Caspofungin as antifungal prophylaxis in pediatric patients undergoing allogeneic hematopoietic stem cell transplantation: a retrospective analysis

The study is a single centre, retrospective survey on antimycotic prophylaxis in pediatric patients under eighteen years of age who underwent allogeneic HSCT between January 2006 and June 2010 at the University Children’s Hospital Tübingen, Germany. The allogeneic stem cell transplantation occurred in high-efficiency particle-arrestance filtered environment. All pediatric patients in our clinic received L-AmB exclusively as antifungal prophylaxis during conditioning and after HSCT until August 2008. As we observed a high incidence of nephrotoxicity, we changed our antifungal prophylaxis beginning from day 1 after allogeneic HSCT, from L-AmB to CAS in September 2008. Other supportive standards remained unchanged during this short period of 4.5 years.

The primary objective of this study was to evaluate safety of CAS as antifungal prophylaxis in pediatric patients following allogeneic HSCT as compared to L-AmB. The secondary objective was the incidence of aspergillosis, candidiasis, and other mycoses.

The observation period in this trial was defined as the time from start of intravenous antimycotic monoprophylaxis at the beginning of conditioning until three weeks after switching to oral antimycotic prophylaxis 3–4 days before inpatient discharge.

All 120 pediatric patients included in the survey received antimycotic prophylaxis with L-AmB (1 mg/kg/day) from the beginning of conditioning until day 0 (day of allogeneic HSCT) (Figure 1).

The first group of pediatric patients receiving stem cell transplants between January 2006 and August 2008 also received antimycotic prophylaxis with L-AmB after HSCT beyond that point, with an initial dose of 1 mg/kg/day. In case of fever greater than or equal to 38.5°C after the first day post HSCT, which was interpreted as the first clinical suspicion of a possible fungal infection, further antimycotic prophylaxis with L-AmB monotherapy was administered at 3 mg/kg/day, and was continued at that dosage up until oral antimycotic prophylaxis a few days before the patient’s discharge.

The second group of patients, who had undergone stem cell transplantation between September 2008 and June 2010, received the same dosage as the pediatric patients in group 1 until day 0, that is, they received L-AmB at a dosage of 1 mg/kg/day, and CAS at a dosage of 1 × 50 mg/m²/day intravenously beginning on day 1 after allogeneic HSCT, but not more than 50 mg/day, even in cases of fever. This dosage was based on pharmacokinetic studies, which showed that 50 mg/m²/day in children was comparable to 50 mg/day in adults, although a further increase to 70 mg/m²/day (maximum, 70 mg/day) would have been possible [26, 27]. A loading dose of CAS was not given, since antimycotic prophylaxis had already been taking place since the start of conditioning with L-AmB at a dosage of 1 mg/kg/day. The infusion time of sixty minutes was the same for each drug. L-AmB (Ambisome®) was manufactured by Gilead Sciences (Martinsried, Germany) and CAS (CANCIDAS®) by MSD Sharp & Dohme GmbH (Haar, Germany). In addition to the systemic antifungal prophylaxis, all of the patients involved in the survey received an oral prophylaxis with amphotericin B (AmphoMoronal®) twice daily. An inhalation of antifungals did not take place.

120 pediatric patients under eighteen years of age with hemato-oncological malignancies, non-malignant hematological diseases, and inborn errors of metabolism undergoing allogeneic HSCT were included in this retrospective analysis. Exclusion criteria were the presence of proven or probable invasive fungal infections before the start of conditioning according to the definitions of invasive fungal disease by the European Organization for Research and Treatment of Cancer Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) [28].

Successful intravenous antifungal monoprophylaxis was defined as the absence of proven or probable fungal infection from the post-transplant period through the date of discharge, as well as three weeks after the end of intravenous antifungal monoprophylaxis.

