Prognostic impact of mitochondrial DNA D-loop variations in pediatric acute myeloid leukemia

• Published in: Oncotarget Vol. 10; no. 13; pp. 1334 - 1343
• Main Authors: Tyagi, Anudishi, Pramanik, Raja, Vishnubhatla, Sreenivas, Bakhshi, Radhika, Bakhshi, Sameer
• Format: Journal Article
• Language: English
• Published: United States Impact Journals, LLC 12.02.2019; Impact Journals LLC
• Subjects: Research Paper; leukemia; D-loop; pediatric AML; prognosis; mitochondrial DNA
• ISSN: 1949-2553; 1949-2553
• DOI: 10.18632/oncotarget.26665

The role of mitochondrial DNA (mt-DNA) changes, especially those in the regulatory D-loop region in Acute Myeloid Leukemia (AML) remains investigational. Consecutive 151 pediatric AML patients, (≤18 yr) were prospectively enrolled from June 2013-August 2016, to assess the prognostic impact of mt-DNA D-loop variations (somatic/germline) on survival. For each patient, D-loop region was sequenced on baseline bone marrow and buccal swab, and mother's blood sample. In 151 AML subjects, 1490 variations were found at 237 positions; 80.9% were germline and 19.1% somatic. The mean number of variations per position was 6.3. Variations with frequency ≥6 were analyzed for their impact on survival and 4 categories were created, namely "somatic-protective", "somatic-hazardous", "germline-protective" and "germline- hazardous". Although, somatic-protective could not predict event free survival (EFS) or overall survival (OS), somatic-hazardous [(OS) HR = 2.33, = 0.06] and germline-hazardous [(OS) HR = 2.85, < 0.01] significantly predicted OS and EFS. Notably, the germline-protective, could significantly predict EFS (HR = 0.31, = 0.03) and OS (HR = 0.19, < 0.01), only when variations at ≥2 positions were present. On multivariate analysis, three positions namely 16111, 16126, 16362 and karyotype were found to be predictive of EFS. A prognostic index (PI) was developed using nomogram PI = (0.8*karyotype) + (1.0*c16111) + (0.7*t16362) + (1.2*t16126). Hazard ratio for EFS increased significantly with increasing PI reaching to a maximum of 3.3 ( < 0.01). In conclusion, the impact of mt-DNA D-loop variations on outcomes in pediatric AML depends on their nature (germline/somatic), position and mutational burden, highlighting their potential role as evolving prognostic biomarkers.