Similar Incidence of Typhlitis in Patients Receiving Various Doses of Daunorubicin or Idarubicin As Induction for Acute Myeloid Leukemia

Background: The current standard of care for the treatment of patients with newly diagnosed AML is an anthracycline plus cytarabine. Both anthracyclines and cytarabine have been associated with the development of typhlitis, a serious adverse event characterized by inflammation of the bowel wall in patients with profound neutropenia, diagnosed via abdominal CT imaging and clinical symptoms. Given the paucity of available data, the aim of our study was to determine the incidence of typhlitis among AML patients receiving induction with either idarubicin 12 mg/m² (IDA), daunorubicin 60 mg/m² (DNA60), or daunorubicin 90 mg/m² (DNA90).

Methods: Adult patients with AML or MDS receiving either daunorubicin or idarubicin along with cytarabine as part of their induction regimen between 1/1/2009 and 6/30/2013 were included. A definition of typhlitis required CT confirmation of inflammation of the cecum, according to CTCAE version 4.03 along with clinical symptoms. Two radiologists (OA, JC) blinded to the treatment and outcomes independently reviewed CT scans. Two additional definitions including inflammation of the ileocecal region only and enterocolitis were also evaluated. All statistical analyses were performed on SAS software version 9.3. P values were calculated using Fisher Exact and Wilcoxon tests. Inter-rater reliability was assessed with Cohen’s Kappa.

Results: Baseline characteristics were similar among the 3 treatment groups with the exception of age. The median age was lower in the DNA90 arm (79 years, 74 years, and 49 years in the IDA group, DNA60 group, and DNA90 group, respectively). A pre-existing GI disorder was reported in 24.1% IDA, 25.7% DNA60, and 22.4% DNA90 patients.

Of the 202 total patients, the two radiologists determined that 40 (20%) and 38 (19%) developed typhlitis, based on the predefined standard. Tables 1 and 2 illustrate the relationship between treatment arm and associated incidence of typhlitis. The incidence in each treatment group did not statistically differ (p=0.23 and p=0.29). When the definition was broadened to include ileocecal region and enterocolitis, the incidence increased (Tables 1 and 2). The inter-reliability ratings of the 2 radiologists’ evaluations for each definition indicated substantial agreement (0.803 cecum, 0.834 ileocecal region only, and 0.752 enterocolitis). Neither the anthracycline chosen, nor the dose had a statistically significant impact on the incidence of typhlitis. Of all patient and treatment specific risk factors assessed for association with development of typhlitis (pre-existing GI disorder, rheumatologic disorder or cancer, prior RT, stem cell transplant or anthracycline exposure, cytarabine regimen or AML risk group), none were found to be statistically significant. In patients that developed typhlitis, approximately 56% and 51% had a concurrent documented infection around the typhlitis episode (Tables 1-2).

Conclusion: To our knowledge, this is the first study to compare the incidence of typhlitis in adult patients receiving idarubicin or daunorubicin for the treatment of AML. While the cumulative incidence of typhlitis was higher than in published literature, the incidence was similar irrespective of the anthracycline chosen or dose. Of the potential factors that may have contributed to the development of typhlitis, none were significantly associated with typhlitis. All patients were managed conservatively with broad-spectrum antibiotics. A more definitive definition of typhlitis may help clinicians identify affected patients sooner and choose appropriate targeted therapy.