Amphotericin B deoxycholate: time to retire our old standard?

Amphotericin B deoxycholate (AmBd) has enjoyed a long lifespan in our antifungal armamentarium. A broad spectrum of activity, relatively low levels of acquired resistance in fungi, low acquisition cost and most importantly, a lack of suitable alternatives, has kept AmBd as a drug of choice for the treatment of systemic fungal infections for over 40 years [1]. Unfortunately, AmBd has many well-known and severe toxicities that can only be partially attenuated through saline loading and pre-medicating, including dose-limiting nephrotoxicity and considerable infusion-related reactions. Despite this significant toxicity profile, more clinical experience has been gained with AmBd than with virtually any other systemic anti-fungal. Recently, several new broad-spectrum antifungal agents have challenged the position of AmBd and other amphotericin B products as the drugs of choice for many fungal infections.

Alternatives to AmBd for systemic anti-fungal therapy have existed for many years, but are laden with significant drawbacks. Early azoles such as clotrimazole and miconazole were introduced in the 1960s, but their systemic use was limited by adverse effects and problematic pharmacokinetics [2]. Ketoconazole was introduced in 1981 and became the drug of choice in nonsevere endemic mycoses. It was the first orally available systemic antifungal agent. However, the oral formulation had problematic inter-patient differences in absorption, and no intravenous formulation was available. Keto-conazole was also shown to be less effective in immunocompromised patients, and lacks activity against many molds. AmBd continued to be used in many clinical situations.

The approval of fluconazole in 1990 represented a true advance in antifungal therapy. Fluconazole was the first of the triazoles to reach the market [2]. With its excellent oral absorption, intravenous formulation, low adverse effects and good activity against many yeasts, fluconazole has supplanted AmBd as a drug of choice for the treatment of many fungal infections. Although fluconazole is a very good anti-yeast drug, it lacks useful activity against most molds, and its popularity in prophylactic regimens in some patient populations has caused a shift towards both intrinsically resistant Candida spp. and fluconazole resistance in Candida albicans in some centers [3].

Fluconazole’s introduction provided a welltolerated and efficacious oral antiyeast medication, although it did not add to our ability to treat filamentous fungi. To address this, itraconazole was introduced to the antifungal market in 1992. With much of the anti-yeast activity of fluconazole and activity against Aspergillus and other fungi, itraconazole had a lot of promise. Unfortunately, its use has been tempered by several of the same issues that held back the early azoles. The initially -introduced oral capsule formulation has highly variable, and often poor, bioavailability. An intravenous formulation was also not initially available. Although both of these issues have since been rectified, use of itraconazole has been limited. The more bioavailable liquid oral formulation of itraconazole is poorly tolerated, and itraconazole has more adverse effects and more severe drug interactions than fluconazole. It has also not been studied as a primary therapy in invasive aspergillosis and has been mostly relegated to a role in consolidation therapy for that invasive fungal infection. Although both fluconazole and itraconazole were shown to be equivalent to AmBd as empiric antifungal therapy in persistently febrile neutropenic patients, the broader antifungal spectrum of AmBd and its extensive history in this patient population led to its continued predominant use for this indication.

The need for an alternative well tolerated, broad-spectrum antifungal with activity against yeasts and molds continued. Increasing numbers of immuno-compromised patients, fueled by increasingly intense oncology regimens, more organ transplant recipients, and the AIDS epidemic, engendered the search for a safer alternative to AmBd that maintained its broad spectrum and good activity. Enter the lipid formulations of amphotericin B (LFABs). In the mid- to-late 1990s, three LFABs were introduced: amphotericin B lipid complex (ABLC), amphotericin B colloidal dispersion (ABCD) and liposomal amphotericin B (L-AmB).

These compounds were designed to attenuate some of the traditional toxicities of AmBd while maintaining its efficacy. They have been successful to varying degrees. All have been shown to be less nephrotoxic than AmBd (L-AmB the least of the three), and some have been shown to also attenuate infusion-related reactions [4]. This reduced toxicity has allowed increased amounts of the drug to be delivered, up to 10 mg/kg/day for L-AmB [5]; a consideration that is especially important for fungal infections that are relatively resistant to amphotericin B, such as zygomycosis. What is the drawback to the LFABs? Cost. Pharmacy costs of LFABs are many times higher than for AmBd [4], and with no obvious efficacy differences, many institutions have been slow to embrace their use. It is the cost difference between AmBd and the LFABs that has led many institutions to restrict the use of LFABs to patients who have developed, or are likely to develop, intolerable adverse effects to AmBd. Recent data shows that a trend towards increased efficacy with LFABs in many invasive fungal infections appears to have emerged [6]. However, this trend is subtle and many institutions continue to restrict LFABs to AmBd-intolerant patients or patients with pre-existing renal dysfunction. This strategy of waiting for toxicity to develop in patients with seemingly normal renal function is only cost justifiable from the perspective of the drug budget, since some pharmaco-economic analyses have shown that the overall healthcare costs for LFABs and AmBd are not as disparate when the cost of nephrotoxicity and other adverse effects are considered [7].

