Introduction

The last two decades have brought significant advances in treatment for invasive fungal disease. These include the availability of multiple oral agents demonstrating broad spectrums of activity with proven efficacy in treating and preventing invasive fungal diseases [1]. This has greatly advanced care for patients with these infections, however has also introduced new uncertainties related to drug exposure. Therapeutic drug monitoring (TDM) is an established practice for many drugs today such as vancomycin, aminoglycosides, antiepileptics, and immunosuppressants. Agents that are ideal candidates for routine TDM demonstrate three key characteristics—a high degree of inter-patient variability in dose-exposure relationship, an established relationship between drug exposure and either efficacy, safety, or both, and an assay that is able to accurately measure drug concentrations. Increased understanding of antifungal pharmacokinetics and pharmacodynamics has led to a growing role for TDM of certain antifungals in routine clinical practice. Currently, there is evidence to support the use of TDM for itraconazole, voriconazole, posaconazole, and flucytosine. A number of review articles have been published that describe the rationale and benefit of TDM for certain antifungal agents [2–5]. More recently, clinical guidelines have been published by the British Society for Medical Mycology (BSMM), as well as the Japanese Society of Chemotherapy and the Japanese Society of Therapeutic Drug Monitoring (JSC/JSTDM) [1, 6]. The purpose of this article is to describe the antifungal agents for which we conduct TDM, and outline its role based on a review of recent literature.

Situations That Favor TDM of Antifungals

There are numerous clinical circumstances that can increase the variability in dose-exposure relationship in antifungal agents. These include patients that display altered or unpredictable pharmacokinetics such as critically ill patients, patients with severe renal or hepatic dysfunction, extremes of age, and obesity [2]. Secondly, given that the azole antifungals are extensively metabolized via the cytochrome p450 enzymes, there are numerous drug-drug interactions that can alter serum concentrations of these medications [7]. Poor oral bioavailability is another factor that plays a large role in the high degree of variability in serum concentrations of certain antifungals, most notably itraconazole and posaconazole. Patients with conditions that impact oral absorption, such as diarrhea or mucositis, are therefore also good candidates for TDM. Severity of infection should also be kept in mind, as TDM is more likely to be useful in treatment of severe disseminated disease or prophylaxis in neutropenic patients at high risk for life-threatening fungal infections. Another situation in which TDM of antifungals has been advocated for is in the assessment of compliance, although less costly alternatives, such as medication refill histories, may be preferable if available.

Perhaps the most important question to ask prior to conducting TDM is whether or not the resulting concentration will alter the treatment plan. For example, if a patient with a confirmed invasive fungal infection is clinically failing therapy on an antifungal agent, it would generally be in the patient’s best interest to switch to an alternative agent as opposed to waiting for a concentration to come back prior to making a decision; particularly when the long turn-around time is considered. Conversely, neutropenic cancer patients receiving posaconazole suspension as prophylaxis against invasive fungal infection represent very good candidates for TDM. Since patients are not yet infected, posaconazole concentration is the only available indication of whether or not they have adequate exposures. In addition, these patients often have poor oral intake and other conditions that predispose them to variable absorption.

Assay Availability

There are two types of assay that have been used for measuring antifungal drug concentrations—microbiological assays or bioassays, and chromatographic assays including liquid chromatography-mass spectrometry (LC-MS) and high-performance liquid chromatography (HPLC). Microbiological assays are relatively inexpensive; however, their major drawback is reduced accuracy due to the potential for interference with other drugs, particularly in the setting of combination antifungal therapy [8].

Overall, concentrations obtained with chromatographic assays and bioassays are similar for patients receiving a single antifungal agent [2]. The exception to this is itraconazole where the production of an active metabolite, hydroxy-itraconazole, cannot be differentiated by bioassay. Caution must be used when converting between itraconazole concentrations measured by bioassay and those measured by HPLC, as the cited degree to which bioassay overestimates HPLC-measured concentration varies in the literature, with a range of a 2–7-fold overestimation being reported depending on the source [1, 2, 9]. This may not have significant clinical impact however, as it is a common clinical practice to combine itraconazole and hydroxy-itraconazole concentrations when describing itraconazole exposure.

