Therapeutic drug monitoring in voriconazole-associated hyponatremia

Abstract

Voriconazole is a second generation triazole antifungal agent and the first choice therapy for invasive aspergillosis (IA). Although voriconazole may be associated with many adverse events, hyponatremia has been rarely reported which potentially could result in death. Therapeutic drug monitoring (TDM) and individualization of therapy by measuring voriconazole plasma concentrations improved the efficacy and safety in patients. We report the effect of TDM to adjust voriconazole dosage in a voriconazole-related hyponatremia patient.

1. Introduction

Invasive pulmonary aspergillosis (IPA) is a necrotizing pneumonia caused by airborne opportunistic fungi of Aspergillus species [1]. Its frequency increases in patients with Chronic Obstructive Pulmonary Disease (COPD), mainly when steroids are prescribed. Voriconazole is a second generation triazole antifungal agent and the first choice therapy for invasive aspergillosis (IA) [2]. Although voriconazole is generally well tolerated, anecdotal case reports have described unexpected severe adverse events related to voriconazole, such as hyponatremia, which potentially could result in death [3–5]. Voriconazole-related hyponatremia is rare and may lead to drug discontinuation, thus confronting clinicians with an antifungal therapy dilemma.

Therapeutic drug monitoring (TDM) comprises the measurement of drug concentrations in biologic fluids to integrate dose, concentration, and pharmacologic effect to understand more precisely the pharmacokinetics and pharmacodynamics of the drug in an individual. General indications for individualization of therapy by means of TDM in specific patient populations include improvement of response to therapy, prevention of drug-related toxicity, and management of drug–drug interactions. TDM has also been suggested for antifungal drug voriconazole, and may contribute to improving therapy in terms of patient outcome, survival, and toxicity [6,7]. In our laboratory, a rapid sensitive and selective HPLC method (water:acetonitrile=65:35 as the mobile phases) has been developed and validated for the routine TDM of voriconazole in plasma. We report a case of successful treatment of IPA in a voriconazole-related hyponatremia patient caused by the syndrome of inappropriate secretion of antidiuretic hormone (SIADH).

2. Case

A 72-year-old man with a 10 years history of COPD was admitted to our respiratory department because of acute exacerbation. Considering the positive aspergillus cultures of sputum before admission (defined as day 0), he had a higher risk of developing IPA. The patient was immediately treated with intravenous voriconazole (two loading doses of 6 mg/kg, then 4 mg/kg every 12 h) until day+14 and then changed to oral voriconazole tablets at a dose of 200 mg every 12 h. The definite diagnosis of IPA was soon obtained from CT guided percutaneous lung biopsy evidencing Aspergillus fumigatus in culture [8]. On day+26 after commencing voriconazole therapy, the patient showed somnolence and malaise symptoms. Electrolyte levels showed that his sodium level was 104 mmol/L, but that his potassium and creatinine levels were normal. The blood osmolality was 263 mOsm/kg and the patient's urinary sodium was 119 mmol/L. Laboratory data are shown in Table 1. TDM was performed and the voriconazole plasma trough concentration (voriconazole-C₀) was high (7.10 μg/mL). Two days after the discontinuation of voriconazole and infusion of 3% saline, the patient's mental status and hyponatremia improved. Our goal voriconazole-C₀ range was 1.0–5.5 μg/mL. The voriconazole-C₀ (0.68 μg/mL), obtained 11 days after a half dose reduction of voriconazole (200 mg/day) on day+39, did not reach the therapeutic range. Since his voriconazole-C₀ remained subtherapeutic (0.68 μg/mL), we increased the dose to 300 mg/day after discharge from hospital. On day+52 after the treatment, voriconazole-C₀ increased to 1.38 μg/mL, which mostly achieved the target concentration of ≥1.0 μg/mL. The patient remained asymptomatic, and repeat CT findings showed near resolution of lung lesions on following up in our outpatient department (Fig. 1). The CYP2C19 genotype was classified as heterozygous extensive metabolizer (CYP2C19⁎1/⁎2).

Table 1.

Laboratory data with voriconazole-associated severe hyponatremia.

Full blood cell count		Urinalysis
White blood cell	6.9×109/L	pH	7.0
Lymphocyte	0.5×109/L	Protein	(+/−)
Eosinophil absolute	0.2×109/L	Occult blood	(−)
Red blood cell	3.73×1012/L	Glucose	(−)
Blood chemistry		Urine chemistry
Urea nitrogen	3.7 mmol/L	Sodium	119 mmol/L
Creatinine	49 µmol/L	Potassium	7.55 mmol/L
Uric acid	68 µmol/L	Osmolality	511 mOsm/kg
Sodium	104 mmol/L	Endocrinological results
Potassium	3.86 µmol/L	Adrenocorticotropic hormone (0 am)	56 ng/L
Calcium	1.83 mmol/L	Adrenocorticotropic hormone (8 am)	69 ng/L
Chloride	89 mmol/L	Adrenocorticotropic hormone (4 pm)	31 ng/L
Albumin	26.8 g/L	Cortisol (0 am)	93.55 nmol/L
Alanine aminotransferase	42 U/L	Cortisol (8 am)	104 nmol/L
Aspartate aminotransferase	117 U/L	Cortisol (4 pm)	126 nmol/L
Alkaline phosphatase	559 U/L
Glucose	6.5 µmol/L
Osmolality	263 mOsm/kg

Fig. 1.

Day+52 after voriconazole treatment the lung lesions are no longer visible, and lung CT shows only minor abnormalities compared to before (left: before treatment, right: after treatment).

