Background
Invasive fungal infections (IFIs) are substantial causes of morbidity and mortality in immunocompromised hosts, such as patients with hematological malignancies and solid-organ transplant recipients [
1]. Prophylaxis was widely accepted as an important intervention in this setting [
2]. Posaconazole is a second-generation triazole agent with antifungal activity against a wide range of yeasts (candida species) and molds (
Aspergillus species,
Zygomycetes, and
Fusarium species) [
3,
4]. It has been strongly recommended as a prophylaxis of IFIs by guidelines from IDSA and ESCMID with high-quality evidence [
5‐
7]. The US Food and Drug Administration (FDA) have approved three formulations, including the oral suspension, the recently delayed-release tablet and intravenous formulations. Due to the large interindividual variability in bioavailability and drug-drug interactions, therapeutic drug monitoring (TDM) is advised by IDSA and FDA in order to ensure adequate exposure and optimize clinical efficacy for posaconazole suspension [
5,
8,
9].
The growing studies reported that there is a significant exposure-response relationship between posaconazole plasma concentrations and prophylactic efficacy [
10‐
12]. Posaconazole TDM is also increasingly used in clinical practice to achieve a plasma concentration target of 0.5 mg/L at steady state which is equivalent to the minimum inhibitory concentration (MIC
90) of posaconazole for most
Aspergillus spp. and was also recommended by the 4th European Conference on Infections in Leukaemia (ECIL-4) [
13]. Thus, a stable drug concentration at 0.5 mg/L has been suggested in several posaconazole TDM studies [
14‐
17]. Nevertheless, Jang et al. recommended a target at 0.7 mg/L which was also adopted in FDA document [
8,
12]. Meanwhile, posaconazole showed a good long-term safety profile compared with voriconazole and itraconazole [
18‐
20]. Therefore, the utility of posaconazole TDM remains a controversial issue and most of related studies are limited by single-center practice and small sample size.
Although exposure-response relationship has been examined in several studies, it is still unclear whether TDM should be routinely performed during the process of posaconazole prophylaxis. Furthermore, there was no final consensus reached about posaconazole concentration target for prophylactic use to date. The aim of this study was to assess the relationship between posaconazole plasma concentration and clinical prophylactic efficacy and to define the optimum posaconazole concentration based on a meta-analysis.
Discussion
This meta-analysis was designed to assess the exposure-response relationship between the reported posaconazole concentration and clinical prophylactic efficacy. Our pooled analysis demonstrated that a steady-state posaconazole target of ≥0.5 mg/L is more predictive for successful prophylaxis than the target of ≥0.7 mg/L. The results indicated that with posaconazole levels at 0.5 mg/L or higher, patients with hematological malignancies were twice more likely to achieve a successful prophylaxis.
We considered possible IFI as failed prevention because patients are usually enrolled in empirical treatment programs once they were diagnosed as possible IFIs. An ideal prophylaxis should prevent even a possible IFI result which could probably degenerate into probable or even proven and cause more medical costs and prolonged hospitalizations [
5,
6]. Before our pooled analysis, there were two putative targets of steady state plasma, 0.5 mg/L and 0.7 mg/L, which were suggested as threshold levels for IFIs prophylaxis. The first target of 0.5 mg/L is based on the MIC
90 for most
Aspergillus species [
23] and the second target of 0.7 mg/L, is considered the efficacy threshold by the FDA. [
24] The evidence of the recommendation from FDA is based on a RCT trial in high risk patients with graft versus host disease (GVHD) after allogeneic hematopoietic stem cell transplantation (allo-HSCT) and the results showed that the successful prophylactic rate would be 98.1% when patients’ posaconazole average concentration reached over 0.7 mg/L, which is similar to our outcome (95.8%) [
12,
21]. Ullmann et al. used the average posaconazole concentrations while most of our data were from trough concentrations which are more available in clinical practice [
21]. As usual, the posaconazole average concentration is higher than the trough concentration although there is no significant fluctuation after steady state [
19]; thus, it is reasonable that an average level of 0.7 mg/L could be considered as the threshold in Jang’s study and a lower trough level at 0.5 mg/L in our analysis [
12].
It has been reported that IFIs are associated with high morbidity and mortality in patients with hematological malignancy [
5,
6,
14]. Our study demonstrated that performing TDM could help posaconazole concentrations reach the threshold 0.5 mg/L and then improve the successful rate from 89.0% (95% CI 87.4%–90.6%) to 95.9% (95% CI 95.2%–96.7%) for prophylactic usage. The successful rate will also be improved from 90.3% (95% CI 89.0%–91.7%) to 95.8% (95% CI 94.9%–96.6%) if the posaconazole concentration reached at 0.7 mg/L or higher, yet there was no statistical significance. The possible explanation of this result is that patients with posaconazole stable concentration at the range of 0.5–0.7 mg/L were mostly identified as a successful prevention; thus, the target of 0.7 mg/L was evaluated with no significance. Additionally, it has been proved that posaconazole concentrations in pulmonary alveolar cells are over 40-fold higher than those in plasma [
53]. So the high posaconazole aggregation pulmonary could be explicable for achieving satisfactory effects under low posaconazole plasma levels.
