on the method of allocation concealment, and were also considered being at unclear risk of bias (Bodhe 2002; Hamill 2010; Thakur 2001; Walsh 1999). Three studies adequately reported concealment methods (Jadhav 2012; Prentice 1997; Sundar 2010) and were assessed at low risk of bias.

Blinding

Three studies (Hamill 2010, Johnson 2002a; Walsh 1999) were double‐blinded and assessed at low risk of bias. Nine studies were assessed at high risk of bias in relation to blinding; of these, four were described as open (Bodhe 2002; Sundar 2004; Sundar 2010; Shah 2005) and five studies did not report blinding participants or study personnel (Jadhav 2012; Leenders 1997; Leenders 1998; Prentice 1997; Thakur 2001).

None of the included studies provided statements about blinding of outcomes assessment. All were assessed at high risk of bias for this domain.

Incomplete outcome data

There was no missing outcome data in six studies (Hamill 2010; Leenders 1997; Leenders 1998; Sundar 2004; Shah 2005; Thakur 2001) and were assessed at low risk of bias. In one study, assessment was not possible due to the lack of information, so it was also assessed as at unclear risk of bias (Walsh 1999). In three studies (Bodhe 2002; Jadhav 2012; Johnson 2002a) missing outcome data were balanced in numbers across intervention groups and were assessed at low risk of bias. In Prentice 1997, missing outcome data for efficacy (an outcome different of the relevant to this review) were explained and the intention‐to‐treat analysis (ITT) was performed, and in one study no patient data were lost in follow‐up (Sundar 2010); these studies were also judged at low risk of bias of attrition bias.

Selective reporting

Most study protocols were unobtainable, so judgements were based on the agreement between the methods and the results' of the study's reports. One study was classified at high risk of bias for this domain (Jadhav 2012); the study failed to include results for a key outcome (nephrotoxicity) that would be expected to be reported, given that in methods, the authors' statement is that haematological and biochemical parameters were monitored twice weekly, including SCr levels.

Other potential sources of bias

Bodhe 2002 was assessed as unclear risk of other sources of bias due to insufficient information to permit judgement.

Pharmaceutical funding was identified in seven studies (Hamill 2010; Johnson 2002a; Leenders 1997; Leenders 1998; Prentice 1997; Shah 2005; Walsh 1999) and these studies were assessed at high risk of bias.

Sundar 2004 and Sundar 2010 were assessed at high risk of bias due to an independent sample t‐test, being used to detect differences in clinical and laboratory results (including change in creatinine) for both study groups. Tests for normal distribution were not shown or reported.

Although there were relatively few studies included in the meta‐analysis (10), a funnel plot was constructed to assess the potential existence of small study bias (Figure 4). The plot showed a lack of small studies in favour of the conventional formulation.

4.

Funnel plot of comparison: 1 Comparison between liposomal and conventional amphotericin B, outcome: 1.1 Increase in serum creatinine level ≥ two‐fold increase from baseline level

Effects of interventions

See: Table 1

Increase in serum creatinine

Ten studies (Bodhe 2002; Hamill 2010; Jadhav 2012; Johnson 2002a; Leenders 1997; Leenders 1998; Prentice 1997; Sundar 2004; Sundar 2010; Walsh 1999) compared liposomal amphotericin B with conventional amphotericin B and were included in the meta‐analysis. There was a significant increase in SCr level: ≥ two‐fold from baseline level with conventional amphotericin B compared to liposomal amphotericin B (Analysis 1.1 (10 studies, 2172 participants): RR 0.49, 95% CI 0.40 ‐ 0.59; I² = 0%).

1.1. Analysis.

Comparison 1 Liposomal versus conventional amphotericin B, Outcome 1 Increase in serum creatinine level ≥ two‐fold increase from baseline level.

Shah 2005 and Thakur 2001 were excluded from the meta‐analysis. Shah 2005 (87 participants) reported the mean CrCl at the start and end of therapy with liposomal and conventional amphotericin B; for liposomal amphotericin B, the starting mean CrCl (mL/min) was 33.65 and at the end was 40.61 (P = 0.661). In the conventional group, the initial mean CrCl was 38.65 and at the end was 44.23 (P = 0.805). Thakur 2001 (34 participants) reported rise in SCr of 23% with conventional amphotericin B and 0% with liposomal amphotericin B. We were unsuccessful in our attempts to contact the authors for clarification.

Infusion‐related reactions

Nine studies (Hamill 2010; Jadhav 2012; Johnson 2002a; Leenders 1997; Leenders 1998; Sundar 2004; Sundar 2010; Thakur 2001; Walsh 1999) compared liposomal amphotericin B with conventional amphotericin B (sodium deoxycholate). There was significant decrease in all infusion‐related reactions in the liposomal group compared with the conventional group (Analysis 1.2): fever (4 studies, 1092 participants): RR 0.39, 95% CI 0.28 to 0.55; I² = 32%); chills and/or rigours (5 studies, 1081 participants): RR 0.27, 95% CI 0.15 to 0.48; I² = 75%); fever and/or rigours (2 studies, 720 participants): RR 0.68, 95% CI 0.52 to 0.90; I² = 58%); nausea (6 studies, 1187 participants): RR 0.50, 95% CI 0.35 to 0.72; I² = 0%); and vomiting (3 studies, 1019 participants): RR 0.51, 95% CI 0.27 to 0.95; I² = 61%).

1.2. Analysis.

Comparison 1 Liposomal versus conventional amphotericin B, Outcome 2 Infusion‐related reactions (as determined by the investigators).

Sensitivity analyses

Sensitivity analyses were performed to explore the influence of the following factors on effect size.

Repeating the analyses taking account of risk of bias: in studies at high risk of bias the difference is not significant (Analysis 2.1; RR 0.16, 95% CI 0.03 to 1.00; I² = 0%); for studies at unclear risk of bias and low risk of bias, the difference is significant (Analysis 2.2; RR 0.39, 95% CI 0.27 to 0.55; I² = 0%) and (Analysis 2.3; RR 0.55, 95% CI 0.44 to 0.69; I² = 0%).

2.1. Analysis.

Comparison 2 Sensitivity analysis for the primary outcome, Outcome 1 Overall risk of bias: high.

2.2. Analysis.

Comparison 2 Sensitivity analysis for the primary outcome, Outcome 2 Overall risk of bias: unclear.

2.3. Analysis.

Comparison 2 Sensitivity analysis for the primary outcome, Outcome 3 Overall risk of bias: low.

Repeating the analyses excluding any very long or large studies to establish how much they dominate the results: We excluded five large studies (1857 participants): Hamill 2010; Prentice 1997; Sundar 2004; Sundar 2010; and Walsh 1999. The difference in the increase in SCr level: ≥ two‐fold from baseline level between the two treatments remained significant (Analysis 2.4; RR 0.36, 95% CI 0.22 to 0.58).

2.4. Analysis.

Comparison 2 Sensitivity analysis for the primary outcome, Outcome 4 Excluding large studies.

