Retrospective evaluation of voriconazole related adverse events in pediatric patients with hematological malignancies

Objective: Voriconazole is a broad-spectrum antifungal agent that is effective in treatment and prophylaxis of invasive aspergillosis in pediatric patient with hematological malignancies. Plasma concentration of the drug varies considerably by inflammatory status, liver functions, age, drug-drug or drug-food interactions and genetic polymorphisms in the drug metabolizing enzymes. Low plasma concentrations lead to therapeutic failure, while higher concentrations may put patient at increased risk of adverse reactions. Due to individual variations in pharmacokinetics, therapeutic drug monitoring and pharmacogenetic-guided dose selection is recommended in pediatric patients. Nevertheless, these are not widely used in routine clinical practice. Safety of voriconazole is usually monitored by clinical assessment of known adverse reactions and laboratory abnormalities. This study primarily aimed to address all adverse events that appeared during the first 30 days of voriconazole use and to evaluate the causal relation between the drug and the event. The data regarding efficacy of voriconazole was also evaluated as secondary objective. Method: Patient data retrospectively obtained from medical records. Severity of the adverse events described in medical records and/ or laboratory abnormalities were graded by the United States National Cancer Institute- Common Terminology Criteria for Adverse Events (NCI-CTCAE) 5.0. Causal relation between voriconazole use and the adverse events was evaluated and categorized by Liverpool Causality Assessment Tool (LCAT), the Naranjo Algorithm and World Health Organization- Uppsala Drug Monitoring Centre (WHO-UMC) Causality Assessment System. Efficacy was assessed according to European Organization for Cancer Treatment and Research/Mycosis Study Group (EORTC/MSG) Consensus Criteria. Categorical variables were displayed as number (n) and percentage (%). Continuous variables were presented by mean ± standart deviation (SD) or median (minimum-maximum). Depending on normality of their distribution, paired samples t-test or Wilcoxon signed rank test was used to compare the laboratory measurements before and during voriconazole use. The Chi Square statistic was used for testing relationship between categorical variables. SPSS-22 (SPSS INC., Chicago, IL, USA) was used for statistical analysis. P-values of <0.05 were considered statistically significant. Findings: At least one adverse event detected in each one of 45 patients included in the study and 23 (51.1%) patients experienced adverse reactions. Hematologic adverse events were anemia in 23 (51.1%), leukopenia in 12 (26.7%), neutropenia in 10 (22.2%) and thrombocytopenia in 12 (26.7%) patients. No causal relationship was detected between hematologic adverse events and voriconazole use. At least one hepatobiliary adverse event identified in 40 (88.9%) patients. In 22 (48.9%) patient, hepatobiliary events were classified as adverse reactions. Hepatobiliary adverse events were alanine aminotransferase elevation (ALT) in 27 (60.0%), aspartate aminotransferase (AST) elevation in 26 (57.8%) and total bilirubin elevation in 21 (46.7%) patients. Of 37 patients whom alkaline phosphatase (ALP) measurements were available, increased level of ALP detected in 12 (32.4%) patients. Twenty (55.6%) of 36 patients with available gamma-glutamyl transferase (GGT) measurements, experienced GGT elevation during voriconazole use. In 14 patients with ALT elevation, 15 patients with AST elevation, nine patients with ALP elevation, 14 patients with GGT elevation and seven patients with total bilirubin elevation, these events and voriconazole use had a "possible" causal relationship. All adverse reaction that lead termination of voriconazole use were related to hepatobiliary system and the drug discontinued in 9 (20.0%) patients for this reason. Hepatobiliary adverse reactions tended to occur within the first two weeks of voriconazole use and reached to maximum grade of severity on the third week. Pancreatic adverse events were lipase and amylase elevation in five (35.7%) and six (42.9%) of the patient with available pancreatic enzyme measurements. A "possible" adverse reaction was detected in a patient with amylase and lipase elevation. Among 44 patients with available calcium and potassium measurements, 17 (38.6%) cases of hypocalcemia, one (2.3%) with hypercalcemia, 18 (40.9%) with hypokalemia and 15 (34.1%) with hyperkalemia identified as adverse events. Of 33 patients with available magnesium measurement, 12 (36.4%) cases with hypomagnesemia and five (15.2%) cases with hypermagnesemia were adverse events. The causality assessment revealed no relationship between voriconazole use and electrolyte abnormalities and none of them were categorized as an adverse reaction. The other adverse events identified during retrospective review were hypoalbuminemia in 16 (36.6%), increased creatinine level is one (2.2%), erythema multiforme in one (2.2%), maculopapular rash in one (2.2%) and bradycardia in one (2.2%) patient. None of these events were causally related to voriconazole use. On the other hand, a case of ventricular extra systoles and another case of hallucination and nightmares were identified as "possible" adverse reactions. Death attributable to hematological malignancy on the eighth day of voriconazole treatment in one patient and progression of invasive fungal infection in 10.7% of the patients were detected among the ones who were included in efficacy evaluation. No findings that would suggest ineffectiveness were identified in the rest of study population. Conclusion: This study individually addressed the frequency of adverse events and adverse reactions in pediatric patients, systematically graded both seriousness/ severity of the events and evaluated causal relation between those events and voriconazole use. Hepatobiliary abnormalities were the most common adverse reactions that occurred during the first month of voriconazole use and were the most common cause of treatment discontinuation. Visual disturbances and dermatological disorders, the commonly reported adverse reactions to voriconazole, could not be identified in our study. As retrospective evaluation entirely relies on the data in medical records, this may have negatively affected our ability to identify the unrecorded adverse reactions that can only be detected by clinical observation. Although the duration of follow up period of the study is inadequate to clearly determine the efficacy, 85.7% of the patients manifested radiological or clinical signs of satisfactory response. To clearly identify the accurate frequency and the causality of all adverse reactions, prospective studies with much larger sample size and with real time plasma concentration measurement of voriconazole are needed.