FormalPara Key Points

EMA-approved antifungals did not or only scarcely include pharmacokinetic/pharmacodynamic indices in their EPAR.

We described the pathway to authorisation for each antifungal authorised since 2000.

We provide an overview of applicable guidelines for antifungal development within the last 23 years.

1 Introduction

Although less common than superficial infections, invasive fungal disease is a big concern since it is linked to very high mortality rates [1]. Twenty-three years ago there were only three classes of antifungal agents available, mainly amphotericin B, flucytosine and the azoles (ketoconazole, itraconazole and fluconazole) [2]. Fungal infections were deemed easily treatable and negligible compared with bacterial infections until the 1970s [3]. This mindset changed with the rise of fungal nosocomial infections in immunocompromised patients [4]. Additionally, the rising resistance of pathogenic fungi against azoles provoked an urgency for newer drugs [5]. While the lead compound of anidulafungin was uncovered in 1974, it took another 27 year until the first echinocandin was authorised for Aspergillus infections [2].

Even with the introduction of echinocandins, there have been few additions to the market; since 2010 we have only seen the introduction of isavuconazole and a few generics [6]. Having a limited arsenal of medication is a serious problem as fungal resistance seems to be on the rise [7].

Antifungal drug development is challenging due to a relatively low incidence of invasive fungal infections, occurrence in immunocompromised patients and/or patients with serious illnesses and diagnostic difficulties [8]. Pharmacokinetic/pharmacodynamic (PK/PD) indices are already an established tool in antibacterial drug development and might aid development in the antifungal sector as well. The indices help create clinical trials that are time and cost efficient and can help to extrapolate efficiency data for rarer pathogens. Therefore, we set out to depict the development pathway of centrally approved antifungal agents and highlight the evolving role of PK/PD in this process.

1.1 Authorisation Process

The European Medicines Agency (EMA) is a scientific body in the European Union. Contrary to popular belief, the EMA does not have the authority to permit marketing but instead makes recommendations to the European Commission. Once granted, the authorisation is valid in all EU member states. While non-prescription and herbal drugs mostly go through what is called a national authorisation process, new antifungal substances must undergo a centralised process together with substances that treat HIV, cancer, diabetes mellitus, neurodegenerative diseases, autoimmune and other immune dysfunctions, viral diseases, etc. [9].

Applicants should provide data as recommended in the guidelines of the EMA and the Committee for Medicinal Products for Human Use (CHMP). They also need to adhere to the international requirements of Good Clinical Practice (GCP), Good Laboratory Practice and Good Manufacturing Practice (GMP). Drug developers may request advice and direction from the EMA on data and study requirements in the form of scientific advice procedures, where different working parties and committees of the agency will provide guidance. Once the application has been submitted to the EMA, the Committee for Medicinal Products for Human Use (CHMP) comprising a panel of experts evaluates the application alongside other EMA committees. After approval, the EMA publishes the European Public Assessment Report (EPAR), a comprehensive set of documents containing the data provided by the applicant and assessed by the CHMP. If the EMA identifies missing data or recognises safety issues, a risk management plan will be added with information on what needs to be evaluated further post-authorisation [9].

1.2 Pharmacokinetic/Pharmacodynamic (PK/PD) Indices

PK/PD indices are predictors of clinical efficacy stemming from a substance’s pharmacokinetic and pharmacodynamic data. For these indices, the MIC (minimum inhibitory concentration), that is, the concentration that inhibits antifungal growth under standardised conditions in vitro, is set into context to the unbound concentration–time profile of the respective drug in vivo. PK/PD indices give valuable information about dosing regimens in clinical trials and later during clinical practice. Three PK/PD categories can be described.

  1. (a)

    In case time (T) over MIC (T>MIC) correlates best with efficacy, the drug should be dosed to stay over the MIC for a long time within the dosing interval.

  2. (b)

    Another predictor can be the AUC (area under the curve) over 24 hours over the MIC (AUC0–24/MIC).

  3. (c)

    In case the concentration peak over MIC (Cmax/MIC) correlates best with efficacy, the drug should be given using high doses administered as short infusion or bolus.

