Therapeutic Drug Monitoring of Posaconazole in Hematology Patients: Experience with a New High-Performance Liquid Chromatography-Based Method

Parallel administration of the proton pump inhibitor (PPI) esomeprazole has been shown to decrease oral bioavailability of posaconazole in healthy volunteers. We prospectively analyzed serum samples (n = 59) obtained from hematology patients (n = 27) under posaconazole prophylaxis. Patients treated concomitantly with pantoprazole had significantly lower posaconazole levels than patients without PPI treatment (median levels of 630 μg/liter versus 1,125 μg/liter, respectively). These results suggest that drug monitoring is relevant when posaconazole and pantoprazole are administered concomitantly.Posaconazole is a broad-spectrum antifungal triazole approved for prophylaxis and treatment of invasive mycoses in hematology patients (11, 12, 23). The drug is available as an oral suspension, and serum level measurements have shown relatively high intra- and interindividual differences in both healthy volunteers (HV) and patient populations (2, 6, 18, 25). The absorption of posaconazole can be increased 2.6-fold by administering it along with food or oral nutritional supplements (3, 13, 22). Gastric pH has also been shown to influence the bioavailability of posaconazole. For instance, parallel administration of posaconazole with cimetidine and esomeprazole has been demonstrated to decrease posaconazole exposure in HV by 40% and 32%, respectively (4, 15). However, similar effects have not been assessed in hematology patients, which would be of great importance to optimize posaconazole administration in these patients. Oral bioavailability of posaconazole has been demonstrated to be especially low in patients after hematopoietic stem cell transplantation (SCT) (10). Possible reasons are high incidences of mucositis and graft-versus-host disease and drug interactions (14, 16, 26). Data in correlation between posaconazole exposure and efficacy are limited, and prospective studies to define target concentrations are missing. Two clinical trials demonstrated that elevated posaconazole levels were associated with improved responses (26, 28). One of these trials demonstrated that mean posaconazole levels in patients who did or did not develop breakthrough infections were 611 μg/liter or 922 μg/liter, respectively (26). Based on these results, the FDA briefing document recommends target posaconazole concentrations of >700 μg/liter (7).Therapeutic drug monitoring (TDM) of posaconazole was carried out with a previously published method based on high-performance liquid chromatography (19). The method''s lower limit of quantification and limit of detection were 0.1 mg/liter and 0.05 mg/liter, respectively. A linear calibration curve from 0.1 to 5 mg/liter was obtained using a 50-μl sample. Relative standard deviations of intraday variations ranged from 2.3% to 9.4%, and intraday accuracy ranged from 88.8% to 114.8%.We prospectively analyzed serum posaconazole levels in samples obtained from consecutive inpatients under posaconazole prophylaxis at our department between November 2008 and July 2009. Samples were drawn before administration of posaconazole as morning trough values. At the time of analysis, all patients had received posaconazole for at least 5 days at a dosage of 200 mg three times daily. Serum posaconazole levels were monitored on a weekly basis. When more than one serum sample obtained from a patient was analyzed, mean posaconazole levels were used for statistics. Of those patients with parallel administration of pantoprazole, four did not receive the proton pump inhibitor (PPI) during the entire period of posaconazole prophylaxis. Depending on whether more serum samples were obtained with or without concomitant administration of pantoprazole, these patient samples were classified within the cohort with the majority of samples.Low posaconazole concentrations were defined as <700 μg/liter (7). Hepatic function impairment was defined as increased alanine transaminase (ALT) levels to higher common toxicity criteria (CTC) grades, according to the National Cancer Institute (NCI) CTC guidelines. Breakthrough infections were defined as the occurrence of proven, probable, or possible fungal infections, according to the criteria of the EORTC/MSG (5). Summary statistics were used to describe patient characteristics and measured posaconazole levels. Box plots were employed to display differences between the analyzed cohorts. Statistical significance was tested with the nonparametric Wilcoxon rank sum test. A P value of <0.05 was considered statistically significant. Statistical analyses were performed by using the R software package (21). The study was approved by the ethics committee of the University of Freiburg.Serum posaconazole levels in 27 hematology patients were analyzed (characteristics summarized in Table ​Table1).1). Fifty-nine measurements were performed, with a median of two samples per patient. Serum posaconazole levels ranged from 130 to 2,300 μg/liter (median, 740 μg/liter) (Fig. ​(Fig.11 A). Two serum levels were below the method''s limit of quantification (100 μg/liter) and were treated as 0 μg/liter for the statistical analysis. Posaconazole levels of <700 μg/liter were observed in 13/27 patients (48%). The effects of age, gender, underlying disease, and number of concomitant drugs on serum posaconazole levels were not significant (Fig. 1B to E).Open in a separate windowFIG. 1.Measured posaconazole levels in all patients (n = 27) (A), patients younger (n = 13) or older (n = 14) than 58 years (B), female (n = 15) versus male (n = 12) patients (C), patients with acute leukemia (n = 22) versus other malignant diseases (n = 5) (D), and patients with eight or less (n = 13) versus more than eight (n = 14) concomitantly applied drugs (co-drugs) (E). The boxes include the interval between the 25% and 75% quantiles, and the extremes represent the 5% and 95% quantiles. Measured values outside the latter range are displayed as small circles, and the median is shown as a horizontal line in the middle of the box.