Given these considerations, chromatographic assays are the preferred and most commonly utilized assays for antifungal TDM. Chromatographic assays demonstrate high sensitivity and specificity [10, 11]. Unfortunately, these assays are only performed by a relatively small number of specialized laboratories, and as a result, it can take a week or more to receive results. Turn-around time is therefore an important consideration when designing an optimal TDM approach for a specific institution.

Antifungals Without a Current Role for TDM

A clear exposure-response relationship has been documented for both amphotericin B and the echinocandins [12]. However, current pharmacokinetic data does not suggest that there is enough variability in dose-exposure relationship to warrant routine TDM for patients on these therapies [4]. As both classes are only available in IV formulations, there are no issues with variable absorption. There is potential for an expanded role for TDM of glucan synthesis inhibitors in the future, as there is currently an oral enfumafungin derivative, a class of glucan synthesis inhibitors structurally distinct from the echinocandins, undergoing a phase 2 trial for treatment of invasive candidiasis, but more data are needed [13].

Fluconazole demonstrates a well-documented exposure-response relationship, with an AUC/MIC >50 associated with clinical efficacy [14–16]. However, due to its excellent oral bioavailability and linear kinetics, it lacks the dose-exposure variability required to consider TDM [17]. Certain populations, such as clinically unstable patients or patients receiving renal replacement therapy, can still experience unpredictable dose-exposure relationships which could make TDM more appropriate. Illustrating this, van der Elst et al. demonstrated a wide range of fluconazole serum concentrations (2.6–46.5 mg/L), 40 % of which were subtherapeutic, in critically ill pediatric cancer patients [18]. However, even among these special populations, the lack of a clear target limits the role of TDM for fluconazole as all of the pharmacodynamic data establishing exposure-efficacy relationship utilizes AUC/MIC, which is not a clinically practical monitoring parameter. The remainder of this review will focus on agents with a demonstrated role for TDM.

Itraconazole

Inter-Patient Variability

Extensive dose-exposure variability of itraconazole has been reported in pharmacokinetic analyses, with the most important contributing factor being oral absorption [19]. The effect of gastric pH and food on absorption of itraconazole varies based on the formulation used—the capsules require an acidic pH and administration with food for optimal absorption, while the solution is absorbed best on an empty stomach [20]. Overall, the solution results in better exposure than the capsules; however, patients on either formulation still experience a high degree of variability [19]. Other factors that can alter oral absorption of itraconazole include mucositis, diarrhea, and acid suppressants. Itraconazole is primarily metabolized by CYP3A4, and inducers or inhibitors of this enzyme can also affect drug concentrations [2].

Concentration-Efficacy Relationship

There are a number of small studies that demonstrate a relationship between itraconazole trough concentration and efficacy for both prophylaxis of invasive fungal infection in immunocompromised patients as well as treatment of invasive fungal infection (Table 1). Studies examining the use of itraconazole for prophylaxis of invasive fungal infections in neutropenic patients have associated concentrations of 0.25–0.5 mg/L with efficacy [21–23]. Concentration breakpoints established as potential targets for successful treatment of invasive fungal infection range from 5 to 6.5 mg/L when measured by bioassay and 0.6–1 mg/L when measured by HPLC [19, 24–28]. There are a number of limitations associated with interpreting results of these studies, as they are retrospective, include a small number of patients, do not have standardized reasons or timing for obtaining concentrations, and are not designed to establish an ideal concentration target.

Table 1 Literature summary of concentration targets associated with clinical efficacy
Full size table

Most clinical practice guidelines recommend that trough concentrations be maintained over 0.5 mg/L for prophylaxis and at least 1 mg/L for treatment, measured by HPLC (Table 2). The BSMM guidelines recommend a minimum trough concentration of 0.5 mg/L for both prophylaxis and treatment, although based on the limited evidence available, there seems to be more support for a treatment trough target of at least 1 mg/L [1]. It is also important to consider that ideal concentrations likely vary based on the MIC of the infecting organism and site of infection, but there is insufficient data to support more tailored targets at this point.

Table 2 Clinical guideline antifungal TDM recommendations
Full size table

Concentration-Toxicity Relationship

Adverse events attributable to itraconazole have been thoroughly described in other antifungal pharmacology reviews and include nausea and diarrhea, liver function abnormalities, a negative inotropic effect, and mineralocorticoid effects leading to edema, hypokalemia, and hypertension [20, 56]. A retrospective chart review by Lestner et al. demonstrated a clear positive correlation between serum concentration and risk of adverse events, and the authors proposed a concentration of 17.1 mg/L measured by bioassay as an upper limit for preventing toxicity. Based on CART analysis, 86 % of patients at or above this concentration would be predicted to experience toxicity versus only 31 % of patients below this concentration [57].

