The challenge of systemic Candida infections

In hospitalised patients, invasive fungal diseases, such as candidaemia, are an important complication in clinical practice.^1,2 Candida infections have contributed significantly to mortality in immunosuppressed patients and treatment outcomes for invasive candidosis are still less than optimal. Candidaemia is the fourth leading cause of bloodstream infections in paediatric patients, cancer patients and hospitalised patients, with mortality rates ranging from 20% in paediatric patients to 47% in intensive care unit (ICU) patients. In a large US cohort study, patients had a number of underlying conditions or risk factors including: urinary catheterisation (60%); endotracheal intubation (50%); surgery (36%); diabetes (26%); neoplasia (26%); and haemodialysis (12%).³ The specific epidemiology with regard to incidence of candidaemia as well as distribution of Candida pathogens

(C. albicans versus non-C. albicans species) varies markedly from country to country as well as from hospital to hospital. C. albicans is still the most common isolate; however, there has been a shift towards more azole-resistant species (for example, C. glabrata) found in some institutions, paralleled by a reduced susceptibility to azole antifungals such as fluconazole. The reasons for this variation are not well understood and may reflect, in part, selection mechanism (for example, due to azole prophylaxis). In some patient groups, such as in haematological patients with prolonged neutropenia, non-C. albicans species are detected much more frequently compared with non-neutropenic patients in the surgical ICUs. Candida species are the primary cause of invasive fungal infections but an increasing number of infections are being attributed to Aspergillus species. In immunocompromised patients, Aspergillus species are an important cause of life-threatening infections. Invasive aspergillosis is a frequent cause of mortality not only in patients with haematological malignancies and recipients of allogeneic haematopoietic stem cell transplants but is also increasingly observed in ICU patients.⁴ Mould infections will not be discussed in this review.

Every blood culture that reveals growth of yeasts, together with clinical signs of infection, represents an infection that needs prompt management, including the initiation of antifungal therapy and the removal of central venous lines. When cultures of catheter tips grow yeasts, while blood cultures remain sterile, systemic antifungals are not indicated in every case, depending on the clinical condition of the patient. Important considerations when choosing the antifungal agent and the mode of application (IV versus oral) in candidaemia and systemic candidiasis include the localisation of the infection, the severity of disease (for example, sepsis, severe sepsis, and septic shock), impairment of organ functions (especially in liver and kidney), previous exposure to antifungals, the identified fungal strain, local resistance patterns and patient characteristics, such as age.

Most importantly, antifungal treatment for uncomplicated candidaemia is recommended for 14 days after the first negative blood culture and resolution of all clinical signs of infection. Ophthalmoscopy is recommended prior to the discontinuation of antifungal chemotherapy to rule out endophthalmitis or chorioretinitis. De-escalation strategies (switch from IV echinocandin therapy after initial response to oral therapy with an azole antifungal drug) are commonly used in clinical practice but criteria for early switch (for example, after three-to-five days) are not clearly established and need further clinical evaluation. After improvement of clinical signs, sterilisation of blood cultures and documented in vitro susceptibility of the causative yeast, step-down therapy after initial treatment with an echinocandin (anidulafungin, caspofungin and micafungin) was shown to be effective with oral fluconazole (or voriconazole) starting on day 10 of antifungal therapy, and may be recommended if oral drug intake and gastrointestinal absorption are possible.¹

Conventional amphotericin B-deoxycholate (cAmB-D) has served as standard treatment for more than five decades, but its toxicity and limited efficacy have underscored the need for alternative antifungal drugs. Fluconazole is as effective as amphotericin B for the treatment of candidaemia and has superior safety, although the reduced susceptibility of some species, such as  C. glabrata and C. krusei, may limit its use in settings where these species are prevalent. Newer options for patients with candidaemia or invasive candidiasis include broad-spectrum azoles (such as voriconazole), and the echinocandins (caspofungin, micafungin and anidulafungin), all of which have excellent activity in vitro against a range of common Candida species. Most importantly, the immediate start of antifungal therapy in a clinically compromised patient with candidaemia after receipt of the culture results from the laboratory is mandatory in order to improve the overall survival of the patient. In daily practice, the time for blood cultures to become positive takes up to 96 hours. This may lead to a substantial delay in the initiation of antifungal therapy in many patients. In addition, it is assumed that only 50–60% of blood cultures may detect disseminated Candida infection. According to this diagnostic dilemma, empirical or pre-emptive antifungal therapy is increasingly used, in particular in severely ill patients with sepsis in the ICU. Despite the lack of prospective, randomised trials, some guidelines have published recommendations for empirical treatment of suspected Candida infections.¹

The 2009 clinical practice guidelines of the Infectious Diseases Society of America (IDSA) reflect significant changes in the management of candidaemia and invasive candidosis with respect to the appropriate use of echinocandins and expanded spectrum azoles.¹ The expert panel of the IDSA favours an echinocandin for first-line treatment of candidaemia in non-granulocytopenic patients with moderately severe-to-severe illness, as first-line treatment for granulocytopenic patients, patients recently exposed to azole, candidaemia caused by C. glabrata, and/or suspected invasive candidosis (empirical therapy) caused by C. glabrata or C. krusei (Figure 1). The three echinocandins are regarded as equally effective for treatment of candidaemia and invasive candidosis. Fluconazole may be considered in less‑critically ill patients. According to the IDSA, amphotericin B is recommended as an alternative therapy in patients with limited tolerance to echinocandins or fluconazole, or in situations where availability of other antifungals is limited.¹ Amphotericin B-induced nephrotoxicity is associated with increased mortality in at-risk individuals, and patients with more than two risk factors for nephrotoxicity are potential candidates for alternative antifungal therapy. Consequently, according to the European guideline, amphotericin B deoxycholate is not recommended for any indication due to severe side effects.⁵ Voriconazole or fluconazole are recommended as step-down therapy in selected cases (either oral or IV).