The pharmacokinetic parameters mostly stem from animal models in earlier preclinical development. PK/PD indices may change when transferred from in vitro or animal species to humans and within different human indications. Furthermore, PK/PD indices are not absolute, may overlap and often do not differ significantly in their predictive power. They also relate to the mode/mechanism of action of a class of compounds and may depend on the pathogen treated. Of course, not only efficacy but also safety and tolerability aspects have to be considered when the clinical dose or dosing regimen is defined.

1.3 Aims

This study aimed to summarise the information on PK/PD indices in the EPARs of antifungals from the last 23 years; to depict the method of authorisation for each antifungal; to investigate the potential variability in content between the EPARs; and to extrapolate the data to potential approval pathways for future antifungals.

2 Methods

2.1 Inclusion and Exclusion Criteria

Figure 1 shows the selection process for this review.

Fig. 1
figure 1

The selection process for this review. Of 10 initially included substances, all generics were excluded. Ketoconazole HRA® was excluded for its primary use in Cushing syndrome. We included Cresemba, an antifungal that has not been assigned a pharmacotherapeutic group yet

Full size image

We conducted this review by extracting data from the first version of the EPARs at the time of first approval. We included antimycotic substances for systemic use that were authorised within the last 23 years.

We screened the EMA website by applying the following filters: ‘’Human’’ AND ‘’European public assessment reports (EPAR)’’ AND ‘’Authorised’’ AND ‘’antimycotics in systemic use’’.

We excluded generics because they require a different authorisation process and substances that were labelled as antimycotics but are used for other indications (Fig. 1).

We included substances that were not assigned a therapeutic group yet but are used for systemic fungal infections.

Since the EPAR does not include all research conducted on a drug but rather data presented by the applicant, we collected missing and background information by searching in PubMed and ClinicalTrials.gov whenever necessary.

2.2 Applied Software

All tables were created with Microsoft® Excel® and all graphs were created using Microsoft® PowerPoint®.

3 Results

3.1 Overview of Included Drugs

In total, we identified 10 drugs that were labelled under ‘antimycotics for systemic use’. We excluded Ketoconazole HRA® for its primary use in Cushing syndrome as well as the four generics: Posaconazole AHCL®, Posaconazole Accord®, Voriconazole Accord® and Voriconazole Hikma® (previously Voriconazole Hospira®). We included Cresemba® (isavuconazole) although it had not yet been assigned a pharmacotherapeutic group. An overview of the drugs is provided in Table 1.

Table 1 Names, marketing authorisation holder, approval date and substance class for all included substances
Full size table

3.2 Preclinical and Phase I Trials

3.2.1 PK/PD Indices

We compiled datasets from the preclinical phase for each authorised drug (Table 2) showing the main pharmacodynamic parameter, PK/PD targets for stasis as well as for 1-log10 and 2-log10 reduction. We distinguished between PK/PD indices published and PK/PD mentioned (indices or information about conducted PK/PD studies written down in the EPAR).

Table 2 Pharmacokinetic/pharmacodynamic (PK/PD) targets for net stasis, 1-log10, 2-log10 reduction CFU
Full size table

In the assessment of the six authorised medications since 2000 we found that one EPAR report did not mention any PK/PD indices at all (micafungin), one mentioned there was no correlation between efficacy and PK/PD indices (voriconazole) and two recommended further investigation post-authorisation (anidulafungin and caspofungin). Only two EPARs described the PK/PD indices investigated in detail (isavuconazole and posaconazole). For both, AUC/MIC was given main pharmacodynamic index, although no direct comparison to the other parameters (Cmax/MIC or T/MIC) was made.

In case the EPARs didn’t have any mention of PK/PD indices, we undertook a search of PubMed to see if there was data available and report data for the individual pathogen in Table 2.

3.2.2 Protein Binding

Protein binding is presented in Table 2. Micafungin plasma protein binding was assessed using the ultrafiltration method. The other EPARs only scarcely describe methods.

3.3 Authorisations

In this section, we look at the number of indications for each drug at the time of their initial authorisation. Voriconazole leads with regard to adult indications (five indications). Micafungin has the same number of authorisations in total (five initial indications) including two paediatric ones. Paediatric indications will not be included for the purpose of this review because their clinical development is different and analysis will often use extrapolated data from adult trials. The mean number of adult indications is 2.67 indications/substance at the time of authorisation.

