Multicentre study of posaconazole delayed-release tablet serum level and association with hepatotoxicity and QTc prolongation

Abstract

Objectives: The association of posaconazole serum concentrations and toxicity is unclear. An assessment of whether levels obtained with the delayed-release tablet (DRT) formulation are correlated with abnormal liver function test (LFT) results and/or QTc prolongation was undertaken.

Methods: This was a multicentre, retrospective, observational study of adult patients with cancer between 26 November 2013 and 14 November 2014. Patients were included if they received posaconazole DRT with a posaconazole level obtained between days 5 and 14. Clinical data, including demographics, hepatotoxic medications, posaconazole levels, LFTs and QTc intervals, were obtained. Association of factors with changes in LFTs and QTc prolongation was assessed using linear and logistic regression.

Results: One hundred and sixty-six study patients were included. The median posaconazole level was 1250 (range 110–4220) ng/mL and the median time until level was 6 (range 5–14) days. There was a statistically significant increase in AST (P < 0.001), ALT (P < 0.001), alkaline phosphatase (ALK) (P < 0.001), total bilirubin (TBILI) (P < 0.001) and QTc (P = 0.05) from baseline. Posaconazole levels were not associated with increases in AST [β (SE) = −0.33 (2.2), P = 0.88], log ALT [β (SE) = −0.02 (0.03), P = 0.63], ALK [β (SE) = 2.2 (2.9), P = 0.46] and TBILI [β (SE) = −0.01 (0.04), P = 0.88]. For each additional hepatotoxic medication, there was a mean change in TBILI of 0.13 mg/dL (P = 0.02) and ALK of 7.1 U/L (P = 0.09). No statistically significant association between posaconazole level and QTc interval prolongation was found.

Conclusions: We did not identify an association between posaconazole serum concentrations and LFT elevations or QTc prolongation. However, some LFTs were found to increase with more hepatotoxic medications administered.

Introduction

Posaconazole is a broad-spectrum azole approved by the FDA for the prevention of invasive fungal infections in HSCT recipients with graft-versus-host disease (GVHD) or those with haematological malignancies with prolonged neutropenia.¹ Unlike posaconazole suspension, absorption of posaconazole delayed-release tablet (DRT) is not reliant upon intake with high-fat meals or an acidic environment.^2–4 The DRT formulation is now the preferred formulation in clinical practice for adults, given its improved absorption, which results in optimized drug exposure.

Posaconazole therapeutic drug monitoring (TDM) has been widely implemented to assess therapeutic efficacy.^5,6 However, the correlation between serum concentration and toxicity is not clear. The primary aim of this study was to assess whether serum concentrations of posaconazole DRT were associated with hepatotoxicity, as measured by changes in liver function test (LFT) values. The secondary aim was to assess whether serum concentrations were associated with the risk of developing QTc prolongation subsequent to posaconazole DRT exposure.

Methods

This was a multicentre, retrospective, observational study that enrolled haematological cancer patients admitted to four academic medical centres between 26 November 2013 and 14 November 2014. These medical centres were: Mayo Clinic, Rochester, MN, USA; Memorial Sloan Kettering Cancer Center, New York, NY, USA; University of Chicago, Chicago, IL, USA; and University of Michigan Health System, Ann Arbor, MI, USA.

Patients were included in the analysis if they met the following criteria: age ≥18 years; receipt of posaconazole DRT (300 mg every 12 h for two doses, then 300 mg daily) for prophylaxis or treatment in the inpatient or outpatient setting; and serum posaconazole level obtained between 5 and 14 days from posaconazole DRT start date. Patients were excluded if they received intravenous or oral suspension within 5 days prior to starting the tablet. Patients without baseline LFTs measured within 1 week prior to or without follow-up LFTs between 5 and 14 days after posaconazole DRT initiation were also excluded.