Repeating the meta‐analysis as many times as RCT included: a sensitivity analysis was conducted as many times as studies included, excluding one study each time. The total effect estimate was statistically significant (Analysis 2.5; RR 0.48, CI 95% 0.45 to 0.52, I² 0%).

2.5. Analysis.

Comparison 2 Sensitivity analysis for the primary outcome, Outcome 5 Sensitivity analysis excluding one study at a time.

Excluding studies with dose or duration of liposomal amphotericin B off‐label: Eight studies were compared and the total effect estimate was also statistically significant (Analysis 2.6; RR 0.49, 95% CI 0.41 to 0.59, I² = 0%).

2.6. Analysis.

Comparison 2 Sensitivity analysis for the primary outcome, Outcome 6 Excluding studies with off‐label use of liposomal amphotericin B.

Subgroup analysis

Two subgroup analyses were also conducted. One looked at the number of days of treatment (fewer than five days and at least 11 days); the second was based in diagnostic or clinical indication of amphotericin B (IFI/neutropenia, AIDS or HIV infected + cryptococcal meningitis or histoplasmosis and visceral leishmaniasis). Only in the subgroup of fewer than five days of treatment and visceral leishmaniasis (both included the same studies), the difference was not significant (Analysis 3.1.1 (2 studies, 565 participants): RR 0.33, 95% CI 0.06 to 1.68; I² = 0%). There is a need to clarify that these two studies (Sundar 2004; Sundar 2010) used a dose and therapy duration different from that recommended for visceral leishmaniasis. The first used a dose of 2 mg/kg/d in five consecutive infusions, and the second used a unique dose of amphotericin B of 10 mg/kg. The recommendations state 3 mg/kg/d on days one to five and 3 mg/kg/d on days 14 and 21 for immunocompetent patients, and 4 mg/kg/d on days one to five and 4 mg/kg/d on days 10, 17, 24, 31 and 38 for immunocompromised patients.

3.1. Analysis.

Comparison 3 Subgroup analysis, Outcome 1 Duration of the treatment.

Discussion

Summary of main results

We included 12 studies (2298 participants) in this review. Of these, 10 were able to be meta‐analysed (2172 participants). Bodhe 2002 evaluated patients with suspected or proven IFI; Hamill 2010 and Leenders 1997 included patients with HIV or AIDS and cryptococcal meningitis; Johnson 2002a also included patients with AIDS but with moderate to severe histoplasmosis. Four studies included patients with neutropenia; Leenders 1998 included patients with neutropenia only; participants in Jadhav 2012 had suspected IFI; Prentice 1997 and Walsh 1999 included neutropenic patients with pyrexia or fever of unknown origin; and Sundar 2004 and Sundar 2010 included patients with signs or symptoms of visceral leishmaniasis.

The primary outcome sought was SCr level increase at least two‐fold from baseline. Ten studies were included in the meta‐analysis for this outcome. For the secondary outcome (infusion‐related reactions) it was possible to obtain information about fever, chills and/or rigours, fever and/or rigours, nausea and vomiting. In all the cases, there were significantly less infusion‐related reactions with liposomal amphotericin B compared to conventional amphotericin B.

Overall, risk of bias in included studies was low or unclear for most domains. However, blinding of participants and personnel, blinding of outcome assessment and other bias (funding) tended to have a high risk of bias. Table 1 provides a concise overview and synthesis of the volume and quality of the evidence for the comparison between liposomal and conventional amphotericin B respect to the increase in SCr level ≥ two‐fold from baseline level. Publication bias was not detected and several sensitivity analyses were performed to check the robustness of the effect estimate.

Overall completeness and applicability of evidence

The studies included in this review vary in several aspects (patients (diagnosis or clinical condition), dose of amphotericin received (treatment group and control group), country, risk of bias, methods). However, the patients that would typically receive this treatment are represented. Although two eligible studies were excluded from the meta‐analysis, the included studies allowed us to obtain a clear response for both the primary and secondary outcomes.

The results reported in the included studies did not enable post hoc comparisons of renal replacement therapy or end‐stage kidney disease. Although the elevation in SCr levels is considered a clinical manifestation of AKI and it is associated with an increased risk of morbidity and mortality (Yang 2014), it is a surrogate outcome with poor specificity and sensitivity with regard to recognition of the early period of AKI (Urbschat 2011), and it is uncertain whether reductions in SCr will translate into clinically important benefits.

Quality of the evidence

The results of this review are robust; 10 studies (2172 participants) were included in the meta‐analysis for the primary outcome of which the overall risk of bias was: high for two studies (Jadhav 2012; Sundar 2004), uncertain for six studies (Bodhe 2002; Leenders 1997; Leenders 1998; Prentice 1997; Johnson 2002a; Sundar 2010 and low for the other two (Hamill 2010; Walsh 1999). The main key methodological limitation of the included studies was the blinding of participants and personnel, blinding of outcome assessment and other bias (funding). Subgroup (Analysis 3.1; Analysis 3.2) and sensitivity analyses (Analysis 2.2; Analysis 2.3; Analysis 2.4; Analysis 2.5) indicated that the magnitude and direction of the effect did not change, except in studies at high risk of bias (Analysis 2.1) and when Sundar 2004 and Sundar 2010 were analysed separately. These studies were conducted in patients with visceral leishmaniasis, with treatment doses and duration that differed from the liposomal formulation product information (off‐label).

3.2. Analysis.

Comparison 3 Subgroup analysis, Outcome 2 Diagnostic or clinical indication.

Johansen 2014 concluded that it was unclear if these formulations (all lipid‐based amphotericin B formulations) offered advantages, if conventional amphotericin B is administered under optimal circumstances; in nine of the 10 included studies in this review, loading saline or premedication with antipyretics or antihistamines were allowed for prevention of infusion‐related reactions and nephrotoxicity in the control groups and in some of the treatment groups.

Potential biases in the review process

In addition to electronic searching of bibliographic databases without language restriction, other strategies included contacting study authors and accessing registers of ongoing studies. Although searches were comprehensive, and we did not find evidence of publication bias (Figure 4), two potentially eligible studies were excluded because we were unable to extract or obtain key outcomes data. However it is unlikely that addition of unpublished data would substantially change our overall review findings.

Prentice 1997 reported nephrotoxicity incidence and accounted for use of concomitant nephrotoxic drugs with amphotericin B. This study did not find any differences in the subgroup of patients receiving concomitant nephrotoxic drugs (P = 0.11). This finding is relevant because it is pharmacologically and clinically plausible that treatment involving more than one nephrotoxic drug increases the likelihood of kidney failure, and this should be considered in treatment choices.

Agreements and disagreements with other studies or reviews

Our findings agree with Johansen 2014, who reported decreased nephrotoxicity (defined as 100% increase in SCr) associated with the liposomal formulation (RR 0.45, 95% CI 0.37 to 0.54). Mistro 2012 also suggested that liposomal amphotericin B reduced nephrotoxicity (RR 0.48; 95% CI 0.36 to 0.64). Mistro 2012 included five studies all of which were included in our review. Johansen 2014 included three studies, all of which were also included in this review.