3. Discussion

Approximately 1–2% of overall fatal cases of IPA occur in COPD patients. Antifungal monotherapy using azole voriconazole is recommended as a first-line treatment of IPA [2]. Voriconazole may be associated with many adverse events, including visual disturbance, encephalopathy, rash, hepatic enzyme elevation, and other adverse effects, which may result in drug discontinuation [9]. Among these adverse events, hyponatremia has been rarely reported [3–5] (Table 2).

Table 2.

Characteristics of related literatures with voriconazole-associated hyponatremia.

Study	Sodium level (mmol/L)	Time to hyponatremia (days)	Voriconazole trough level (μg/mL)	Mechanism of hyponatremia	CYP2C19 polymorphism
Isobe et al. [3]	122	15	18	SIADH	CYP2C19⁎1/⁎2
Teranishi et al. [4]	113	16	/	SLN	/
Kim et al. [5]	114	10	9.94	SIADH	CYP2C19⁎1/⁎2
	116	7	6.12	SIADH	CYP2C19⁎1/⁎1
	118	6	4.14	SIADH	CYP2C19⁎1/⁎2
This case	104	26	7.10	SIADH	CYP2C19⁎1/⁎2

In the present case, hyponatremia was suggested to be a result of the syndrome of inappropriate secretion of antidiuretic hormone (SIADH) rather than adrenal insufficiency. Adrenal insufficiency might not be suspected as a cause of hyponatremia in our patient, because a maintenance dose of steroid was administered. Therefore, the level of cortisol was slightly low. Moreover, other laboratory data, including lower lymphocyte, eosinophil absolute and calcium levels, and higher blood glucose along with normal urea nitrogen and potassium, could not explain adrenal insufficiency in our patient. Voriconazole-induced hyponatremia due to salt-losing nephropathy (SLN) has been reported [4], such as cisplatin. However, SLN could not be definitively established because the patient's hyponatremia as a result of compromised re-absorption of sodium at the proximal tubules was not confirmed. Our clinical and laboratory findings suggested that voriconazole-associated severe hyponatremia may be associated with SIADH, the same with previous report [3], rather than SLN (see Table 2). Decreased plasma osmolality, albumin, sodium and chloride levels, increased urinary sodium as well as normal creatinine, urea nitrogen, uric acid levels and thyroid function are in accordance with the criteria for SIADH. The precise mechanism of hyponatremia in this case is unknown for the cause of SIADH itself has not yet been fully elucidated.

There is a remarkable interindividual variability of voriconazole blood concentrations, mainly owing to non-linear saturable pharmacokinetics, drug–drug interactions, liver disease, patient age, or CYP2C19 genetic polymorphism [10]. This variability makes it particularly difficult to predict exposure to voriconazole and its potential dose-dependent toxicity. Indeed, voriconazole-C₀≥1.0 μg/mL have been associated with improved response to therapy and survival [11,12]. Increased adverse events have been associated with voriconazole-C₀>5.0–6.0 μg/mL [6,13]. As a consequence, TDM may be a useful tool to optimize voriconazole therapy. In our case, voriconazole-C₀ was 7.10 μg/mL, which is considered to be in the toxic range. Therefore, voriconazole-associated hyponatremia may be concentration-dependent. Two days after the discontinuation of voriconazole, this symptomatic progressive hyponatremia got resolved immediately. Instead of discontinuing antifungal therapy, it was decided to reduce the voriconazole dose to 200 mg/day and voriconazole-C₀ was subtherapeutic (0.68 μg/mL). Finally, TDM revealed an adequate voriconazole-C₀₀ (1.38 μg/mL) 13 days after dose adjustment to 300 mg/day, suggesting that the dose regimen for this patient was appropriate. So, voriconazole-related hyponatremia suggests the clinical utility of routine TDM of voriconazole reduces drug-related adverse events and improves treatment outcome in invasive fungal infections.

In vitro data show that the hepatic metabolism of voriconazole is mediated primarily by CYP2C19, and to a lesser extent by CYP2C9 and CYP3A4 [14]. It is well known that CYP2C19 exhibits genetic polymorphisms and influences the blood concentration of voriconazole [15]. The frequency of CYP2C19 genetic polymorphism possesses considerable interethnic difference: poor metabolizers of CYP2C19 alleles are more common in Asians (15–20%) than the Caucasian populations (3–5%) [16]. If voriconazole-associated hyponatremia is concentration-dependent, this adverse event may be more common in Asians. In our case, an analysis of genetic polymorphism showed a mutation of cytochrome P450 (CYP2C19⁎1/⁎2). Furthermore, the patients did not receive medications known to interfere with voriconazole metabolism. These findings suggest that voriconazole-associated hyponatremia may be related to CYP2C19 genetic polymorphism. There are similar cases where CYP2C19 genotypes were classified as a heterozygous extensive metabolizer [3,5]. However, there was no homozygous poor metabolizer among all patients.

In conclusion, this case suggests that fatally severe hyponatremia can develop after initiating voriconazole antifungal therapy. Furthermore, SIADH and genetic polymorphism of cytochrome P450 may be considered as the cause of voriconazole-related hyponatremia. In addition, this experience confirms the appropriateness of voriconazole dose reduction considered according to voriconazole-C₀ instead of therapy interruption . We believe that TDM is useful to determine voriconazole dosage in a voriconazole-related hyponatremia patient.

Conflict of interest statement

The authors declare no conflict of interest.

Appendix A. Supplementary materials

Supplementary Material