This endpoint was comprised of clinical and laboratory parameters. Clinical side effects were recorded according to current Common Toxicity Criteria by the US National Cancer Institute [29]. Hepatic toxicity was analyzed by assessment of alanine aminotransferase (ALT, normal value ≤ 39 U/L) and aspartate aminotransferase (AST, normal value ≤ 39 U/L), alkaline phosphatase (AP, normal value ≤ 320 U/L), total and direct bilirubin in the patients’ sera. Renal function was assessed by plasma creatinine (normal value ≤ 0.7 mg/dl), urea (normal value ≤ 46 mg/dl) and potassium. Increases of ≥ 1.5 and ≥ 2.5 times the normal values (beginning on day 1 after allogeneic HSCT) were considered significant, or with regard to potassium, a decrease of plasma concentrations ≤ 3.4 mM or ≤ 2.4 mM. Furthermore, we appraised the need for oral substitution of potassium, bicarbonate, and calcium at discharge. We also assessed if cessation of either prophylaxis was necessary due to toxicity.

All 120 pediatric patients who received monoprophylaxis with L-AmB or CAS were included in the assessments of safety and efficacy. The statistical comparison of the difference between the results and normal values was done with the one-sample t-test. The Wilcoxon matched pairs signed rank test was applied for a statistical comparison of these parameters: between the baseline before conditioning and day 0, the baseline and maximum/minimum during intravenous antifungal prophylaxis, the baseline and end of intravenous antifungal prophylaxis, day 0 and maximum/minimum, day 0 and end, and maximum and end. The hepatic and renal function data are presented as mean values ± standard deviation. P values of ≤ 0.05 (*), P ≤ 0.01 (**) and P ≤ 0.001 (***) were defined as statistically significant. The Two Independent Proportion Z-Test was used to compare percentage of patients with hypokalemia during intravenous antimycotic prophylaxis as well as the percentage of electrolytes substitution between two groups. Graphs were created with GraphPad Prism 4 for Windows, version 4.03 (GraphPad Software, San Diego, California, USA). The statistical analysis was carried out with the statistical program XLStat2010 (AddinSoft, Paris, France).

In total, 120 pediatric patients under the age of eighteen years, divided into two groups, were evaluated in this analysis. All pediatric patients included in the survey, a total of 60 patients in the first group and an additional 60 patients in the second group, received antimycotic prophylaxis with L-AmB (1 mg/kg/day) from the beginning of conditioning until day 0. The 60 pediatric patients in group 1 underwent allogeneic HSCT between January 2006 and August 2008, and received L-AmB at a dosage of 1 mg/kg/day as antifungal prophylaxis from day one after allogeneic HSCT. In 34 (56.7%) of the 60 pediatric patients in the L-AmB group, the antimycotic prophylaxis with L-AmB was adjusted after HSCT to 3 mg/kg/day from the initial dosage of 1 mg/kg/day due to fever greater than or equal to 38.5°C. The second group also consisted of 60 pediatric patients, all of whom were treated intravenously with CAS at a dosage of 1 × 50 mg/m²/day beginning on day one after allogeneic HSCT, but no more than 50 mg/day, between September 2008 and June 2010. Fever greater than or equal to 38.5°C came about in 27 (43%) of the 60 pediatric patients treated with CAS.

The median age in the L-AmB group was 7.5 years (range 4 months to 17.6 years), while the CAS group had a median age of 9.5 years (range 8 months to 17.5 years) (Table 1). Males were slightly over-represented (L-AmB: n = 38, CAS: n = 33) in both groups. The most common primary diagnoses in both groups were acute lymphoblastic leukemia (ALL) (n = 18, n = 17) and solid tumors including Ewing sarcoma, neuroblastoma, and rhabdomyosarcoma (n = 9, n = 10). 28.3% (n = 17) of patients in the L-AmB group and 25% (n = 15) of patients in the CAS group received a myeloablative conditioning regimen with total body irradiation.

Leukocytopenia was observed for a median duration of 12 days in both groups. 56.7% (n = 34) of the pediatric patients in the L-AmB group and 21.7% (n = 13) in the CAS group received methylprednisolone during conditioning. 61.7% (n = 37) of patients in the CAS group were treated with prednisolone after transplantation in comparison to 38.3% (n = 23) in the L-AmB group. The median duration of intravenous antifungal monoprophylaxis was approximately the same for each group, 23.0 days (range 9–72 days) in the L-AmB group and 24.0 days (range 14–49 days) in the CAS group after allogeneic HSCT.