Figure 2 depicts the indications associated with the initial authorisation of the antifungal agents. Furthermore, we present how many drugs were authorised as first-line or second-line treatment (meaning refractory and/or intolerant to the first-line medication). Six of 16 authorised indications are second-line treatments.

Fig. 2
figure 2

The number of substances authorised as first- and second-line therapy for each disease. (1) Candida ssp. infections include invasive candidiasis, invasive candidiasis in adult non-neutropenic patients, candidemia in non-neutropenic patients as well as fluconazole-resistant serious invasive Candida infections (including C. krusei). (2) Second-line therapy means being refractory and/or intolerant of the first-line medication. Refractoriness defines a progression or failure to improve after a minimum of 7 days after prior therapeutic doses of an effective antifungal

Full size image

More than a quarter of indications (five authorisations) are for infections with Candida ssp. (subspecies) and one quarter (four authorisations) for infections with Aspergillus ssp. Although PK/PD indices are directional for drug development, authorisation also depends on external factor such as marketing considerations.

3.4 Phase II and III Clinical Trials

Numbers as extracted from the EPARs can be found in Tables 3 and 4.

Table 3 Substances authorised for the main indications (aspergillosis, invasive candidiasis, oesophageal candidiasis, prophylaxis of fungal infections) including the no. of subjects in phase II + III if the EPAR reports the exact number
Full size table
Table 4 The number of subjects in phase II + III if the EPAR reports the exact number for rare indications
Full size table

While the absolute number of studies conducted and number of subjects included is hard to extract (with especially voriconazole offering mostly pooled data from across their programme), it can be said that oesophageal candidiasis (EC; for the purpose of this review, this also includes indications like oropharyngeal candidiasis and candida esophagitis) was the most investigated disease overall (2315 subjects), followed by aspergillosis (1688 subjects), invasive candidiasis (1323 subjects) and prophylaxis (905 subjects). For posaconazole, we were unable to find an indication for the phase II study mentioned in the EPAR, therefore the studied 98 subjects could not be included in the tables.

Oesophageal candidiasis was also the leading indication in phase II (686 subjects) and in phase III (1578 subjects). Voriconazole offered pooled data on 51 more cases with oropharyngeal candidiasis where we could not determine the origin. All substances showed a diverse number of subjects in their phase II and III trials.

Oesophageal candidiasis is one of the most studied diseases in phase II trials, however it has only led to one authorisation (micafungin). At that time, the aim was to determine the dosage and to assess safety and toxicity in trials in oesophageal candidiasis rather than to find treatment options against EC. Meanwhile, there were five authorisations for Candida infections (four first-line and one prophylaxis indication) as well as four for Aspergillus infections (two first-line, two second-line).

Isavuconazole, voriconazole and posaconazole included very rare pathogens across their program (Table 4). The number of studied subjects was relatively small but led to some first-line authorisations (voriconazole for Scedosporium and Fusarium infections) as well as some second-line authorisations (posaconazole for chromoblastomycosis and mycetoma, coccidioidomycosis and fusariosis and isavuconazole for mucormycosis).

4 A Closer Look

4.1 Azoles

4.1.1 Voriconazole

Voriconazole showed in-vitro activities against Candida ssp., Aspergillus ssp. as well as rarer pathogens like Fusarium ssp. and Scedosporium ssp. [10, 11]. Animal models were mainly conducted in guinea pigs [12]. The EPAR mentions PK/PD indices were investigated by comparing MIC and clinical outcome to mean plasma concentrations in each patient, but claims not to have detected a relationship between efficacy and PK/PD indices. In vivo models of voriconazole show AUC/MIC correlates with efficacy for candidiasis [13, 14].

Most indications (candidiasis, fusariosis and scedosporiosis) were based on pooled efficacy data across the programme and additional data for refractory candidiasis, EC, fusariosis and scedosporiosis where no source could be identified.

The Aspergillus indication was based on a phase II invasive aspergillosis study with a historical control (control: 101 subjects), a phase III invasive aspergillosis study and pooled data across the programme [15, 16].

