Title: Pediatric Acute Pancreatitis Related to Tigecycline
Abstract: To the Editors: Tigecycline is the first member of the glycylcycline class of antibiotics and has been recently approved for complicated skin, soft-tissue, and intra-abdominal infections in adults. According to official guidelines, tigecycline is not recommended for children less than 18 years of age in France because of the lack of safety and effectiveness evidence in pediatric patients. However, when no alternative treatment for severe infections is available, tigecycline may be used considering the risk-benefit ratio. In this study, we report the first case of tigecycline-induced pancreatitis in children. A 9-year-old boy, with sickle-cell anemia, was admitted in October 2008 for left leg vaso-occlusive crisis requiring intravenous hydration and morphine. The patient was treated with hydroxyurea for 3 years for a severe form of sickle-cell anemia and had undergone cholecystectomy 1 year ago. After 5 days of hospitalization, he developed a septic syndrome, with very high fever (40°C), shivers, and abdominal pain. Enterobactercloacae producing extended spectrum betalactamase was isolated from blood culture (minimum inhibitory concentrations [MICs]: ciprofloxacin = 0.5 mg/L, amikacin = 3 mg/L, imipenem = 0.5 mg/L, meropenem = 0.023 mg/L, tigecycline = 0.75 mg/L). Intravenous antibiotics (imipenem-cilastatin, ciprofloxacin, and amikacin) failed to prevent secondary dissemination of the infection in the 2 femurs, the 2 hips, and the 2 shoulders. During treatment, E.cloacae acquired an imipenem-cilastatin outer membrane permeability defect (MIC = 16 mg/L) and decreased susceptibility to ciprofloxacin (MIC = 1.5 mg/L). Antibiotics were changed to tigecycline (100 mg infusion, then 50 mg twice a day), colistin (1.25 MUI tid, then qid), amikacin (60 mg/kg/d), and rifampin (25 mg/kg/d). The high doses of the antibiotics used were monitored by measurement of serum concentrations except for tigecycline. Two weeks after the beginning of this therapy, gastro-intestinal symptoms, including abdominal pain, recurrent vomiting, and moderate pancreatic reaction, were noted. Transient and slight amylase and lipase increase was observed (Fig., Supplemental Digital Content 1, https://links.lww.com/INF/A525). The patient was given parenteral nutrition. Amylase and lipase values returned to normal within a few days, and no imaging was performed during this episode. After 8 weeks of tigecycline treatment and in the setting of persistent digestive symptoms (anorexia, vomiting, and recurrent abdominal pain), a tigecycline-induced reaction was suspected and tigecycline was stopped for 10 days. No improvement was noted and tigecycline was then restarted. Thirteen days later, the patient suddenly started vomiting and developed stronger epigastric paroxystic abdominal pain which was not relieved by antalgics (WHO level 1 and 2 analgesic ladder). His lipase concentration was about 4.5 times higher than the upper limit of normal range (134 U/L), but the amylase value was normal (Fig., Supplemental Digital Content 1, https://links.lww.com/INF/A525). Acute pancreatitis was confirmed by an abdominal computed tomography scan showing inflammation involving pancreas and peripancreatic fat without necrosis (Ranson Score 2 and Balthazar stage 2). Tigecycline therapy was stopped again and the colistin dose tapered. The serum amylase and lipase values quickly returned to normal in 5 days (Fig., Supplemental Digital Content 1, https://links.lww.com/INF/A525). Abdominal pain and vomiting resolved allowing the patient to receive oral alimentation after 3 days. All pancreatic signs and symptoms resolved without any long-term complications. In this study, we report the first pediatric case of acute pancreatitis associated with tigecycline therapy. All signs and symptoms finally resolved after tigecycline treatment was stopped. No long-term complications occurred. This possible adverse reaction has to be considered for the risk-benefit ratio of tigecycline use in children. All cases reported in the literature described nausea and abdominal pain within 7 to 14 days before the diagnosis of acute pancreatitis.1–4 In our case, the first pancreatic reaction was mild, and started 2 weeks after treatment initiation. Total parenteral nutrition may have had a positive effect on this pancreatic reaction. Abdominal symptoms were however persistent and, 4 weeks after the first pancreatic reaction, tigecycline was stopped without clinical improvement. After 10 days and because of the lack of alternative antibiotic, the patient was rechallenged with tigecycline. In the published data, no patient had been rechallenged with tigecycline or tetracycline. This rechallenge demonstrated the probable association of tigecycline with pancreatitis. For this second episode, abdominal symptoms resolved within 3 days after withdrawal of the drug. In other cases, the time to recovery of symptoms and abnormal pancreatic enzymes ranges from 13 to 21 days.1–4 Lipase and amylase concentrations returned to 173 UI/L and 151 UI/L, respectively, much lower than other reported cases (700–1400 UI/L). Hung et al recommend that "clinicians monitor for symptoms of abdominal pain during treatment with tigecycline and have a low threshold to order amylase and lipase concentrations if the clinical presentation is compatible with acute pancreatitis."2 As a result of this case, in the presence of clinical symptoms compatible with pancreatitis in a tigecycline-treated child, we also recommend rapidly performing CT imaging despite mild elevated amylase and lipase. Other tetracyclines may be contra-indicated for such patients considering the possible cross-reactivity. In the pediatric context, more precautions are needed, because children's sensibility to tigecycline-induced pancreatitis remains unknown. Sonia Prot-Labarthe, Pharm D, PhD Pharmacie APHP, Hôpital Robert Debré Université Paris Descartes Rym Youdaren Pharmacie APHP, Hôpital Robert Debré Malika Benkerrou, MD Centre de Drépanocytose APHP, Hôpital Robert Debré Romain Basmaci Pédiatrie Générale APHP, Hôpital Robert Debré Mathie Lorrot, PD, PhD Pédiatrie Générale APHP, Hôpital Robert Debré Faculté de médecine Denis Diderot Paris VII Paris, France