The newest agents to enter the antifungal market are the echino-candins and voriconazole. Voriconazole is an extended-spectrum triazole with activity against most clinical Candida spp. and many molds. It is indicated in the USA for primary therapy of invasive aspergillosis, esophageal candidiasis, candidemia in non-neutropenic patients and serious infections caused by Scedo-sporium apiospermum and Fusarium spp. Voriconazole was proven to be more effective than AmBd in the treatment of invasive aspergillosis by both increased cure rates at 12 weeks and decreased mortality [8]. Significantly, it is only one of two drugs approved for primary therapy of invasive aspergillosis in the USA, the other being AmBd. Voriconazole was only recently approved in the USA for the treatment of candidemia, but its broad spectrum against Candida including C. glabrata and C. kruseii, its activity against some fluconazole-resistant C. albicans, and its oral formulation makes it a good alternative in these infections. The oral formulation of voriconazole is a highly bioavailable tablet, giving it an advantage over itraconazole and the echinocandins, since preferential use of highly bioavailable oral antifungal therapy has been shown to decrease institutional costs [9]. Voriconazole has been studied as empiric antifungal therapy in persistently febrile neutropenic patients in a controversial trial [10]. In this study, voriconazole did not achieve noninferiority by the investigators’ predefined composite end point, but did show significantly less breakthrough fungal infections than L-AmB. These seemingly contradictory findings have caused many to question the design and primary end points of this type of study, and many clinicians are using voriconazole for this indication.

Voriconazole does have a down side, namely a large number of significant drug interactions, nonlinear pharmacokinetics and an adverse-effect profile that includes visual effects in a high percentage of patients and liver function abnormalities in others. It also has an antifungal spectrum that, while very broad, does not include the zygomycetes, a group of emerging fungi that are somewhat covered by amphotericin B formulations. However, its high tolerability, highly bioavailable oral formulation and strong clinical data for several indications have led to it displacing amphotericin B formulations for many uses.

Two echinocandins have been approved for clinical use – caspofungin and micafungin. Caspofungin has been available in the USA for several years; micafungin was only recently approved by the US Food and Drug Administration (FDA), but has been available in Japan for some time. These drugs have excellent activity against most clinical Candida spp., as well as activity against Aspergillus. They have a low incidence of adverse reactions, and have been studied in refractory therapy for invasive aspergillosis, candidiasis, antifungal prophylaxis in cancer patients and as empiric antifungal therapy in persistent febrile neutropenia. Caspofungin has been compared with AmBd in a large, randomized, double-blinded trial for the treatment of invasive candidiasis and demonstrated at least equivalent efficacy with less toxicity [11]. Noninferiority was demonstrated with less adverse effects when caspofungin was compared with L-AmB for empiric therapy in persistently febrile neutropenic patients in another large double-blinded trial [12]. Less robust data are available for the treatment of invasive aspergillosis, but the novel mechanism of action of echinocandins combined with their low toxicity has renewed interest in combination antifungal therapy for this disease.

The echinocandins are unfortunately limited to intravenous formulations. They also have activity against Aspergillus that is neither classically fungicidal, nor fungistatic. Instead, they inhibit the rapidly dividing tips of growing hyphae, leading to the production of aberrant hyphal forms [13]. This mechanism has led to concern that they may not be ideal choices of monotherapy for patients with invasive aspergillosis that has been present for some time. This concern has not been proven, but they have taken a back seat to voriconazole for the treatment of this disease due to voriconazole’s wealth of clinical trial data.

So where does this leave AmBd? It is still an inexpensive option that is effective for many mycoses. To try to improve AmBd’s margin of safety, some investigators have studied continuous infusions of AmBd [14,15]. One group studied AmBd given over 4 h to continuous infusion AmBd in 80 patients with cancer [14]. The patients were randomized, and each group received approximately 1 mg/kg/day of AmBd in an unblinded fashion. Most patients in the study had acute leukemia, although the indications for AmBd therapy were not listed. The investigators found that the continuous infusion AmBd patients experienced significantly less nephrotoxicity and infusion-related toxicity than the 4 h infusion patients. The mortality rate in the 4 h infusion arm was also significantly higher, with seven patients dying during treatment (three from proven fungal infections) compared with none in the continuous infusion arm. Although this study introduces an interesting premise, it is likely that many of these patients did not have fungal infections. Also, the relatively small sample size makes interpretation questionable. A separate noncomparative study demonstrated that the reduced toxicity of continuous infusions of AmBd allows for increased dosing. This paper demonstrated that up to 2.0 mg/kg/day of AmBd was well tolerated. Both of these studies prove that continuous-infusion AmBd is less toxic than intermittent infusions, but they are unable to prove greater or even equal efficacy than the 2–6 h intermittent infusions that most patients receive. Amphotericin B is a concentration-dependent drug, suggesting that better killing may occur with high intermittent dosing. Until studies in patients with proven fungal infections are conducted, continuous infusion AmBd remains a potentially efficacious therapy, not a proven one.

AmBd has survived years of challenges from many new and different antifungals, but has remained our de facto default therapy for decades. However, that has finally changed. The introduction of the echinocandins has given us a class of agents with broad anticandidal activity and few adverse reactions. Voriconazole has proven to be more efficacious than AmBd in the treatment of invasive aspergillosis, with less serious adverse events and a convenient oral formulation. For the indications where AmBd remains a drug of choice, the LFABs are available for the patients who do not or cannot tolerate AmBd. Continuous infusions of AmBd have been shown to considerably reduce toxicity, but the efficacy of this method remains unproven in invasive fungal infections. Although there are indications for AmBd where alternatives have not been well studied, these are few in number and represent the minority of patients who require systemic antifungal pharmacotherapy.

After a long period of few effective antifungal options, we now find ourselves with many. We need to decide if the cost savings to our drug budget is worth exposing our patients to the significant and likely toxicity of AmBd therapy.