Voriconazole was initially approved for the treatment of invasive aspergillosis, candidemia in non-neutropenic patients, fluconazole-resistant serious Candida infections and serious fungal infections caused by Scedosporium and Fusarium.

At the time of this review, voriconazole was also approved for the prophylaxis of invasive fungal infections in haematopoietic stem cell transplantation (HSCT) recipients.

4.1.2 Posaconazole

Posaconazole was authorised in 2005. In vitro, it was active against various fungi including Aspergillus ssp. and Zygomycetes [17]. The EPAR mentions a correlation between AUC/MIC and efficacy in in vivo models for candidiasis [18]. Additionally, the EPAR claims to have found the same correlation in an analysis of 189 clinical cases.

The clinical evaluation of posaconazole included one phase II study (98 subjects) where we could not determine the indication and therefore could not include it in the table. Phase III consisted of one study including patients that had proven or probable fungal infections refractory to standard treatment. This study used an external control (279 subjects). The majority had Aspergillus infections as well as rarer diseases such as fusariosis, chromoblastomycosis or mycetoma, coccidioidomycosis and cryptococcosis [19]. There was additional data submitted from a pilot study with coccidioidomycosis (20 subjects).

In the end, posaconazole was authorised as second-line medication for invasive aspergillosis, fusariosis, chromoblastomycosis and mycetoma and coccidioidomycosis. Later, this authorisation was broadened to also include oropharyngeal candidiasis in severely sick or immunocompromised patients with little hope to benefit from a topical agent, prophylaxis in patients receiving remission-induction chemotherapy for AML (acute myeloid leukaemia) or MDS (myelodysplastic syndrome), expected to develop neutropenia and in HSCT patients undergoing immunosuppressive therapy for graft-versus-host disease.

4.1.3 Isavuconazole

Marketed in 2015, isavuconazole is the most recent of the drugs marketed in the last 23 years. Isavuconazole showed promising effect in vitro against Aspergillus ssp. as well as against Mucorales [20, 21]. In terms of the PK/PD relationship, the EPAR describes a correlation between efficacy with the AUC/MIC, although there was no comparison amongst the indices. This information stems from studies of murine pulmonary aspergillosis [22, 23].

Phase II consisted of one study in uncomplicated EC and one study in patients with neutropenia undergoing chemotherapy for AML [24].

The phase III programme included two studies: one double-blind, randomised study comparing isavuconazole with voriconazole in patients with invasive aspergillosis and other filamentous fungi [25]. The other phase III study was an uncontrolled (although the applicant did provide amphotericin B studies in order to compare data), single-arm study including patients with invasive fungal disease caused by rare fungi [26].

These studies led to the authorisation for invasive aspergillosis and for treatment of refractory mucormycosis in adults. The applicant had first sought a first-line indication but changed it to second-line.

To this day, no further authorisation for other indications has been granted.

4.2 Echinocandins

4.2.1 Caspofungin

Caspofungin was the first echinocandin to be launched on the market, in 2001. In in vitro studies it performed very well against azole-susceptible and resistant Candida ssp.as well as other rare moulds including Aspergillus fumigatus and Aspergillus flavus [27, 28]. Animal models were conducted in mice and a pharmacokinetic profile was evaluated in rabbits, as well as in monkeys and rats [29, 30]. Even though pharmacokinetic studies were conducted, there is no mention of a preclinical analysis of PK/PD indices in the EPAR. The EPAR claims two population PK/PD analyses including 142 patients with candidiasis and a similar analysis in patients with aspergillosis were conducted but did not come to a conclusive answer. The applicant committed to further investigate it as part of the post-authorisation measure. Published in 2005, an investigation in mouse models suggested a correlation of efficacy with AUC/MIC for candidiasis as well as Cmax/MEC (minimum effective concentration) for pulmonary aspergillosis [31, 32].

The clinical path included 12 pharmacokinetic studies in phase I (312 subjects) and two phase II studies with refractory extrapulmonary aspergillosis and definite/probable pulmonary aspergillosis that were compared with a historical control (206 subjects). Supportive data came from three phase II studies in EC, additional safety data came from one finished EC trial as well as from two ongoing trials at the time (one trial with invasive candidiasis and one for empirical treatment of fever and neutropenia [33, 34]. A compassionate use study in invasive aspergillosis was also included in the EPAR as the only phase III study, although it included only three patients at that time.