Authors' conclusions

Implications for practice.

Liposomal amphotericin B is less nephrotoxic than conventional amphotericin B, but the high cost precludes routine use in many healthcare settings. Patients being considered to receive the liposomal formulation should be selected on the basis of diagnosis, concomitant nephrotoxic drugs, duration of therapy and dose of amphotericin B. In most included studies, premedication or saline loading was permitted to prevent infusion‐related reactions and nephrotoxicity, therefore, renal effect of each amphotericin B formulation is hard to establish.

Implications for research.

Larger, well‐designed and adequately powered studies comparing conventional amphotericin B with liposomal amphotericin B (including other lipid based formulations) and final endpoints are needed to inform clinical decision making. If this approach is considered to be unethical, new biomarkers (neutrophil gelatinase‐associated lipocalin, cystatin‐C, N‐acetyl‐β‐d‐glucosaminidase, kidney injury molecule‐1, and α₁‐microglobulin, among others) with better performance in early stage AKI should be considered for incorporation in clinical practice and investigated in few studies.

Investigators should also consider adopting the recommendations of the CONSORT statement (Moher 2010) in the planning and conduct of new studies.

Analyses of cost‐effectiveness would beneficial; there are currently insufficient efficacy, safety and quality data to inform cost‐effectiveness analyses.

Appendices

Appendix 1. Electronic search strategies

Database	Search terms
CENTRAL	1. MeSH descriptor Amphotericin B, this term only 2. deoxycholate*:ti,ab,kw 3. (#1 AND #2) in trials 4. amphotericin b deoxycholate*:ti,ab,kw in Trials 5. fungizone:ti,ab,kw or fungisone:ti,ab,kw in Trials 6. ("c‐amb":ti,ab,kw or amb:ti,ab,kw) in Trials 7. conventional next amphotericin b:ti,ab,kw in Trials 8. ambd:ti,ab,kw in Trials 9. (#3 OR #4 OR #5 OR #6 OR #7 OR #8) 10. MeSH descriptor Amphotericin B, this term only 11. liposom*:ti,ab,kw 12. (#10 AND #11) 13. "l‐amb":ti,ab,kw in Trials 14. ambisome:ti,ab,kw in Trials 15. liposomal amphotericin b:ti,ab,kw in Trials 16. (#12 OR #13 OR #14 OR #15) 17. (#9 AND #16)
MEDLINE	1. Amphotericin B/ and Deoxychol$.tw. 2. Amphotericin B deoxycholate$.tw. 3. fungizone.tw. 4. ("c‐amb" or amb).tw. 5. (conventional adj2 amphotericin b).tw. 6. ambd.tw. 7. polyene antifungal$.tw. 8. or/1‐7 9. Amphotericin B/ and Liposom$.tw. 10. ("l‐amb" or ablc).tw. 11. ambisome.tw. 12. liposomal amphotericin b.tw. 13. polyene antifungal$.tw. 14. or/9‐13 15. and/8,14
EMBASE	1. Amphotericin B Deoxycholate/ 2. Amphotericin B deoxycholate$.tw. 3. fungizone.tw. 4. ("c‐amb" or amb).tw. 5. (conventional adj2 amphotericin b).tw. 6. ambd.tw. 7. polyene antifungal$.tw. 8. or/1‐7 9. Amphotericin B/ and Liposom$.tw. 10. ("l‐amb" or ablc).tw. 11. ambisome.tw. 12. liposomal amphotericin b.tw. 13. polyene antifungal$.tw. 14. or/9‐13 15. and/8,14

Appendix 2. Risk of bias assessment tool

Potential source of bias	Assessment criteria
Random sequence generation Selection bias (biased allocation to interventions) due to inadequate generation of a randomised sequence	Low risk of bias: Random number table; computer random number generator; coin tossing; shuffling cards or envelopes; throwing dice; drawing of lots; minimization (minimization may be implemented without a random element, and this is considered to be equivalent to being random).
	High risk of bias: Sequence generated by odd or even date of birth; date (or day) of admission; sequence generated by hospital or clinic record number; allocation by judgement of the clinician; by preference of the participant; based on the results of a laboratory test or a series of tests; by availability of the intervention.
	Unclear: Insufficient information about the sequence generation process to permit judgement.
Allocation concealment Selection bias (biased allocation to interventions) due to inadequate concealment of allocations prior to assignment	Low risk of bias: Randomisation method described that would not allow investigator/participant to know or influence intervention group before eligible participant entered in the study (e.g. central allocation, including telephone, web‐based, and pharmacy‐controlled, randomisation; sequentially numbered drug containers of identical appearance; sequentially numbered, opaque, sealed envelopes).
	High risk of bias: Using an open random allocation schedule (e.g. a list of random numbers); assignment envelopes were used without appropriate safeguards (e.g. if envelopes were unsealed or non‐opaque or not sequentially numbered); alternation or rotation; date of birth; case record number; any other explicitly unconcealed procedure.
	Unclear: Randomisation stated but no information on method used is available.
Blinding of participants and personnel Performance bias due to knowledge of the allocated interventions by participants and personnel during the study	Low risk of bias: No blinding or incomplete blinding, but the review authors judge that the outcome is not likely to be influenced by lack of blinding; blinding of participants and key study's personnel ensured, and unlikely that the blinding could have been broken.
	High risk of bias: No blinding or incomplete blinding, and the outcome is likely to be influenced by lack of blinding; blinding of key study participants and personnel attempted, but likely that the blinding could have been broken, and the outcome is likely to be influenced by lack of blinding.
	Unclear: Insufficient information to permit judgement
Blinding of outcome assessment Detection bias due to knowledge of the allocated interventions by outcome assessors.	Low risk of bias: No blinding of outcome assessment, but the review authors judge that the outcome measurement is not likely to be influenced by lack of blinding; blinding of outcome assessment ensured, and unlikely that the blinding could have been broken.
	High risk of bias: No blinding of outcome assessment, and the outcome measurement is likely to be influenced by lack of blinding; blinding of outcome assessment, but likely that the blinding could have been broken, and the outcome measurement is likely to be influenced by lack of blinding.
	Unclear: Insufficient information to permit judgement
Incomplete outcome data Attrition bias due to amount, nature or handling of incomplete outcome data.	Low risk of bias: No missing outcome data; reasons for missing outcome data unlikely to be related to true outcome (for survival data, censoring unlikely to be introducing bias); missing outcome data balanced in numbers across intervention groups, with similar reasons for missing data across groups; for dichotomous outcome data, the proportion of missing outcomes compared with observed event risk not enough to have a clinically relevant impact on the intervention effect estimate; for continuous outcome data, plausible effect size (difference in means or standardized difference in means) among missing outcomes not enough to have a clinically relevant impact on observed effect size; missing data have been imputed using appropriate methods.
	High risk of bias: Reason for missing outcome data likely to be related to true outcome, with either imbalance in numbers or reasons for missing data across intervention groups; for dichotomous outcome data, the proportion of missing outcomes compared with observed event risk enough to induce clinically relevant bias in intervention effect estimate; for continuous outcome data, plausible effect size (difference in means or standardized difference in means) among missing outcomes enough to induce clinically relevant bias in observed effect size; ‘as‐treated’ analysis done with substantial departure of the intervention received from that assigned at randomisation; potentially inappropriate application of simple imputation.
	Unclear: Insufficient information to permit judgement
Selective reporting Reporting bias due to selective outcome reporting	Low risk of bias: The study protocol is available and all of the study’s pre‐specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre‐specified way; the study protocol is not available but it is clear that the published reports include all expected outcomes, including those that were pre‐specified (convincing text of this nature may be uncommon).
	High risk of bias: Not all of the study’s pre‐specified primary outcomes have been reported; one or more primary outcomes is reported using measurements, analysis methods or subsets of the data (e.g. subscales) that were not pre‐specified; one or more reported primary outcomes were not pre‐specified (unless clear justification for their reporting is provided, such as an unexpected adverse effect); one or more outcomes of interest in the review are reported incompletely so that they cannot be entered in a meta‐analysis; the study report fails to include results for a key outcome that would be expected to have been reported for such a study.
	Unclear: Insufficient information to permit judgement
Other bias Bias due to problems not covered elsewhere in the table	Low risk of bias: The study appears to be free of other sources of bias.
	High risk of bias: Had a potential source of bias related to the specific study design used; stopped early due to some data‐dependent process (including a formal‐stopping rule); had extreme baseline imbalance; has been claimed to have been fraudulent; had some other problem.
	Unclear: Insufficient information to assess whether an important risk of bias exists; insufficient rationale or evidence that an identified problem will introduce bias.