Data were collected on: patient demographics; primary malignancy; concomitant hepatotoxic medications (including any medication with an associated risk of hepatotoxicity) up to 7 days prior to starting posaconazole DRT; AST, ALT, alkaline phosphatase (ALK) and total bilirubin (TBILI) at baseline and up to 14 days following initiation of posaconazole; QTc interval at baseline and up to 14 days following initiation of posaconazole; date of posaconazole initiation; date of discontinuation; posaconazole dose; date and time of first posaconazole level; and reported posaconazole concentration. Levels were categorized as troughs when they were measured >20 h after the dose administration, peaks if obtained within 1–4 h after the dose administration and random if measured within 4–20 h after the dose.

For the purposes of this study, baseline hepatotoxicity was defined as initial LFTs 2 × the upper limit of normal. LFTs were evaluated at baseline (within 1 week prior to initiation of posaconazole DRT) with repeat measurements up to 14 days from the time of initiation. QTc prolongation was defined as an increase in QTc interval to >500 ms while on posaconazole therapy if initial QTc interval was <500 ms. Patients were only included in this secondary analysis if a baseline QTc interval and follow-up QTc interval were available.

Ethics

Institutional review boards (IRBs) at each medical centre reviewed and approved the study. IRB approvals for each centre included a waiver of consent for retrospective review of medical charts.

Statistical analysis

The change in LFT measures (AST, ALT, ALK and TBILI) from baseline was assessed using the Wilcoxon signed rank test. Associations between LFT change and serum posaconazole level, number of hepatotoxic medications and underlying hepatic dysfunction were assessed using linear regression models. The β coefficient is interpreted as the change in LFT per unit increase in a continuous variable or the relative change in LFT of one stratum compared with a reference stratum for a categorical variable; β > 0 indicates a mean increase, while β < 0 indicates a mean decrease in LFT. We checked the distribution of all LFT outcomes and a natural log transformation was applied to ALT change values to normalize the skewed distribution. A P value <0.05 was considered significant.

For the association of serum posaconazole with QTc prolongation, QTc was analysed as a continuous variable using linear regression and as elevated repeat QTc (≥500 versus <500) using logistic regression. All statistical analyses were performed using SAS 9.4 (SAS Institute, Inc., Cary, NC, USA) and R version 3.2 (www.R-project.org).

Results

There were 440 patients identified within the study period. Patients were excluded if: there was no serum level between days 5 and 14 (n = 152); they received another posaconazole formulation immediately prior to posaconazole DRT (n = 43); they did not receive a loading dose or received alternative dosing (n = 69); or they did not have baseline or repeat LFTs (n = 10).

Median (range) age was 59.5 (18–89) years, median (range) BMI was 25.9 (16.6–66.7) kg/m², 104 (62.6%) were male, 75 (45%) were seen as inpatients and 91 (55%) were seen as outpatients. There was a diagnosis of haematological malignancy in 161 (97%) patients. Forty-four patients had a history of bone marrow transplant, of which 41 underwent allogeneic transplant and 3 underwent autologous transplant. Baseline hepatic dysfunction was noted in 53 (32%) patients and 19 patients had a history of GVHD.

Indications for posaconazole DRT were prophylaxis in 142 patients (85.5%) and empirical or definitive treatment in 24 (14.5%). Median number (range) of concomitant hepatotoxic medications was 2 (0–6) in 163 patients for whom records were available.

Of the 166 levels, 111 were categorized as troughs, 3 were categorized as peaks and 52 were categorized as random levels. The median posaconazole level was 1250 (range 110–4220) ng/mL. Serum levels were obtained between days 5 and 14 (median 6 days) after initiation of posaconazole DRT. A statistically significant increase in LFT levels from baseline was observed. Initial and repeat median results were as follows: AST 23 (range 6–193) and 34 (range 7–294) U/L (P < 0.001); ALT 25 (range 6–300) and 43 (range 7–2057) U/L (P < 0.001); ALK 82 (range 21–596) and 106 (range 24–804) U/L (P < 0.001); and TBILI 0.6 (range 0.1–8.1) and 0.9 (range 0.1–7.4) mg/dL (P < 0.001). Posaconazole levels were not associated with increases in AST, ALT, ALK or TBILI (Figure 1). Receipt of more hepatotoxic medications was significantly associated with increasing TBILI and marginally associated with increasing ALK from baseline (Table 1). For an increase of one additional hepatotoxic medication, there was a mean change in TBILI of 0.13 mg/dL (P = 0.02) and in ALK of 7.1 U/L (P = 0.09).