In line with these results, the BSMM guidelines recommend that 17 mg/L (measured by bioassay) is a reasonable limit in order to reduce risk of toxicity [1]. Other clinical guidelines and review articles on antifungal TDM do not take a strong stance on upper limit itraconazole concentrations to reduce toxicity (Table 2).

Recommendations

Both Andes et al. and the BSMM guidelines recommend routine monitoring of itraconazole concentrations after the initiation of therapy or dose changes independent of indication [1, 2]. Due to the highly variable oral absorption of itraconazole, TDM is appropriate for most patients receiving the drug. Because of the long half-life of itraconazole, it can often take a week for the drug to reach steady state conditions, and the first measurement should not be taken before this point. [20] Concentrations should ideally be measured as trough concentrations; however, due to the long half-life of itraconazole, concentrations drawn in the middle of the dosing interval do not differ substantially from troughs.

After a patient reaches steady state therapeutic concentrations, recommendations on when to check serum concentrations are less defined, but rechecking can be considered for patients who are initiating or discontinuing major interacting medications, who have new concerns about oral absorption, or who demonstrate lack of response or signs of toxicity [1, 2]. As previously mentioned, concentrations should only be monitored if the results are likely to impact clinical decision-making while keeping in mind that the evidence supporting specific targets is limited.

Dose Adjustments

Dose adjustment algorithms for patients with sub- or supra-therapeutic concentrations of other azole agents have been published recently; however, recommendations for itraconazole are sparse. The BSMM guidelines recommend that the dose be increased from 200 mg twice daily to 300 mg twice daily for patients with subtherapeutic concentrations, but no citation is given [1]. Initial interventions for patients with subtherapeutic itraconazole concentrations include switching from the capsule formulation to the liquid, ensuring the liquid is taken on an empty stomach and the capsule with food, and elimination or reduction of interacting medications.

Recommendations for addressing supra-therapeutic itraconazole concentrations are even more limited. It is important to keep in mind that dose reduction should typically be reserved for patients who are actually experiencing adverse events and for whom alternate agents would be inappropriate.

Voriconazole

Inter-Patient Variability

Unlike itraconazole, voriconazole has dependable oral bioavailability [20]. However, there is still a high level of both inter- and intra-patient variability in dose-concentration relationship [41, 58]. This is attributed primarily to genetic polymorphisms in CYP2C19, which vary in prevalence based on ethnicity and can lead to dramatic differences in drug exposure [59, 60]. No dosage adjustments are currently recommended based on genotype, as the influence of genotype on voriconazole serum concentration is confounded by other factors such as drug-drug interactions [60]. Drugs that induce or inhibit CYP2C19, such as omeprazole, can also affect voriconazole concentrations [7].

Despite its good oral bioavailability, IV to oral transition also affects serum concentrations due to differences in approved dosing regimens. The approved intravenous dose is 4 mg/kg twice daily, while the approved oral dose is 200 mg twice daily. This can lead to large differences in dosing and subsequently exposure, particularly in overweight patients, which has prompted some expert recommendations for oral dosing to approximate the 4-mg/kg-IV dosing rounded to the nearest available tablet strength [61]. More recent data also suggests that oral bioavailability of voriconazole may be lower in patients being treated for invasive fungal infection than in healthy patients [62].

Concentration-Efficacy Relationship

Several studies have been published that support a relationship between voriconazole concentration and efficacy, while only one study failed to demonstrate a correlation (Table 1). Specific concentrations associated with successful therapy in these studies ranged from 0.25 to 2.2 mg/L, with most studies supporting a concentration target in the 1–2-mg/L range. A meta-analysis of 12 studies including patients with deep-seated mycosis found that trough concentrations over 1 mg/L were significantly associated with success in eradicating the infection (OR 7.23, 95 % CI 2.84–18.36, p < 0.0001) [63].

Perhaps the strongest evidence supporting TDM for voriconazole is a recent randomized controlled trial in which patients receiving voriconazole for treatment of invasive fungal infection who had their doses adjusted based on TDM to a target trough concentration of 1–5.5 mg/L had a significantly higher rate of either complete or partial response than patients without TDM [33].