Echinocandins are cyclic lipopeptide molecules that inhibit the synthetic cell wall enzyme complex, β-1,3-D-glucan synthase. Inability of the organism to build β-1,3-D-glucan destabilises the integrity of the fungal cell wall, leading to osmotic instability and cell death. Echinocandins are known to be active against Candida and Aspergillus species but are also active against Pneumocystis jirovecii. Echinocandins are fungicidal against yeast and fungistatic against mould and have a similar spectrum of antifungal activity. They are not active at clinically relevant concentrations against Zygomycetes, Cryptococcus neoformans, or Fusarium species. In addition, echinocandin activity can persist in a drug-free environment following drug exposure (post-antifungal effect). Pharmacological properties are shown in Table 1.

However, the echinocandins have structural and pharmacokinetic differences.⁶ Differences between the three echinocandins with regard to the route of metabolism, requirement for a loading dose, dose adjustment in patients with moderate-to-severe hepatic disease and different dosing schedules for different types of Candida infections have to be considered (Figure 2). Drug–drug interactions are rare. None are major p-glycoprotein substrates and none are metabolised significantly by cytochrome P450.⁶ All echinocandins have good tissue penetration, but none penetrate sufficiently into the central nervous system or the eye, therefore they are not recommended as the drug of choice for treatment of patients with fungal meningitis or endophthalmitis.

Unlike fluconazole, the echinocandins have also shown activity against biofilms. Importantly, with all echinocandins, no dose modification is required in patients with renal insufficiency or moderate hepatic impairment. However, the area under the curve may be increased by up to 75% for caspofungin and a dose reduction to 35mg/day may be required. These agents have an excellent safety profile that is similar to that of fluconazole and a low potential for drug interactions.

Data from prospective randomised trials comparing either: (i) micafungin (100mg/day) with L-AmB (3mg/kg/day); (ii) micafungin (at 100mg/day and 150mg/day, respectively) with caspofungin (70mg/day on day 1 followed by 50mg/day from day 2); and (iii) anidulafungin (200mg IV loading dose followed by 100mg IV daily) compared with fluconazole (800mg IV loading dose followed by 400mg IV daily) for invasive candidosis and candidaemia have been published. According to the results from these trials, echinocandins are strongly recommended for targeted initial treatment of candidaemia, whereas liposomal amphotericin B and voriconazole are supported with moderate strength. A direct comparison between fluconazole and anidulafungin showed a similar safety profile, a superior treatment response and a trend toward better survival in patients treated with anidulafungin. This was evaluated in a randomised, double-blind, multicentre, multinational, phase III study of patients with candidaemia and/or other forms of invasive Candida infections.⁷ Global success at the end of IV therapy for anidulafungin was 75.6% compared with 60.2% for fluconazole-treated patients. The results of this study led to the decision of the ESCMID/EFISG to give fluconazole a marginal recommendation only⁵ (see Figure 3).

An unmet medical need, with respect to candidaemia and invasive Candida infections, is the development of treatment strategies with echinocandins in specific ICU patient populations, in order to improve their outcome and survival. Most recently, anidulafungin was studied in a non-comparative trial in high-risk patients treated in the ICU.⁸ The patients required at least one of each cofactor/underlying illness to be included in this trial (post-abdominal surgery, elderly individuals >65 years, renal insufficiency/failure or dialysis, solid tumour, solid organ (liver, kidney, lung, heart, pancreas) transplant recipients, hepatic insufficiency, or neutropenia (neutrophil count <500/mm3) including haematology/oncology patients). Anidulafungin was equally effective in all subgroups, but to a lesser extent in transplant and neutropenic patients.

A direct comparison of caspofungin and micafungin showed similar efficacy and safety. In addition, no difference in safety or efficacy was seen in patients treated with two different dosages of micafungin (100mg/day or 150mg/day).⁹

Further studies comparing different echinocandins are lacking. Higher dosages of caspofungin (150mg/day versus 70/50mg/day) and micafungin (150mg/day versus 100mg/day) showed a trend toward improved efficacy in subgroups of patients (APACHE-II score >20, granulocytopenia) and might be used in selected patients.^9,10 Owing to increased minimal inhibitory concentrations and a higher rate of persistent fungaemia, the use of echinocandins in fungal infections from C. parapsilosis may not be recommended. In this clinical situation, fluconazole is the preferred antifungal drug. However, it has to be stressed that only the combined strategy using an echinocandin together with the removal of an infected catheter may improve Candida-related mortality in severely ill patients with candidaemia/invasive candidosis.¹¹

In summary, it may be concluded that options for initial therapy of candidemia and other invasive Candida infections in non-granulocytopenic patients include one of the three approved echinocandin compounds; liposomal amphotericin B and voriconazole are secondary alternatives because of their less favourable pharmacological properties. European and US guidelines differ with regard to the use of fluconazole. In the ESCMID Candida guideline, fluconazole is regarded as drug of second choice only, but in the IDSA guideline as a potential alternative. In granulocytopenic patients, an echinocandin or liposomal amphotericin B is recommended as initial therapy based on the fungicidal mode of action. Indwelling central venous catheters serve as a main source of infection independent of the pathogenesis of candidemia in the individual patients and should be removed whenever feasible.