Figure 1.

Increasing levels of serum posaconazole were not significantly associated with changes in AST, ALT, ALK or TBILI from baseline.

Table 1.

Increase in risk of hepatotoxicity for each additional concomitant hepatotoxic medication (n = 163)

Number of hepatotoxic medications	ALK (U/L)				TBILI (mg/dL)
	baseline mean (SD)	repeat mean (SD)	β (SE)	P	baseline mean (SD)	repeat mean (SD)	β (SE)	P
0	92 (55)	99 (51)	reference	0.09	0.7 (0.3)	0.7 (0.3)	reference	0.01
1	118 (97)	149 (98)	24.3 (12.5)		0.9 (0.8)	1.3 (1.2)	0.4 (0.2)
≥2	98 (74)	130 (102)	24.4 (12.0)		0.8 (1.0)	1.3 (0.9)	0.5 (0.2)

Univariate linear regression analysis of change in LFTs from baseline to post-therapy.

β is the regression coefficient for that stratum relative to the reference stratum (no medications): β > 0 indicates a mean increase in LFT and β < 0 indicates a mean decrease in LFT.

QTc was available for 149 patients at baseline and 119 post-therapy The median interval at baseline was 444 ms (range 383–529) and median interval after posaconazole DRT exposure was 449 ms (range 377–583), P = 0.05. Baseline and repeat QTc data were available for 118 of 166 (71%) patients, excluding 1 patient with baseline QTc >500 ms. There was no statistically significant relationship between posaconazole level and change from baseline QTc [per 500 ng/mL β (SE) = −0.35, P = 0.87]. There was no increased likelihood of elevated QTc ≥500 ms after posaconazole DRT exposure (OR 0.96, 95% CI 0.62–1.47, P = 0.85).

Discussion

Variable pharmacokinetics, LFT abnormalities and QTc prolongation are significant limitations of azole antifungals.^7–9 This study did not find any association between posaconazole levels and elevations in LFTs or QTc intervals. In addition, clinical relevance of observed changes in AST, ALT, ALK and TBILI was limited. The occurrence of elevated TBILI and ALK was more common as additional hepatotoxic medications were administered. Patients receiving posaconazole DRT have an underlying haematological malignancy and may receive other therapy associated with hepatotoxicity. Therefore, clinicians should be aware of the risk additional hepatotoxic medications are posing.

There are several limitations of this study worth noting. As this was a multicentre study, there was site-to-site variability in the concomitant chemotherapy regimens used in the treatment of haematological malignancies. Therefore, the agents were categorized according to their relative risk of hepatotoxicity. To account for variances in practices between centres for ordering posaconazole levels, LFTs and QTc monitoring, we set stringent exclusion criteria. Patients who receive posaconazole for salvage therapy may have more risk factors for the development of hepatotoxicity. However, only 24 (15%) of 166 received posaconazole for empirical or definitive treatment, so the number is too small to draw conclusions. Similarly, patients with differing cancers may have varying risk factors for developing hepatotoxicity. Additionally, we did not follow LFTs serially over an extended period of time and may not have captured longer-term toxicities.

The use of antifungal TDM has been increasing rapidly with the advent of the expanding azole antifungal arsenal.¹⁰ In order to justify the cost of TDM, a benefit of improved efficacy and/or reduced toxicity needs to be demonstrated. This study did not find an association of serum levels and elevations in LFTs or QTc prolongation among patients receiving posaconazole DRT. Our findings suggest that routine serum level monitoring to assess for posaconazole toxicity may not be necessary with the DRT formulation.