Data and analyses

Comparison 1. Liposomal versus conventional amphotericin B.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Increase in serum creatinine level ≥ two‐fold increase from baseline level	10	2172	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.40, 0.59]
2 Infusion‐related reactions (as determined by the investigators)	9		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Fever	4	1092	Risk Ratio (M‐H, Random, 95% CI)	0.39 [0.28, 0.55]
2.2 Chills and/or rigours	5	1081	Risk Ratio (M‐H, Random, 95% CI)	0.27 [0.15, 0.48]
2.3 Fever and/or rigours	3	720	Risk Ratio (M‐H, Random, 95% CI)	0.68 [0.52, 0.90]
2.4 Nausea	6	1187	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.35, 0.72]
2.5 Vomiting	3	1019	Risk Ratio (M‐H, Random, 95% CI)	0.51 [0.27, 0.95]

Comparison 2. Sensitivity analysis for the primary outcome.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Overall risk of bias: high	2	218	Risk Ratio (M‐H, Random, 95% CI)	0.16 [0.03, 1.00]
2 Overall risk of bias: unclear	6	1000	Risk Ratio (M‐H, Random, 95% CI)	0.39 [0.27, 0.55]
3 Overall risk of bias: low	2	954	Risk Ratio (M‐H, Random, 95% CI)	0.55 [0.44, 0.69]
4 Excluding large studies	5	315	Risk Ratio (M‐H, Random, 95% CI)	0.36 [0.22, 0.58]
5 Sensitivity analysis excluding one study at a time	10		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
5.1 Excluded Leenders 1997	9	2144	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.40, 0.59]
5.2 Excluded Prentice1997	9	1834	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.38, 0.61]
5.3 Excluded Leenders 1998	9	2066	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.41, 0.61]
5.4 Excluded Walsh 1999	9	1485	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.32, 0.55]
5.5 Excluded Johnson 2002a	9	2095	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.41, 0.60]
5.6 Excluded Bodhe 2002	9	2133	Risk Ratio (M‐H, Random, 95% CI)	0.46 [0.36, 0.58]
5.7 Excluded Sundar 2004	9	2019	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.41, 0.59]
5.8 Excluded Hamill 2010	9	1915	Risk Ratio (M‐H, Random, 95% CI)	0.45 [0.34, 0.58]
5.9 Excluded Sundar 2010	9	1770	Risk Ratio (M‐H, Random, 95% CI)	0.47 [0.37, 0.59]
5.10 Excluded Jadhav 2012	9	2117	Risk Ratio (M‐H, Random, 95% CI)	0.48 [0.39, 0.59]
6 Excluding studies with off‐label use of liposomal amphotericin B	8	1607	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.41, 0.59]

Comparison 3. Subgroup analysis.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Duration of the treatment	10	2172	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.40, 0.59]
1.1 Fewer than 5 days of treatment	2	565	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.06, 1.68]
1.2 At least 11 days of treatment	8	1607	Risk Ratio (M‐H, Random, 95% CI)	0.49 [0.41, 0.59]
2 Diagnostic or clinical indication	10		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Invasive fungal infection/neutropenia	5	1235	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.40, 0.62]
2.2 AIDS or HIV infected + cryptococcal meningitis or histoplasmosis	3	372	Risk Ratio (M‐H, Random, 95% CI)	0.13 [0.03, 0.50]
2.3 Visceral leishmaniasis	2	565	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.06, 1.68]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Bodhe 2002.

Methods	Study design: open‐label parallel RCT Study duration: NS
Participants	Setting: inpatients Country: India Health status: patients of any age or sex with clinically and radiologically suspected and histologically or microbiologically proven systemic fungal infection Number (randomised/analysed): treatment group (20/17); control group (19/17) Age (number of participants): age cut‐off was not defined in the study Treatment group: adults (8); children (2); neonates (10) Control group: adults (6); children (4); neonates (9) Sex (M/F): treatment group (11/9); control group (10/9) Exclusion criteria: pregnant women; moribund; other antifungal drugs co‐administered
Interventions	Treatment group Liposomal amphotericin B (L‐AMP‐LRC‐1) Dose: 1 mg/kg Duration: 3 to 8 weeks Frequency: daily Administration: IV, diluted with normal saline IP. 60 minutes infusion (three escalating doses were administered the first 3 days of therapy) Control group Conventional amphotericin B (Fungizone) Dose: 1 mg/kg (a test dose of 1 mg administered over 20 to 30 minutes) Duration: 3 to 8 weeks Frequency: daily Administration: IV, diluted with 5% dextrose over 2 to 6 hours
Outcomes	Infusion‐related reactions Fever Chills Headache Backache Nausea Vomiting Palpitation Apnoea Tachypnoea Dizziness Nephrotoxicity (SCr level)
Notes	Support and financial assistance: Department of Biotechnology, Indian Government
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Insufficient information to permit judgement
Allocation concealment (selection bias)	Unclear risk	No information on method included
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding of personnel and patients
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	Missing outcome data balanced between intervention groups (three in the intervention group and two in the control group)
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes were reported
Other bias	Unclear risk	Insufficient information to permit judgement

Hamill 2010.