Prophylaxis was effective with L-AmB as well as with CAS. There was no incidence of proven invasive aspergillosis or another invasive fungal infection in either group. This also remained true during the conditioning phase, when both groups received liposomal amphothericin B, and also during the post-transplant period, in which group one received L-AmB and group two CAS. Furthermore, no proven fungal breakthrough infections were observable in either group three weeks after the conclusion of intravenous antifungal prophylaxis.

In the CAS group, one severely immunocompromised pediatric patient showed clinical and serological indicators of a probable invasive fungal infection, i. e. increase of galactomannan antigen in more than two consecutive blood samples and signs of pulmonary aspergillosis in computed tomography thirteen days after start of monoprophylaxis with CAS. However, bronchopulmonary lavage failed to detect fungi by culture or PCR. The patient died on day 94 after allogeneic HSCT of veno-occlusive disease (VOD), the parents did not consent with an obduction. There was no incidence of probable invasive fungal infections in the L-AmB group.

None of the patients in both cohorts died from an invasive fungal infection during the observation period. In addition to the single pediatric patient from the CAS group who succumbed to VOD, one patient from the L-AmB group died from a relapse of AML three weeks after the end of the intravenous antifungal prophylaxis.

Clinical side effects directly related to intravenous treatment with L-AmB were observed in 5 (8.3%) and directly related to CAS in 2 (3.3%) pediatric patients (Table 2). The clinical side effects in der L-AmB group occurred equally as frequently in the pediatric patients receiving L-AmB at a dosage of 1 mg/kg/day (n = 2) as in those receiving 3 mg/kg/day (n = 3).

In the L-AmB group this included one incidence each of fever, nausea, and bone pain, as well as two cases of headache. L-AmB prophylaxis was discontinued in the latter four of these cases and switched to CAS. In the CAS group, fever occurred in one case and headache in another. We did not stop the administration of a drug in either of these two cases. Laboratory parameters of hepatic function showed a statistically significant, yet transient increase in ALT and AST in both groups, beyond the upper normal limit during treatment post-transplant in relation to baseline (P < 0.001) as well as to day 0 (P < 0.001) (Figure 2). An increase ≥ 1.5 times the normal value of 39 U/L occurred almost with the same frequency in both groups for ALT (L-AmB: n = 13, CAS: n = 14) and more frequently in the CAS group for AST (L-AmB: n = 15, CAS: n = 19) (Table 2). ALT values ≥ 2.5 times the normal value of 39 U/L were observed in 3 patients under L-AmB and in 8 patients under CAS. 6 patients in the L-AmB group and 2 patients in the CAS group showed AST levels at least 2.5fold above the normal limit. Total bilirubin and direct bilirubin also increased insignificantly during post-transplant treatment. AP did not increase during post-transplant period. In both groups, renal function parameters showed an increase in creatinine and urea during treatment with L-AmB, but not significantly beyond the upper normal limit. Hypokalemia ≤ 3.4 mmol/L occurred significantly (P = 0.006) more often in the L-AmB group (48 patients, 80%) than under CAS (34 patients, 57%) (Figure 3).

A total of 25% (n = 15) of pediatric patients in the L-AmB group required oral potassium supplementation and spironolactone upon discharge. This compares to only 11.7% (n = 7) in the CAS group. Sodium bicarbonate substitution was required in 5 (8.33%) and calcium in 3 (5%) cases upon discharge in the L-AmB group. In the CAS group, calcium was given in 2 (3.3%) cases and sodium bicarbonate in one (1.7%) case. In the L-AmB group, there was a trend towards higher percentage of patients requiring potassium (P = 0.0591) and bicarbonate substitution (P = 0.094) after end of intravenous antifungal prophylaxis. All pediatric patients in the L-AmB group who required oral supplementation with potassium and bicarbonate had received L-AmB after HSCT at a dosage of 3 mg/kg/day.