In the end, caspofungin was authorised for the treatment of aspergillosis in patients refractory for amphotericin B.

Additional indications were added on later. At the time of this review, caspofungin was also authorised as treatment for invasive candidiasis in adult and paediatric patients and as empirical therapy for presumed fungal infections in neutropenic adult and paediatric patients.

4.2.2 Anidulafungin

Anidulafungin was authorised nearly 6 years after caspofungin. As with other echinocandins, it showed promising in vitro effects against Candida ssp. [35]. In models of candidiasis in mice, anidulafungin was very effective [36]. Even though there were animal PK studies investigating the pharmacokinetic properties, no index is mentioned. A population PK/PD analysis was conducted which associated AUC but not AUC/MIC with probabilities of success at global response at the end of therapy and at the 2-week follow-up. Overall, the EPAR considers data too limited to properly characterise the PK/PD index of anidulafungin. In vivo models published in 2008 show a correlation of efficacy with AUC/MIC as well as Cmax/MIC ratios for candidiasis [37].

The phase II studies were conducted in invasive candidiasis [38]. The main phase III study compared anidulafungin and fluconazole [39]. Another study was discontinued due to insufficient enrolment. Further phase II and III studies mentioned in the EPAR were only included in the population pharmacokinetic analysis mentioned above. They were also included in the safety analysis.

Anidulafungin was authorised for the treatment of invasive candidiasis in adult non-neutropenic patients. No further authorised indication has been added to date.

4.2.3 Micafungin

Being the newest member of the echinocandin class, micafungin demonstrates potent in vitro activity against Candida ssp., being fungicidal at MIC and above as well as showing fungistatic but not fungicidal activity against Aspergillus spp. [40]. Although the EPAR does not mention PK/PD indices being investigated, animal models of disseminated Candida infections show a correlation between AUC/MIC and efficacy [41].

The potent activity against Candida ssp. was investigated in phase II and III studies looking at the effect on oesophageal candidiasis and systemic Candida infections [42, 43]. Additionally micafungin was investigated as prophylaxis in a phase III study in patients undergoing HSCT [44].

Micafungin is authorised for treatment of invasive candidiasis patients ≥ 16 years of age, treatment of oesophageal candidiasis in patients aged ≥ 16 years, treatment of invasive candidiasis in children and neonates < 16 years of age and prophylaxis of Candida infection in patients undergoing HSCT or patients who are expected to have neutropenia (absolute neutrophil count < 500 cells/µL) for 10 or more days in adults and children (and neonates) < 16 years of age.

At the time of this review, micafungin has not been authorised for any additional indications.

5 Discussion

5.1 EPAR Content

Guidelines help applicants identify and investigate relevant data in order to be authorised. In the last 23 years the EMA has changed their guidelines regarding antifungal drugs. From 2003 to 2010, the guideline entitled ‘Points to consider in the clinical evaluation of new agents for invasive fungal infections’ was in place.

It was replaced in 2010 with the ‘Guideline on the clinical evaluation of antifungal agents for the treatment and prophylaxis of invasive fungal disease’. Fig. 3 illustrates the timeline.

Fig. 3
figure 3

The timeline of approval and guidelines for antifungal agents at the EMA over the last 23 years

Full size image

Many changes in clinical practice and research led to this update, including accumulating evidence of the importance of PK/PD indices in selecting doses for anti-infective drugs [45].

As a result, the authorised substances in the last 23 years adhere to different guidelines and therefore include distinct data.

While the two antifungal guidelines have similarities, the guideline from 2003 included less non/preclinical data (see Table 5). How this translates into data represented in the EPAR is shown in Table 6. Although the EPAR represents how the EMA handled PK/PD indices and presents data submitted by the applicant, this does not mean no information was available at the time. In addition, due to the long duration of drug development, preclinical data, concepts and earlier clinical trials for antifungals were conducted at a time when there were no guidelines, templates and no prior experience.