Methods	Study design: parallel RCT with three arms Study duration: June 1995 to April 1998
Participants	Setting: inpatients, multicentre Countries: USA and Canada (35 institutions) Health status: patients with AIDS and acute cryptococcal meningitis Primary disease (number): treatment group 1 (81, 94.3%); treatment group 2 (86, 91.5%); control group (85, 97.7%) Mean baseline SCr (mg/dL): treatment group 1 (1.1); treatment group 2 (1.0); Control group (1.0) Number: treatment group 1 (86); treatment group 2 (94); control group (87) Mean age, range (years): treatment group 1 (38.7, 22 to 61); treatment group 2 (40.1, 21 to 68); control group (38.5, 10 to 59) Sex (M/F): treatment group 1 (76/10); treatment group 2 (81/13); control group (81/6) Exclusion criteria: received cumulative dose > 12.4 mg/kg amphotericin B, > 15 mg/kg amphotericin B lipid complex or > 1200 mg fluconazole; other systemic fungal infection; SCr > 2 x ULN (1.2 mg/dL); serum hepatic transaminase level > 10 x ULN; pregnant or lactating women; aged < 1 month; history of anaphylactic reaction to amphotericin B
Interventions	Treatment group 1 Liposomal amphotericin B (AmBisome) Dose: 3 mg/kg/d Duration: minimum 11 days (11 to 21 day induction. Targeted induction phase was 14 days of uninterrupted therapy) Frequency: daily Administration: IV, infusion 2 to 4 hours Treatment group 2 Liposomal amphotericin B Dose: 6 mg/kg/d Duration: minimum 11 days (11 to 21 day induction. Targeted induction phase was 14 days of uninterrupted therapy) Frequency: daily Administration: IV, infusion 2 to 4 hours Control group Amphotericin B Dose: 0.7 mg/kg/d Duration: minimum 11 days (11 to 21 day induction. Targeted induction phase was 14 days of uninterrupted therapy) Frequency: daily Administration: IV, infusion 2 to 4 hours
Outcomes	Infusion–related reactions Increase in temperature ≥ 1° C Chills and/or rigours Nausea Vomiting Respiratory system (any adverse event) Nephrotoxicity (SCr)
Notes	IV saline loading was permitted in all arms Funding source included Astellas Pharma
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation via centralised telephone centre
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Each infusion bag or bottle and tubing was modified to prevent distinction between liposomal amphotericin B and amphotericin B, because the drugs differ in appearance
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing outcome data
Selective reporting (reporting bias)	Low risk	Include all expected outcomes
Other bias	High risk	Financial support: Astellas Pharma and, in part, the Department of Veterans Affairs and the National Center for Research Resources, General Clinical Research Center Manuscript preparation. Astellas Pharma provided assistance with study design and execution, as well as statistical analyses

Jadhav 2012.

Methods	Study design: parallel RCT with three arms Study duration: July 2006 and June 2008
Participants	Setting: inpatients Country: India (multicentre), two centres Health status: men and women aged from 2 and 60 years with neutropenia (ANC ≤ 500 neutrophil mm3); suspected fungal infection demonstrated by fever after at least 96 hours of broad spectrum antibacterial therapy Number: treatment group 1 (22); treatment group 2 (23); control group (20) Mean age ± SD (years): treatment group 1 (16.2 ± 13.4); treatment group 2 (16.0 ± 10.9); control group (22.7 ± 16.2) Sex (M/F): not reported Exclusion criteria: pregnancy; evidence of liver disease, SGOT or SGPT > 10 x ULN or total bilirubin > 5 x ULN; known allergy to any product containing amphotericin B; patients with concomitant medical condition (e.g. HIV), whose participation may create unacceptable additional risk; patients with life expectancy < 1 week; patients who had evidence of fungal infection before enrolment (documented invasive fungal infections)
Interventions	Treatment group 1 Liposomal amphotericin B Dose: 1 mg/kg/d Duration (mean ± SD): 17.0 ± 9.8 days Frequency: daily Administration: IV, test dose at day 1 of 1 mL over 5 min. Infusions over 1 hour Treatment group 2 Liposomal amphotericin B Dose: 3 mg/kg/d Duration (mean ± SD): 16.2 ± 8.3 days Frequency: daily Administration: IV, test dose at day 1 of 1 mL over 5 min. Infusions over 1 hour Before administration the study drug liposomal amphotericin B (stored at 2 to 8°C) was sonicated using a bath sonicator with thermostat, at ambient temperature for 45 min to convert multilamellar vesicles into unilamellar vesicles Control group Conventional amphotericin B Dose: 1 mg/kg/d Duration (mean ± SD): 14.7 ± 10.7 days Frequency: daily Administration: IV, test dose at day 1. Infusions over 4 hours
Outcomes	Infusion–related reactions Vomiting Nausea Fever Chills, rigours Headache Nephrotoxicity (SCr)
Notes	The study author was contacted to seek numbers of patients whose SCr increased two‐fold from base‐line in each study arm (one patient in the control group and none in the intervention group). The investigators determined which, if any, appropriate premedication should be used to prevent IRR on subsequent infusions. The premedication, if required, was started from treatment day 2 of the study i.e. 24 hours after first dose of study medication Funding source: Department of Biotechnology, New Delhi
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated central randomisation in blocks of six
Allocation concealment (selection bias)	Low risk	Sealed envelopes with a unique centre and identification number. Participants who met the inclusion criteria were assigned ID numbers and randomised as per randomisation plan
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding of personnel and participants
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	Missing outcome data balanced among interventions
Selective reporting (reporting bias)	High risk	The study did not include results for a key outcome (i.e. nephrotoxicity) that would be expected, given that the methods reported recording of haematological and biochemical parameters including SCr twice weekly
Other bias	Low risk	The study appears to be free of other sources of bias Funding source: Department of Biotechnology, New Delhi

Johnson 2002a.

Methods	Study design: parallel RCT Study duration: not reported
Participants	Setting: 21 sites of the US National Institute of Allergy and Infectious Diseases Mycoses Study Group Country: USA Health status: AIDS and moderate to severe disseminated histoplasmosis Median CD4 cell count (range),109 cells/L: treatment 0.018 (0.001 to 0.182); control 0.018 (0.004 to 0.094) Number: treatment group (53); control group (24) Age: median 33 years (range, 16 to 68 years) Sex: only reported for 73 male participants who were evaluated for efficacy Exclusion criteria: SCr > 2 x ULN; other uncontrolled opportunistic infections or malignant disease; treated ≥ 3 days with ketoconazole, itraconazole, fluconazole or amphotericin B
Interventions	Treatment group Liposomal amphotericin B (AmBisome) Dose: 3.0 mg/kg Duration: 2 weeks Frequency: daily Administration: IV infusion over 2 hours Control group Amphotericin B Dose: 0.7 mg/kg Duration: 2 weeks Frequency: daily Administration: IV infusion over 2 hours
Outcomes	Acute infusion‐related reactions Discontinuation of therapy due toxicity Nephrotoxicity (SCr)
Notes	Pre‐medication not permitted before first dose but could be used for later doses; infusion times modified depending on side effects Supported by a research grant from the National Institute of Allergy and Infectious Diseases (NIAID) and the National Center for Research Resources, and by Gilead Sciences, Inc
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method not reported
Allocation concealment (selection bias)	Unclear risk	Sealed envelope
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double blinded
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	Missing outcome data balanced between intervention groups
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes reported
Other bias	High risk	This study was supported by a research grant from the National Institute of Allergy and Infectious Diseases (NIAID) and the National Center for Research Resources, and by Gilead Sciences, Inc

Leenders 1997.