Clinical practice guideline recommendations on efficacy targets for voriconazole are reported in Table 2. The target of 0.5 mg/L recommended by the IDSA guidelines on treatment of histoplasmosis may be outdated, and guidelines incorporating more recent data recommend a target of at least 1 mg/L for treatment of invasive fungal infections [1, 2, 6]. A higher trough target of 2 mg/L may be considered based on the clinical situation; however, attempts at increasing voriconazole exposure to this point in all patients may lead to unacceptable toxicity. One potential strategy noted in the BSMM guidelines is to utilize a trough:MIC ratio in place of a static trough target, which is supported by an in vivo pharmacokinetic model as well as a Monte Carlo simulation of nine clinical trials [64, 65]. Recommendations for target trough concentrations for prophylactic use of voriconazole range from 0.5 to 1 mg/L, although support for these targets is much more limited than for targets for treatment of established disease [1, 2, 29, 30].

Concentration-Toxicity Relationship

Toxicities attributable to voriconazole include visual disturbances, hepatotoxicity, encephalopathy, and cutaneous phototoxicity [20]. Risk of visual disturbance, hepatotoxicity, and encephalopathy, all appear to correlate with increasing voriconazole concentration [35, 63, 66, 67]. While visual disturbances are dose dependent, they are also typically transient and self-limiting, and thus do not create a need for TDM in order to limit risk [2]. Neurologic toxicity has been associated with concentrations greater than 4–5.5 mg/L [35, 63]. Hepatotoxicity is more challenging, because while it does appear to be dose dependent, there is no clear concentration cut-off at which patients experience a significantly increased risk [63].

Published guidelines on TDM recommend that concentrations above 4–6 mg/L should be avoided (Table 2). Some experts argue that regular TDM, in order to avoid hepatotoxicity, is unnecessary because routine monitoring of LFTs should be sufficient [6, 68]. However, utilization of TDM still has potential utility in establishing the most probable cause of hepatotoxicity in patients who have multiple risk factors, as well as in avoiding other adverse effects, such as encephalopathy, that would not otherwise be identified by routine laboratory monitoring of these patients.

Recommendations

Similarly, to itraconazole, clinical practice guidelines recommend consideration of TDM of voriconazole in the majority of patients [1, 2, 6]. Situations in which TDM is particularly attractive include populations with unpredictable pharmacokinetics, such as pediatric patients, patients receiving strongly interacting medications, or patients displaying signs of toxicity. Due to the disparity between intravenous and oral dosing regimens, the intravenous to oral transition is another point when TDM should be considered, particularly in overweight patients. Because much of the concentration variability with voriconazole is due to genetic polymorphisms in CYP2C19 and not due to absorption issues, even patients receiving IV therapy may have variable concentrations, and thus TDM can still be considered. However, if a patient is failing to respond to appropriate IV dosing of voriconazole for an invasive fungal infection, it may be preferable to switch to an alternate therapy as opposed to waiting for results of concentration monitoring.

While voriconazole reaches steady state earlier than itraconazole, Michaelis-Menten kinetics lead to variable half-life, so it is still important to wait an appropriate period of time prior to monitoring concentrations [20]. Both the JSC/JSTDM guidelines and Andes et al. recommend waiting a minimum of 4–5 days prior to obtaining a concentration, while the BSMM guidelines state that a concentration can be obtained earlier than this as long as a second concentration is repeated a few days later to ensure that the drug is at steady state (Table 2). It is important to note that an apparent autoinduction of voriconazole metabolism can be seen in some patients, which has the potential to lead to large fluctuations in concentrations in patients who were initially therapeutic [58, 62]. Therefore, concentrations should be followed over time in patients with invasive fungal diseases being managed with voriconazole. Similar to other azole antifungals, monitoring should be via trough concentrations.

Dose Adjustments

All patients with subtherapeutic voriconazole concentrations should be assessed for adherence, as well as interacting medications that can induce metabolism of voriconazole, such as rifampin. Dose adjustment algorithms have been published for voriconazole, which, along with the BSMM guidelines, recommend doubling or halving the dose of voriconazole for sub- or supra-therapeutic concentrations, respectively [1, 33, 35]. This was also the dose adjustment strategy used by Park et al. in their randomized controlled trial of TDM versus no TDM [33]. However, less aggressive dose titrations than this may be appropriate, particularly for patients close to target, since the Michaelis-Menten kinetics mean that small-dose adjustments can lead to dramatic changes in drug exposure, and smaller titrations are possible due to the availability of 50-mg tablets [69].