TABLE 1.

Patient characteristics

Characteristica	Value
No. of patients	27
Sex
No. of males (%)	12 (44)
No. of females (%)	15 (56)
Median yr of age (range)	58 (28-75)
Underlying disease
No. of patients with leukemia (%)	25 (93)
No. of patients with AML	18
No. of patients with ALL	4
No. of patients with CML	2
No. of patients with MDS	1
No. of patients with lymphoma (%)	2 (7)
% median Karnofsky index (range)	80 (60-90)
SCT
No. of patients without SCT (%)	20 (74)
No. of patients with allogeneic SCT (%)	7 (26)
Median no. of coadministered drugs (range)	8.5 (5-13)

Open in a separate window^aAML, acute myeloid leukemia; ALL, acute lymphoblastic leukemia; CML, chronic myeloid leukemia; MDS, myelodysplastic syndrome.A total of 17 of 27 patients (63%) received pantoprazole (median dosage, 40 mg per day; range, 20 to 80 mg). Low serum posaconazole levels were seen more frequently in patients who received pantoprazole (12/17 patients [71%]) than in those who did not (1/10 patients [10%]). Moreover, there was a significant difference in median posaconazole levels between the two cohorts (630 μg/liter versus 1,125 μg/liter; P = 0.005) (Fig. ​(Fig.22 A).Open in a separate windowFIG. 2.Measured posaconazole levels in patients without (n = 10) versus with (n = 17) parallel administration of pantoprazole (A), patients without (n = 20) versus with (n = 7) SCT (B), and patients with (n = 13) versus without (n = 14) elevation of ALT levels during posaconazole prophylaxis (C). The boxes include the interval between the 25% and 75% quantiles, and the extremes represent the 5% and 95% quantiles. Measured values outside the latter range are displayed as small circles, and the median is shown as a horizontal line in the middle of the box.In the nontransplant cohort, 8/20 patients (40%) showed posaconazole levels of <700 μg/liter, compared to 5/7 patients (71%) in the allogeneic SCT group. However, differences in median serum levels (805 μg/liter versus 640 μg/liter) were not significant (Fig. ​(Fig.2B2B).Elevated ALT levels during posaconazole prophylaxis were observed in 13/27 patients (48%). Low posaconazole levels were equally seen in patients with (6/13 patients [46%]) and without (7/14 patients [50%]) elevated ALT levels. Moreover, median posaconazole levels did not significantly differ (830 μg/liter versus 700 μg/liter) (Fig. ​(Fig.2C2C).Breakthrough fungal infections during posaconazole prophylaxis were observed only in two patients with acute myeloid leukemia (AML), thereby reflecting the drug''s high efficacy in hematology patients. Serum posaconazole levels were 900 μg/liter when the first radiological signs of a fungal infection were detected, and both patients died of progressive respiratory insufficiency within 30 days. Fungal infections could not be confirmed by autopsies and, therefore, only met the EORTC/MSG criteria for possible fungal infections.The available data suggest that TDM of posaconazole in patients with poor oral absorption, with progressive disease under posaconazole, and on concomitant medication with significant drug interactions is clinically beneficial (1, 8, 24). As these prerequisites are commonly found in hematology patients, determining serum posaconazole levels may be a valuable option in this high-risk cohort.PPIs belong to the most frequently applied drugs in hematology patients. We found that the parallel administration of pantoprazole results in significantly decreased serum posaconazole levels, which is in agreement with data from Krishna et al., who reported similar effects for esomeprazole in HV (15).According to previous data (16), mucositis is an important cause of low levels of serum posaconazole. However, the examination of mucositis was not within the scope of our analysis and was therefore not performed.Our data on the efficacy of posaconazole use are limited by the fact that only two possible breakthrough infections were detected, with host factors and clinical features being the only evidence of fungal infections in these patients.In contrast to previously published data (25), our posaconazole levels did not significantly differ between patients with allogeneic SCT and nontransplant patients. These results are most likely due to the fact that nontransplant patients were treated with intensive induction chemotherapy regimens, with mucositis rates presumably as high as those in the transplant group. Furthermore, the numbers of supportive drugs used were comparable in both cohorts, resulting in similar potentials for drug interactions.Nearly half of our patients (13/27) had elevated ALT levels during posaconazole prophylaxis, which is much higher than previously described. However, previous trials applied tight exclusion criteria (2, 10, 25, 28), leading to a preselection of patients who are not necessarily representative of those seen in “everyday clinics” and may therefore show more favorable treatment results. This particularly applies to side effects that emerge more often in patients with comorbid conditions and poor performance statuses. In agreement with data from Walsh et al. (28), we did not observe significant differences between serum posaconazole levels in patients with and without elevated ALT levels.The possible limitations of our study are that only patients from a single institution were included and that the number of analyzed serum samples was limited. However, the determination of serum posaconazole levels in consecutive patients reflects clinical routine much better than previous trials with preselected patients.Our study clearly demonstrates that serum posaconazole levels can be substantially reduced with concomitant pantoprazole use in hematology patients and suggests that TDM of posaconazole and immediate feedback to physicians are important contributions to understanding pharmacokinetics and drug interactions, especially in patients receiving multiple medications.