Table 5 Summary of the differences between the guideline from 2010 and the guideline from 2003
Full size table
Table 6 The substances in the context of applicable guidelines at the time and how the guidelines effected data inclusion in the EPARs
Full size table

5.2 Additional PK/PD Guidelines

PK/PD indices have been somewhat neglected in drug development over the last decade but are becoming more important as it becomes apparent that they benefit drug development in various ways—not just investigating dosages but also in determining potency and intrinsic activity as well as suppression of resistance development [45]. Better preclinical methods for determining PK/PD indices as well as a better understanding of dose-optimisation importance might be responsible for the rise in importance over a very short period of time [46].

For other anti-infective drugs, recommendations on exploring PK/PD indices have existed since the year 2000. The guideline ‘Points to consider on pharmacokinetics and pharmacodynamics in the development of antibacterial medicinal products, CPMP/EWP/2655/99’ was put in place that year for antibacterial substances, stating that there is sufficient evidence to support a recommendation that the PK/PD relationship should be investigated for antibacterial medicinal products.

No such recommendation existed for antifungals, though the new guideline on clinical evaluation of antifungals in 2010 did mention PK/PD indices to a larger extent than before. This might be one reason why antibacterial substances that were authorised about the same time as antifungals include substantially more PK/PD data [47].

In 2016, the CPMP/EWP/2655/99 was revised and replaced by the ‘Guideline on the use of pharmacokinetics and pharmacodynamics in the development of antimicrobial medicinal products’ (see Fig. 3). This guideline now applies to antibacterials, antimycobacterials and antifungals alike.

New requirements in the PK/PD sections include the following: applicants should investigate specific nonclinical pharmacokinetic targets (including net static effect, 1-log10 drop in CFU [colony forming units], 2-log10 drop in CFU), provide details on the analysis methods for those indices and must provide urinary concentrations, epithelial lining fluid (ELF) drug concentrations and cerebrospinal fluid concentrations (CSF) when the drug is to be used against relevant indications.

Pipeline antifungals might adhere to the more comprehensive guideline from 2016 and will therefore have to include a lot more PK/PD data.

Currently in the late-stage pipeline (completion of at least one phase III trial) are a few mentionable substances [48]. There is rezafungin (an echinocandin, completed one phase III trial in patients with candidemia and/or invasive candidiasis), VT-1161 (an azole, completed three phase III trials for VVC [vulvovaginal candidiasis]) and ibrexafungerp (completed three phase III studies for VVC) with the latter being a first-in-class inhibitor of the glucan synthesis that may offer the opportunity of oral application in the future [48]. Although detailed information about their preclinical and PK/PD development pathway are not publicly available at the moment, PK/PD evaluations show AUC/MIC correlations for rezafungin and ibrexafungerp while VT-1161 seems to correlate with the free concentration [49,50,51].

5.3 Comparison with Antibacterial Sector

Although there was a big breakthrough in antifungal therapy with the introduction of echinocandins into the market in 2001, the output of the antifungal sector appears low compared with the antibacterial sector. For antifungals, there have been 0.43 newly authorised substances/year including generics (four substances in the last 23 years); without generics there were only 0.26 new substances/year within the last 23 years.

Five of six antifungals have been authorised in the first decade of the 21st century (2000–2010) with only one substance authorised since 2010 (isavuconazole in 2015); as a result, the actual output for 2000–2010 is 0.5 new substances/year and 0.07 new substances/year since 2010.

In comparison, the antibacterial sector has seen 1.6 new substances/year in the last 5 years only; in addition, the antibacterial market remains even more dynamic with the invention of fixed combinations as treatment options [47]. It is noteworthy, however, that in contrast to antifungals the antibacterial sector has seen no new class being introduced in the last 10 years [47].

6 Conclusion

Antifungal drug development presents many challenges. Only six new substances have been approved in the last 23 years, all with very different levels of information presented in the EPARs. This may be partly explained by differences in the applicable guidelines at the time. The standardisation of EPARs and the data required for authorisation may be less developed for antifungals than for antibacterial substances. Although key PK/PD data for antifungals was not studied or was studied only to a limited extent at the time of their authorisation, information on the PK/PD relationship is now available for each substance. There has been an important change in requirements with the introduction of a new PK/PD-specific guideline in 2016 but we still have to wait for a new substance to be approved to see the impact.