The BSMM guidelines recommend avoiding doses over 6 mg/kg twice daily for the IV formulation or 300 mg twice daily for the oral formulation [1]. Per pharmacokinetic modeling however, many patients may need a higher oral dose than 300 mg twice daily to achieve therapeutic concentrations [62, 69].

Posaconazole

Inter-Patient Variability

Similar to itraconazole and voriconazole, posaconazole exhibits significant inter-patient variability in dose-exposure relationships [70, 71]. This is predominantly due to inconsistent absorption of the oral suspension formulation. Posaconazole is a lipophilic molecule with very poor aqueous solubility, which requires administration with a high-fat meal or, less optimally, an acidic beverage [20, 42, 72]. Poor oral absorption has led to unacceptably high rates of subtherapeutic concentrations, with data published from a reference lab demonstrating that 16.3 % of samples analyzed had undetectable concentrations, and 70.3 % had concentrations below 0.7 mg/L [73].

A relatively new development that could dramatically alter TDM for posaconazole is the approval of delayed release posaconazole tablets, as well as an intravenous formulation. The tablet formulation achieved average steady state concentrations of 1.46 mg/L in a phase I trial of patients with acute myelogenous leukemia (AML) or myelodysplastic syndrome (MDS), which is substantially higher than the average posaconazole concentration of 0.58 mg/L seen in the same patient population receiving the suspension [74]. The tablet is also less affected by food or stomach pH, although it is still recommended to be administered with a meal [75]. The intravenous formulation also achieved average serum posaconazole concentrations greater than 1 mg/L [76]. These more dependable exposure results reduce the need for TDM with the newer formulations.

Concentration-Efficacy Relationship

Several small observational studies have supported an exposure-efficacy relationship for posaconazole when used for prophylaxis of invasive fungal infections in immunocompromised patients, most of which support a target above 0.5–0.7 mg/L (Table 1). The most convincing evidence comes from an FDA briefing document compiled from the results of two large observational studies, which found that 12 out of 15 breakthrough fungal infections occurred in patients with serum posaconazole concentrations below 0.7 mg/L [47]. These results are reflected in the target concentrations recommended by published guidelines for posaconazole prophylaxis (Table 2).

Data are much more limited on the exposure-response relationship for posaconazole when used to treat established invasive fungal infections. Walsh et al. found a positive correlation between exposure and response in patients receiving salvage treatment for invasive aspergillosis, with the highest response rate in patients with a mean average concentration of 1.25 mg/L [50]. Other published studies did not reflect a relationship between initial posaconazole concentrations and response in the treatment of invasive fungal infections; however, mean posaconazole concentrations in both studies were greater than 1.25 mg/L, which may have limited the ability to detect a relationship [51, 52]. As can be expected from the small amount of data available, guideline recommendations vary widely in the recommended target posaconazole concentration for these patients (Table 2), but a trough concentration of 1 mg/L appears to be a reasonable minimum goal.

Concentration-Toxicity Relationship

Unlike other antifungals, a relationship between concentration and toxicity has not been documented for posaconazole. In the previously referenced FDA briefing document, adverse effects did not vary significantly based on posaconazole concentration [47]. Thus, at this point, no upper limit is recommended in order to limit toxicity (Table 2). Questions remain about whether the higher posaconazole exposures achieved with the intravenous and delayed release tablet formulations will lead to new dose-dependent toxicities. While a much larger proportion of patients receiving these formulations experience concentrations over 2.5 mg/L, thus far there have been no differences seen in adverse events between these formulations and the oral suspension [76].

Recommendations

The BSMM guidelines recommend that TDM be conducted for the majority of patients receiving posaconazole at the initiation of therapy [1]. Based on the known exposure-response relationship and the substantial issues with oral absorption of posaconazole suspension, TDM of patients receiving this formulation is recommended. This is especially true of patients who cannot comply with the dietary recommendations or are receiving acid suppressants.

