The Clinical Significance of Non-Tuberculosis Mycobacterium (NTM) Infection in Pediatric Hematopoietic Cell Transplant (HCT) Recipients

Introduction:  Infections with NTM species are reported in immunocompromised patients (pts) post-HCT with little information on associated risk factors. In HIV+ pts, routine prophylaxis is indicated for those with CD4 counts of 100-200cells/uL.  We hypothesize CD4 counts ≤200 as a possible risk factor for development of NTM infection in the pediatric alloHCT setting. We report our center’s experience with NTM infections to help identify risk factors, clinical significance and need for prophylactic therapy.

Methods: Pts ≤21 years undergoing HCT from 2001 to 2014 were included. Definition of NTM infection: Positive culture for NTM species in blood, cerebrospinal fluid (CSF), lungs, BM or tissue. Only pts with Mycobacterium Avium Complex (MAC) infection were included. Controls (n=19) were identified by matching age, gender, and underlying disease. Descriptive statistics was calculated and comparison between 2 groups was carried with t-test for continuous variables and Fisher exact test for categorical variables.

Results:  Of 272 pts undergoing HCT, NTM was diagnosed in 11 pts with 17 HCT (autologous=1; allogeneic=16) performed among these pts. Indications for HCT:  malignant=6 (55%), non-malignant=5 (45%). Donor sources:  sibling BM (n=3), sibling PB (n=2), unrelated PB (n=3) unrelated BM (n=3), or unrelated UCB (n=5). Conditioning regimens: myeloablative (n=7), non-myeloablative (n=6), no conditioning (n=3). Pts had MAC isolated from the following sites: lungs (n=5), CSF (n=1), BM (n=2), lymph node (n=1), and blood (n=3). Median date from HCT to NTM diagnosis was 233 days (range: 15 to 724 days). Mean absolute CD4 count at diagnosis was 137 +/- 143 cells/uL (1 unreported). All pts began therapy with azithromycin, ethambutol, and rifabutin/rifampin.  Additive therapy included levofloxacin (n=3), imipenem (n=1), amikacin (n=3), and gatifloxacin (n=1). Treatment duration ranged from 53-459 days.  Eight of 11 pts had graft-vs-host disease (GVHD) with 75% (n=6) diagnosed prior to the onset of NTM. Overall survival was 73%(n=8). Causes of death were:  complications due to NTM, progressive underlying disease, and pulmonary failure secondary to chronic GVHD. Of the 8 surviving pts, 1 had recurrence, 5 cleared the infection, and 2 are in treatment.

No statistically significant risk factors for NTM were found, including presence of GVHD, CD4 count at diagnosis, CMV viremia, or donor source. Similar results were obtained when not controlling for these variables. No statistically significant difference was found between pts who developed MAC infections and those that did not.

Discussion

These results were unable to identify low CD4 counts as a potential risk factors for NTM necessitating the need for multicenter collaborations. NTM appears to be a treatable disease, with the role of prophylaxis unclear.  With the extensive treatment duration, healthcare utilization costs should also be evaluated.