For the result of subgroup analysis based on the underlying disease, a target of 0.5 mg/L on the steady state is recommended in patients with hematological malignancies but not for cardiothoracic transplant recipients. There was only one study involved in the solid-organ transplant group at the target of 0.7 mg/L, so whether 0.7 mg/L could be a target of this cardiothoracic transplant population still need further studies to confirm. Our results showed that the prophylactic efficacy have no significant difference when patients’ posaconazole levels were over or below 0.7 mg/L in cardiothoracic transplant recipients. The chief reasons are that posaconazole concentrations of this population showed great variability [
32] and studies on this population are limited at present. Therefore, the utility of TDM in cardiothoracic transplant recipients and the specific posaconazole target warrant further investigation.
Both heterogeneity and the publication bias in our pooled analysis are low. Sensitivity analysis was done by exclusion of studies with NOS ≤ 6, intervention following TDM results and small sample size with
n < 30. It shows that the significance of main outcome at 0.5 mg/L remained stable after excluding studies mentioned above or removing each individual study, which confirmed the high reliability and stability of our meta-analysis. However, after excluding 6 studies following TDM intervention [
35,
45,
47,
48,
51,
52], the insignificant outcome (
P = 0.08) of the target at 0.7 mg/L turned into significant (
P = 0.03). This change might attribute to the influence of dosing adjustment, which could interfere the standard-compliant samples or the prevention outcome.
A particular section of our meta-analysis was the result verification using of TSA. According to the TSA results, it is noticeable that the Z curves were not stable with the growing data. For the TSA result of 0.7 mg/L, the Z curve escaped out of the conventional boundary even after reaching the optimal sample size. Study from Sengar was supposed to be the main reason contributing to this reverse change because the Z curve turned back to the conventional boundary after adding the other two studies (Vanstraelen and Liebenstein) into the analysis. However, only the abstract is available in Sengar’s study and maybe the race of the participants (Asian) could be a possible reason to explain this result. In brief, the TSA results under two posaconazole concentration targets indicated that there are sufficient information to support our conclusion. However, only two RCTs are available in our study, a well-designed prospective trial is needed to verify our results.
According to the involved studies, we found that posaconazole concentration is higher when patients are administered with those two new formulations, delayed-release tablet and injection formulation due to the stable bioavailability [
5]. Thus, the number of patients with subtherapeutic levels is less and there were five and two studies involved in groups of tablet and injection in our analysis, respectively. Since posaconazole delayed-release tablet and the injection forms could increase the possibility of achieving the target, whether TDM is useful in this case still needs future investigation with large sample size. To date, posaconazole oral suspension formulation is still widely used and remains available worldwide, and this form is still an important option for patients with nasogastric tubes or those unable to take tablets [
11]. Therefore, the target of 0.5 mg/L is required and should be recommended during the TDM process for patients administered with posaconazole oral suspension.
It was reported that posaconazole showed a good safety profile during a standard long-term administration. The most common adverse event which related to the treatment is gastrointestinal tract disturbances [
20,
54,
55] and the incidence of serious adverse events is 6% - 13%, including the hepatotoxicity, QTc prolongation, etc. [
20,
21]. Thus, studies about the relationship between posaconazole exposure and adverse events are still limited, which make it not viable to explore the threshold for safety concentration like voriconazole [
56,
57]. To date, posaconazole appears to have a more favorable safety and tolerability profile than voriconazole [
58,
59]. Only a few studies accessed the relationship between posaconazole exposure and treatment efficacy, which make it infeasible to define a posaconazole concentration target by meta-analysis. Walsh et al. conducted a study in which a cohort of 67 patients who received posaconazole for salvage treatment of invasive
Aspergillosis, the results demonstrated that the cure rates increased with growing posaconazole average concentration quartiles. The cure rates could achieve 75% when posaconazole average concentration reached at 1.25 mg/L; thereby this quartile value was subsequently accepted as a threshold for IFIs treatment. Further research like a prospective and well-powered study is required to investigate the optimum posaconazole concentration for ensuring safety of posaconazole and efficacy of salvage therapy.
Strengths and weaknesses
Our study has several strengths. This is the first pooled analysis comparing two commonly used but disputed cut-off values for prophylactic efficacy of posaconazole and the results recommended an optimal target for posaconazole usage in IFIs prevention. Besides, we implemented subgroup analysis to seek the suitable targets for patients in different underlying diseases. Our results recommended 0.5 mg/L as a target concentration for IFIs prophylaxis in patients with hematological malignancy, which is more likely to achieve than 0.7 mg/L; hence, it may help to reduce the posaconazole dosage and financial burden for patients and simultaneously ensure the prophylactic efficacy in the long-term usage of posaconazole.
Potential limitations of our study merit discussion. First, we did not investigate the relationship between treatment efficacy, safety and posaconazole exposure owing to the limited number of published studies. Second, survival benefit on each cut-off values have not been explored due to the low mortality and short follow-up time of studies involved. Third, studies concerning the direct comparison of the clinical outcomes of patients taking posaconazole for prophylaxis of IFI with and without TDM are limited, so we are not able to validate the practical benefit of posaconazole TDM in clinical up to date. Further studies are needed in this respect. Finally, the inevitable limitation of all meta-analysis is that the quality of the results are directly related to the quality of individual studies included in the analysis. Except for two randomized controlled trials, all were cohort studies, many of which used a small sample size and focused on a single center. However, we provide the largest pooled analysis of the relationship between posaconazole TDM and clinical efficacy of IFIs prevention. The present study highlights high quality studies in this area is poor and emphasizes the remaining controversy regarding the relationship between posaconazole TDM and treatment efficacy and safety. A well-designed prospective trial to assess the utility of posaconazole TDM, especially in reference to survival, successful response and toxicity, is warranted.