Methods	Study design: parallel RCT Study duration: June 1992 to June 1995
Participants	Setting: inpatients Countries: Netherlands and Australia Health status: HIV‐infected patients aged > 18 years with primary episode of cryptococcal meningitis Number: treatment group (15); control group (13) Median age, range (years): treatment group (40, 29 to 55); control group (41, 28 to 49) Sex: not reported Exclusion criteria: previous treatment for cryptococcal infection; SCr > 250 µmol/L
Interventions	Treatment group Liposomal amphotericin B (AmBisome) administered at full dose with no escalation schedule Dose: 4.0 mg/kg Duration: 3 weeks Frequency: daily Administration: IV infusion over 45 min Control group Conventional amphotericin B Dose: 0.7 mg/kg Duration: 3 weeks Frequency: daily Administration: IV infusion over 6 hours
Outcomes	Acute infusion‐related reactions Chills Nausea Other acute reactions Nephrotoxicity (SCr level before treatment and twice weekly thereafter)
Notes	This study was supported by a grant from NeXstar Pharmaceuticals, Inc, San Dimas, California, USA Duration of follow‐up 3 weeks followed by fluconazole 400 mg daily for 7 weeks Sodium supplementation prior to conventional amphotericin B infusion to prevent nephrotoxicity was recommended and administration of concomitant medication to prevent acute reactions permitted
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method not reported
Allocation concealment (selection bias)	Unclear risk	Sealed envelope
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing outcome data
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes reported
Other bias	High risk	The study was supported by grants from NeXstar Pharmaceuticals Inc, San Dimas, California, USA

Leenders 1998.

Methods	Study design: parallel RCT Study duration: January 1992 to January 1996
Participants	Setting: inpatients aged > 16 years Countries: Netherlands and France Median, range baseline SCr (µmol/L): treatment group (83, 21 to 195); control group (70, 30 to 187) Health status: severely neutropenic (neutrophilic granulocytes < 0.5 x 109/L) Number (randomised/analysed): treatment group (52/32); control group (54/34) Median age, range (years): treatment group (52.5, 18 to 74); control group (48, 20 to 68) Sex (M/F): treatment group (6/26); control group (10/24) Exclusion criteria: not reported
Interventions	Treatment group Liposomal amphotericin B (AmBisome) Dose: 5 mg/kg Duration: median duration 14.5 days Frequency: daily Administration: IV at full dose infused over 45 min Control group Conventional amphotericin B Dose: 1 mg/kg Duration: median duration 16.4 days Frequency: daily Administration: IV infusion over 6 hours Sodium supplementation prior to conventional amphotericin B infusion to prevent nephrotoxicity and medication to prevent acute reactions permitted
Outcomes	Adverse events Fever/chills Nausea Exanthema Chest pain/tachypnoea Nephrotoxicity (SCr)
Notes	40 patients were ineligible for definitive enrolment and measured for efficacy This study was supported by grants from NeXstar Pharmaceuticals Inc., San Dimas, California, USA, and the Erasmus University Medical Centre Rotterdam/Dijkzigt Hospital
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described
Allocation concealment (selection bias)	Unclear risk	Sealed envelopes
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding of personnel and patients
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing outcome data
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes were reported
Other bias	High risk	The study was supported by grants from NeXstar Pharmaceuticals Inc., San Dimas, California, USA

Prentice 1997.

Methods	Study design: parallel, open‐label RCT with three arms Study duration: not reported
Participants	Setting: inpatients Countries: 19 European centres Health status: patients with pyrexia of unknown origin or neutropenia (neutrophilic granulocytes < 0.5 x 10⁹/L) Number (adults/children): treatment group 1 (48/70); treatment group 2 (47/71); control group (39/63) Age median (years, adults/children): treatment group 1 (43/7); treatment group 2 (42/7); control group (45/7) Sex (M; adults/children): treatment group 1 (27/36); treatment group 2 (31/40); control group (26/39) Exclusion criteria: not reported
Interventions	Treatment group 1 Liposomal amphotericin B (amBisome) Dose: 1 mg/kg Duration: not reported Frequency: daily Administration: was started immediately at the randomised study dose, without prior test doses Treatment group 2 Liposomal amphotericin B (amBisome) Dose: 3 mg/kg Duration: not reported Frequency: daily Administration: was started immediately at the randomised study dose, without prior test doses Control group Conventional amphotericin B (Fungizone) Dose: 1 mg/kg Duration: not reported Frequency: daily Administration: a gradual dose increase of c‐AMB was allowed over 2–3 d until a total planned daily dose of 1 mg/kg was reached There were no restrictions on use of premedication such as antihistamines, steroids, pethidine or paracetamol
Outcomes	Infusion‐related allergic reactions Dose reduction or study drug discontinuation Nephrotoxicity (SCr)
Notes	Antifungal prophylaxis with itraconazole, fluconazole and/or an oral polyene permitted until day of enrolment Dose reduction for nephrotoxicity was allowed to 50% of the prescribed daily dose. After kidney function recovery, the dose could be restored stepwise to the originally assigned level Educational grant and free AmBisome supplies, by Nexstar Inc
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Central randomisation; each participating centre provided with blinded, numbered envelopes which required sequential opening
Allocation concealment (selection bias)	Low risk	Sequentially numbered, sealed envelopes
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding of personnel and patients
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	Missing outcome data for efficacy were explained and ITT analysis performed
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes were reported
Other bias	High risk	These studies were made possible through an educational grant and free AmBisome supplies, by Nexstar Inc, USA

Shah 2005.