There is no apparent need to perform TDM for the intravenous formulation of posaconazole as inter-patient variability is due to variable oral absorption, and 100 % bioavailability is ensured with intravenous administration. Necessity of TDM with the delayed release tablet formulation is less clear at this point. In the phase I study in AML or MDS patients, the 300 mg once daily dose of the tablet achieved average serum posaconazole concentrations over 0.5 mg/L in 100 % of patients [77]. It therefore seems that routine TDM would be unnecessary for patients receiving prophylaxis with the tablet formulation. It is possible that there may still be a need for TDM in patients requiring higher exposures for treatment of invasive fungal infections, but this can likely be limited to situations where there is significant concern for altered oral absorption, such as severe mucositis or prolonged severe diarrhea.

Posaconazole suspension does not reach steady state until 1 week after initiation or dose changes [5]. Therefore, it is recommended to wait until day 7 of therapy prior to obtaining a concentration [1]. One published dosing algorithm uses a concentration of at least 0.35 mg/L 48 h after initiation as a surrogate for therapeutic steady state concentrations and recommends dose adjustments at that point. [47] Since its pharmacokinetics are relatively predictable, this is a more reasonable strategy for posaconazole relative to the other azoles. Similar to other azole antifungals, it is recommended that concentrations be drawn as troughs; however, most studies documenting concentration-response relationship utilize average concentrations. Similarly, to itraconazole, posaconazole’s long half-life means that random and trough concentrations should not differ significantly [20].

Dose Adjustments

A recently published dosing algorithm for posaconazole demonstrated that patients with subtherapeutic concentrations on standard prophylaxis dosing of 200 mg three times daily can reach target concentrations after a dose increase to either 300 mg three times daily or 200 mg four times daily [78]. However, due to saturable oral absorption, dose increases past this point are unlikely to result in further increases in serum concentration [46]. Because there has not been a demonstrated upper limit for posaconazole concentrations to prevent toxicity, dose reduction based on TDM is not warranted.

Due to its saturable oral absorption, it is important to consider strategies outside of dose titration to increase posaconazole exposure, including ensuring the patient is receiving the suspension with a fatty meal or acidic beverage, eliminating interacting medications such as acid suppressants, and dividing the daily dose so that it is administered four times daily. Smaller, more frequent administrations (i.e., 200 mg four times daily) achieve higher serum concentrations than larger, less frequent administrations (i.e., 400 mg twice daily). Patients who cannot achieve desired concentrations on posaconazole suspension despite optimal administration can be transitioned to the delayed release tablet formulation—this was demonstrated to dramatically increase exposure in one case series [79].

Isavuconazole

A new azole antifungal, isavuconazole, has recently been approved for use based on r phase III trials in which it demonstrated non-inferiority to voriconazole in the treatment of invasive aspergillosis with significantly fewer adverse effects [80]. Isavuconazole has a number of attractive features, including both intravenous and oral formulations, a long half-life that allows for once daily dosing, excellent oral bioavailability, few adverse effects, and a wide spectrum of activity including Candida, Aspergillus, and some Mucorales [81–84].

Many of the factors contributing to variability in serum concentrations of other azole antifungals are not present with isavuconazole, as it demonstrates excellent oral bioavailability that is unaffected by food as well as linear pharmacokinetics [85, 86]. In line with this information, Schmitt-Hoffmann et al. demonstrated that there was only low to moderate intersubject variability in serum concentrations [87]. Thus, while any definite conclusions must be deferred until there is more clinical experience with isavuconazole, early evidence suggests that the apparent need for TDM of the drug is low.

Flucytosine

Inter-Patient Variability

Flucytosine is almost exclusively eliminated as unchanged drug via the kidneys [17]. Thus, changes in renal function can have a dramatic effect on flucytosine serum concentrations. Given that current use of the drug is nearly always in combination with the highly nephrotoxic amphotericin B for treatment of cryptococcal meningitis, changes in renal function and therefore changes in flucytosine exposure can be expected [88].

High levels of variability in serum concentrations of flucytosine have been documented, with retrospective reviews of concentrations measured in both adult and pediatric populations demonstrating that only about 20 % of serum concentrations analyzed were within therapeutic range [88, 89]. However, there are significant limitations to these data given that flucytosine doses were not recorded. In addition, definitions for therapeutic serum concentrations used were largely arbitrary due to an overall lack of clinical data supporting concentration targets.

