Response to “: A Validation Study in a Spanish Cohort”PNPLA3: A Validation Study in a Spanish Cohort” rs738409 and Hepatotoxicity in Children With B‐cell Acute Lymphoblastic Leukemia: A Validation Study in a Spanish Cohort”

    loading  Checking for direct PDF access through Ovid


To the Editor: The association between PNPLA3 rs738409 and higher liver enzyme levels in children with acute lymphoblastic leukemia (ALL) identified in our study of ALL in children in the United States1 was replicated in a Spanish cohort,2 despite differences in treatment and analysis. First, the Spanish study used a cohort of patients with ALL with mostly European ethnicity, whereas our study included patients from diverse ethnicity groups with different PNPLA3 rs738409 minor allele frequency and predisposition to hepatotoxicity. However, the Spanish study used the highest grade of hepatotoxicity observed rather than toxicity grade at the end of induction. This difference explains at least partially why the Spanish study showed more patients with higher hepatotoxicity grades than ours. Third, the Spanish study used not only alanine aminotransferase but also aspartate aminotransferase levels to decide hepatotoxicity, which might have also resulted in more toxicity events than our study. Moreover, the remission induction regimens in the two studies are not the same. Compared to our US cohort, the Spanish cohort was treated with higher doses of daunorubicin (120 mg/m2 vs. 25 mg/m2) and prednisone (tapering from 60 mg/m2/day vs. 40 mg/m2/day) and exclusively used native asparaginase compared to our use of mostly pegylated asparaginase.
It should be acknowledged that the PNPLA3 rs738409 variant has been associated with alanine aminotransferase elevation in the general population, nonalcoholic fatty liver disease, and hepatic fibrosis3 in population studies that are not necessarily linked to drug therapy. In fact, this is one of the few genomic variants that seem to predispose to risk of both drug‐induced as well as constitutive liver disease.
Challenges remain in identifying and replicating pharmacogenetic risk factors due to diverse criteria of adverse events grading. Defining the upper limit of normal (ULN) range of liver enzymes in infants, children, and adolescents with the same value for adults may not be appropriate. Many current ULN standards are higher than what is needed for sensitive monitoring.5 The ULN for liver enzymes varies from hospital to hospital, increasing the difficulty for collaboration and meta‐analysis. Reaching consensus definitions of age‐dependent and gender‐dependent ULN and grading criteria may improve comparability and replicability between study groups.
The replication of our discovery in a different cohort highlights the impact of PNPLA3 and liver lipid metabolism on hepatotoxicity caused by ALL treatment. Monitoring hepatotoxicity along with genetic interrogation may facilitate better elucidation of risk factors for drug‐induced adverse effects in patients with cancer.
    loading  Loading Related Articles