Methods	Study design: parallel, open‐label RCT Study duration: 14 June 2000 to 16 May 2003
Participants	Setting: inpatients Country: UK Health status: patients undergoing liver transplantation at high risk (defined as patients requiring ICU admission beyond post‐transplant day 4, transplanted for fulminant hepatic failure or patients re‐admitted to ICU within 3 months of liver transplant Number: treatment group (43); control group (44) Age mean (years): treatment group (50.6); control group (49) Exclusion criteria: not reported
Interventions	Treatment group Liposomal amphotericin B (amBisome) Dose: 50 mg Duration: until patient discharged from ICU, death or suspected or confirmed IFI Frequency: daily Administration: not reported Control group Conventional amphotericin B Dose: 15 mg Duration: until patient discharged from ICU, death or suspected or confirmed IFI Frequency: daily Administration: not reported
Outcomes	Nephrotoxicity (mean CrCl, mL/min)
Notes	Study of low dose for prevention of IFI Not included in quantitative synthesis (was the only study which assessed CrCl in mL/min) Funding source: Gilead, Cambridge, UK
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Information about the method of randomisation is described (random selection of sealed envelopes)
Allocation concealment (selection bias)	Unclear risk	Sealed envelopes
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding of personnel and patients
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing outcome data
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes were reported
Other bias	High risk	Funding by Gilead

Sundar 2004.

Methods	Study design: parallel, open‐label RCT with three arms Study duration: May to July 2001
Participants	Setting: Kala‐Azar Medical Research Center Country: India Health status: Symptoms and signs of visceral leishmaniasis (fever, weight loss, and splenomegaly); microscopic analysis of a splenic aspirate smear revealed parasites Number: treatment group 1 (51); treatment group 2 (51); control group (51) Mean age ± SE (years): treatment group 1 (17 ± 2); treatment group 2 (19 ± 2); control group (20 ± 2) Sex (M): treatment group 1 (63%); treatment group 2 (73%); control group (73%) Baseline SCr (mg/dL): treatment group 1 (0.8); treatment group 2 (0.8); control group (0.9) Exclusion criteria: pregnant or breastfeeding; HIV seropositive; serious concurrent infection; granulocyte count < 1000 granulocytes/mm3; Hb < 3.5 g/dL; platelet count < 40,000 platelets/mm3; hepatic transaminase or total bilirubin levels 13 x ULN; SCr > 2.0 mg/dL; prothrombin time > 5 seconds above control time
Interventions	Treatment group 1 Liposomal amphotericin B (amBisome) Dose: 2 mg/kg Duration: 5 infusions on consecutive days Administration: 4 h IV infusion Treatment group 2 Amphotericin B lipid complex (Abelcet) Dose: 2 mg/kg Duration: 5 infusions on consecutive days Administration: 4 h IV infusion Control group Conventional amphotericin B (Fungizone) Dose: 1 mg/kg Duration: 15 infusions on alternate days Administration: 4 h IV infusion Patients in control group received antipyretic (paracetamol 500 mg) and antihistamine treatment (chlorpheniramine 50 mg) before each infusion
Outcomes	Infusion‐associated reactions Fever and/or rigours (more than 1 episode) Nephrotoxicity (SCr level)
Notes	No losses to follow‐up Because 5 day liposomal amphotericin B or amphotericin B lipid complex regimens induced little toxicity laboratory testing was not repeated for treatment group 1 and treatment group 2 patients until the apparent cure evaluation 2 weeks after initiation of treatment (day 19) Financial support: The Liposome Company; World Health Organization/World Bank/United Nations Development Programme Special Programme in Research and Training in Tropical Diseases (grant 990106); and Sitaram Memorial Trust, Muzaffarpur, India
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random sequence generation by computer‐based random number generator
Allocation concealment (selection bias)	Unclear risk	Sealed envelopes.
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding of personnel and patients
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing outcome data
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes were reported
Other bias	High risk	Paired‐sample t‐test was used to compare pre and post treatment values in each group, the test for normal distribution is not shown

Sundar 2010.

Methods	Study design: parallel, open‐label RCT Study duration: February 2008 to March 2009
Participants	Setting: field centre of Kala‐Azar Medical Research Center Country: India Patients between 2 and 65 years old with symptoms and signs of leishmaniasis; parasites were shown on microscopy of a splenic aspirate smear Baseline SCr (mg/dL): treatment group (0.7 ± 0.3); control group (0.7 ± 0.2) Number: treatment group (304); control group (108) Mean age, range (years): treatment group (19, 2 to 60; ≤ 12 years: 54%); control group (18, 2 to 65; ≤ 12 years: 66%) Sex (M/F): treatment group (182/122); control group (67/41) Exclusion criteria: HIV seropositive; serious concurrent infection; WCC < 750/mm3; Hb < 3.5 g/dL; platelet count < 40,000/mm3; levels of hepatic aspartate aminotransferase or total bilirubin > 5 times the normal range; SCr > 1.5 x ULN; prothrombin time > 4 seconds above control
Interventions	Treatment group Liposomal amphotericin B Dose: 10 mg/kg Duration: unique dose Frequency: unique dose Administration: IV over 1 hour Patients were discharged home 24 hours later Control group Conventional amphotericin B (Fungizone) Dose: 1 mg/kg Duration: 15 infusions Frequency: every other day during a 29‐day hospitalisation Administration: IV over 6 hours Patients in the conventional therapy group who had IRR required subsequent oral pretreatment with an antipyretic (paracetamol) and an antihistamine (chlorpheniramine) with the use of standard paediatric or adult doses
Outcomes	Infusion‐related reactions Fever and/or rigours Nephrotoxicity (SCr)
Notes	No patient was lost to follow‐up Supported by the Sitaram Memorial Trust and the Varanasi Physicians Research and Education Foundation
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐based random number generator
Allocation concealment (selection bias)	Low risk	Sealed envelopes using a computer‐based random number generator
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding of personnel and patients
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	No losses to follow‐up
Selective reporting (reporting bias)	Low risk	Protocol available and all prespecified outcomes were reported
Other bias	High risk	An independent sample t‐test was used to detect differences in clinical and laboratory results (included SCr) for the two study groups, test for normal distribution was not shown Supported by the Sitaram Memorial Trust and the Varanasi Physicians Research and Education Foundation

Thakur 2001.

Methods	Study design: parallel RCT Study duration: not reported
Participants	Setting: not reported Country: India Cases of febrile splenomegaly with presence of Leishman – Donovan bodies in their splenic aspirates; 12 and 60 years Number: treatment group (17); control group (17) Mean age (years): treatment group (16.4); control group (18.2) Sex (M/F): treatment group (14/3); control group (14/3) Exclusion criteria: HIV positive cases; cases with tuberculosis; kidney, liver and heart diseases; pregnant and lactating women; Refusal to give informed consent; patients previously treated with amBisome were excluded
Interventions	Treatment group Liposomal amphotericin B (AmBisome) Dose: 15 mg/kg Duration: single dose Frequency: single dose Administration: IV over 2 hour Control group Conventional amphotericin B Dose: 1 mg/kg Duration: 20 days Frequency: daily Administration: IV over 2 hours
Outcomes	Infusion‐related reactions Shivering during infusion Shaking chill during infusion Nephrotoxicity (rise in SCr)
Notes	No patient was lost to follow‐up Not included in quantitative synthesis because only reported rise in SCr (is not possible to know the magnitude of the change)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No information on method used is described
Allocation concealment (selection bias)	Unclear risk	not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	not reported
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	No patient was lost to follow‐up
Selective reporting (reporting bias)	Low risk	All outcomes pre‐specified reported
Other bias	Low risk	The authors did not identified another possible source of bias Funding source: Balaji Utthan Sansthan

Walsh 1999.