Concentration-Efficacy Relationship

There is a paucity of data supporting specific concentration targets for efficacy of flucytosine, with the limited data available coming from murine models of invasive candidiasis [90, 91]. These models demonstrate that time >MIC is the parameter most associated with success of flucytosine therapy, with the optimal target being 25 % of the dosing interval in one study [91] and 45 % in another [90]. It is unknown if these targets can be applied to cryptococcal infections, and there is currently no clinical efficacy data in humans supporting a specific flucytosine concentration target.

One potential rationale for minimum flucytosine concentration targets in the absence of strong correlations with efficacy is the prevention of resistance. In an in vitro study, concentrations below 25 mg/L were associated with development of resistant mutants in Candida albicans and Candida glabrata [92]. This is the evidence cited by major clinical practice guidelines to justify current minimum concentration targets (Table 2). It is important to note that this study was examining flucytosine alone, and it is possible that lower concentration targets are acceptable when the drug is used in combination in accordance with its current role in therapy [88].

Concentration-Toxicity Relationship

TDM for flucytosine has routinely been considered the standard of care primarily for the reason of toxicity prevention. Elevated flucytosine concentrations have been linked to both myelosuppression and hepatotoxicity [93–95]. In the largest study examining flucytosine toxicity, 62 % of patients with prolonged flucytosine peak concentrations of over 100 mg/L experienced flucytosine toxicity versus 31 % without concentrations over 100 mg/L [93]. Of note, the authors of this study defined peak concentrations as concentrations drawn 2 h after a dose as opposed to the more traditional definition of 30 min after a dose. Also, they used a dose of 150 mg/kg/day in patients with normal renal function, which is higher than the currently recommended dose of 100 mg/kg/day for cryptococcal meningitis [88]. In a small study, out of 16 patients that received oral flucytosine dosed at 100 mg/kg/day, none had peak concentrations over 75 mg/L [96].

Clinical practice guideline recommendations for minimum flucytosine concentrations to prevent toxicity are presented in Table 2. The cut-offs recommended vary significantly and, given that the only concentrations definitively linked to toxicity are those exceeding 100 mg/L, exact cut-offs below this point are somewhat arbitrary. However, there is evidence that lower doses than those currently used may provide equivalent efficacy in the treatment of cryptococcal meningitis [97]. Given that flucytosine demonstrates time-dependent killing and has very weak concentration-dependent effects, it is reasonable to target peak concentrations well below the traditional cut-off of 100 mg/L so long as adequate time >MIC is achieved.

Recommendations

All major published guidelines commenting on TDM of flucytosine recommend that concentrations be monitored early in the course of therapy in the majority of patients to minimize the risk of toxicity (Table 2). Because myelosuppression is a delayed, progressive toxicity, checking a peak concentration early in the course of therapy is an important step to preventing this from occurring [95]. In patients who have acceptable concentrations, it is not necessary to continue to check concentrations so long as renal function is monitored closely, and the flucytosine dose is appropriately renally adjusted.

In patients receiving flucytosine orally at recommended doses, peak concentrations and trough concentrations are not significantly different [95]. Thus, it is reasonable to only monitor peak concentrations.

Dose Adjustments

Recommendations regarding how to adjust flucytosine dose based on sub- or supra-therapeutic concentrations are sparse. The BSMM guidelines recommend that the dose should be increased by 50 % in patients with a subtherapeutic concentration [1]. Given the extremely limited data supporting minimum flucytosine concentrations and the definite risk associated with supra-therapeutic concentrations, dose increases should be implemented with caution. No recommendations are available on how to address a supra-therapeutic peak concentration, but given the linear pharmacokinetics and the minimal concentration-dependent killing of the drug, doses can be lowered fairly aggressively and proportionally in patients with concentrations over 100 mg/L.

Conclusion

There is a growing body of literature that supports the utility of TDM for flucytosine, itraconazole, voriconazole, and posaconazole suspension. All of these agents demonstrate the variability in drug exposure and clear correlations between exposure and efficacy, toxicity, or both that make them ideal candidates for TDM. The role of TDM in newer agents such as posaconazole delayed release tablets and isavuconazole is still unclear, but preliminary data suggest that it may not be needed. Additionally, clinicians must remember that current concentration target recommendations are based on observational studies with varying patient populations and significant limitations. Thus, the decision on whether or not to conduct TDM as well as the specific concentration to target must be assessed on a patient-by-patient basis.