Methods	Study design: parallel RCT Study duration: January 1995 to May 1996
Participants	Setting: multicentre (32) Country: USA Patients in chemotherapy for leukaemia, lymphoma, or other cancers or had undergone bone marrow or peripheral‐blood stem‐cell transplantation; and had received empirical antibacterial therapy for at least 5 days while continuing to have fever and neutropenia (absolute neutrophil count < 500 mm3) Number: treatment group (343); control group (344) Mean age, range (years): treatment group (41, 2 to 79); control group (42, 2 to 80) Sex (M/F): treatment group (184/159); control group (190/153) Exclusion criteria: known uncontrolled bacteraemia or invasive fungal infection at the time of randomisation; received any form of parenteral amphotericin B within 10 days before administration of the study drug; serum levels of aspartate aminotransferase, alanine aminotransferase or alkaline phosphatase > 10 x ULN; total serum bilirubin levels > 3 mg/dL; aminotransferase levels ≥ 2 x ULN or > 5 mg/dL; SCr > 2 x ULN; history of anaphylactic reaction to conventional amphotericin B
Interventions	Treatment group Liposomal amphotericin B (amBisome) Dose: 3 mg/kg Duration: mean duration of therapy was 10.8 days Frequency: daily Administration: standard care of the institution Control group Conventional amphotericin B Dose: 0.6 mg/kg Duration: mean duration of therapy was 10.3 days Frequency: daily Administration: standard care of the institution Before the first infusion, no premedication for the prevention of infusion‐related reactions were permitted. If a patient had an infusion‐related toxic reaction during the first infusion, it was treated. For subsequent infusions, appropriate premedication were administered at the discretion of the blinded investigator or primary physician
Outcomes	Infusion‐related reactions Fever following infusion (increase ≥ 1 °C) Chills or rigours Nausea Vomiting Headache Dyspnoea Hypotension Tachycardia Nephrotoxicity (SCr)
Notes	Mean daily doses throughout the study were 3.0 ± 0.9 mg/kg for liposomal amphotericin B and 0.6 ± 0.2 mg/kg for conventional amphotericin B Supported in part by a grant (N01‐Al‐65296) from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, and by Fujisawa USA, Deerfield, Ill., through grants provided to study centres
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A computerised system randomly assigned treatment according to centre and risk stratum
Allocation concealment (selection bias)	Unclear risk	The pharmacist at the patient’s institution telephoned a central randomisation centre to obtain a drug assignment. No information about the concealment of the allocation is stated
Blinding of participants and personnel (performance bias) All outcomes	Low risk	To maintain the blinded conditions, the study drug concentrations were adjusted so that the volume of solution in the IV bag was the same for both study drugs. Infusion bottles were concealed by opaque bags and infusion tubing was either opaque or covered with opaque wrapping
Blinding of outcome assessment (detection bias) All outcomes	High risk	not reported
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Insufficient information to permit judgement
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported
Other bias	High risk	Pharmaceutical funding

CrCl ‐ creatinine clearance; ICU ‐ intensive care unit; Hb ‐ haemoglobin; HIV ‐ human immunodeficiency virus; IFI ‐ invasive fungal infection; IV ‐ intravenous; M/F ‐ male/female; RCT ‐ randomised controlled trial; SCr ‐ serum creatinine; SGOT ‐ serum glutamic oxaloacetic transaminase; SGPT ‐ serum glutamic pyruvic transaminase; WCC ‐ white cell count; ULN ‐ upper limit of normal

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Anaissie 1995	Wrong intervention/comparison: amphotericin B lipid complex versus conventional amphotericin B
Cordonnier 2009	Wrong intervention/comparison: broad spectrum betalactam and antifungal prophylaxis as needed
Drew 2004	Wrong intervention/comparison: comparative safety of amphotericin B lipid complex and amphotericin B deoxycholate as aerosolized antifungal prophylaxis
Echevarria 2006	Wrong intervention/comparison: oral rehydration solution to prevent nephrotoxicity of amphotericin B
Eriksson 2001	Wrong intervention/comparison: comparison of effects of amphotericin B deoxycholate infused over 4 or 24 hours
Fleming 2001	Wrong intervention/comparison: liposomal amphotericin B versus amphotericin B lipid complex
Heinemann 1994	Study of pharmacokinetic characteristics
Isnard 2007	Wrong intervention/comparison: switch from conventional to liposomal amphotericin B formulation
Llanos 1991	Wrong intervention/comparison: effect of salt supplementation on amphotericin B nephrotoxicity
Luke 1998	Wrong intervention/comparison: amphotericin B lipid complex
Moreau 1992	Wrong intervention/comparison: amphotericin B mixed with intralipid compared with conventional amphotericin B
NCT00002019	Wrong intervention/comparison: safety and efficacy of amphotericin B lipid complex
NCT00523965	Wrong intervention/comparison: different combination treatment regimens of liposomal amphotericin B, paromomycin and miltefosine
NCT01122771	Wrong intervention/comparison: different combination treatment regimens of liposomal amphotericin B, paromomycin and miltefosine
NCT01310738	Wrong intervention/comparison: active comparator meglumine antimoniate
Sharkey 1996	Wrong intervention/comparison: amphotericin B lipid complex compared with amphotericin B
Sorkine 1996	Wrong intervention/comparison: administration of amphotericin B in lipid emulsion
Subira 2004	Wrong intervention/comparison: amphotericin B lipid complex versus conventional amphotericin B
Sundar 2014	Wrong intervention/comparison: amphotericin B emulsion versus liposomal formulation
Tollemar 1995	Placebo‐controlled trial
Wingard 2000	Wrong intervention/comparison: liposomal amphotericin B versus amphotericin B lipid complex

Differences between protocol and review

This review differs from the published protocol in the following points:

• Measurement of the secondary outcome was only possible to infusion‐related reactions; the others were not possible due to lack of data

• A sensitivity analysis was performed investigating overall risk of bias

• A sensitivity analysis was carried out excluding studies with dose or duration of liposomal amphotericin B off‐label.

Contributions of authors

1. Draft the protocol: JPBA

2. Study selection: JPBA, AMRH

3. Extract data from studies: JPBA, AMRH

4. Enter data into RevMan: JPBA

5. Carry out the analysis: JPBA

6. Interpret the analysis: JPBA, AMRH

7. Draft the final review: JPBA, AMRH

8. Disagreement resolution: JPBA, AMRH

9. Update the review: JPBA

Sources of support

Internal sources

• Hospital Pablo Tobón Uribe, Colombia.

External sources

• None, Other.

Declarations of interest

Hospital Pablo Tobón Uribe provides monetary incentives to published review authors to promote research studies.

New