Pharmacogenomics in acute lymphoblastic leukemia

Abstract

Pharmacogenomics is a fast-growing field of personalized medicine using a patient's genomic profile to determine drug disposition or response to drug therapy, in order to develop safer and more effective pharmacotherapy. Childhood acute lymphoblastic leukemia (ALL), being the most common malignancy in childhood, which is treated with uniform and standardized clinical trials, is remarkably poised for pharmacogenomic studies. In the last decade, unbiased genome-wide association studies have identified multiple germline risk factors that strongly modify host response to drug therapy. Some of these genomic associations (e.g. TPMT, NUDT15 and mercaptopurine dosing) have accumulated a significant level of evidence on their clinical utility such that they are warranted as routine clinical tests to guide modification of treatment. Most of these germline associations however, have not yet reached such actionability. Insights have also been gathered on germline factors that affect host susceptibility to adverse effects of antileukemic agents (eg, vincristine, asparaginase, methotrexate). Further large-scale studies are required, along with the assimilation of both germline and somatic variants, to precisely predict host drug response and drug toxicities, with the eventual aim of executing genomic-based precision-pharmacotherapy in the treatment of ALL.

Introduction

The study of pharmacogenomics is a fast-growing research field, combining human genetics and molecular pharmacology, to determine how inter-individual genetic variations contribute to a patient's interaction with or response to drugs [1]. By using a patient's genomic profile to elucidate functionally relevant genomic determinants for drug disposition or response to drug therapy, we can aim to develop safer and more effective pharmacotherapy [2]. The idea is that the choice of treatment can be evaluated by weighing the probability of cure against the probability of adverse effects based on genetic characteristics. Host genomic profiles may also indicate drugs with little importance for cure but which are associated with a high risk for side effects for specific patients, allowing for better tailored therapy.

With rapid advances in genomic profiling technology, we are now able to efficiently examine variants across the entire genome in genome wide association studies (GWAS), with testing and analyzing a few million genetic markers per patient to determine the association of these variants with a particular phenotype of interest [3]. These large-scale studies have provided invaluable insight into the genetic basis of drug efficacy and toxicity during leukemia treatment.

Due to the exceedingly high number of variants analyzed in GWAS, the required level of significance for association between these variants and any particular phenotype is generally set at a much more stringent level of significance (P < 5 × 10⁻⁷), compared to the usual level of 0.05. Thus, it is expected that only genomic variations with relatively large effect sizes are detected in these studies, especially when the sample size is modest [4].

The majority of pharmacogenomics studies in leukemia are performed in childhood acute lymphoblastic leukemia (ALL), as ALL is the most common childhood cancer comprising up to 25% of all childhood malignancies and also a model disease for pharmacogenomics research, for many unique reasons. The translation of theis knowledge into new chemotherapeutic paradigms can serve as good models for optimizing treatment of other malignancies [5].

Firstly, childhood ALL generally has very standardized treatment protocols and most children under treatment are enrolled on prospective and uniform protocols where important data including presenting features, medication history, treatment outcome, and adverse events are systematically recorded.

Second, combination chemotherapy is highly efficacious in childhood ALL with a large majority of the patients being cured on pharmacotherapy alone. With medications playing such a large role in treatment, one can infer that pharmacogenomics would significantly influence the outcome of these patients. Furthermore, these medications operate on narrow ranges of therapeutic indices, hence it is imperative for us to understand the underlying genetic basis of antileukemic response and adverse effects, so that we can carefully navigate patients into the optimal therapeutic range of these drugs.

Most contemporary childhood ALL protocols utilize genomics of the ALL blasts to risk-stratify patients in order to decide the intensity of treatment. Hence, clinicians are already used to adjusting treatment regimens according to patient genomics.

Finally, many medications commonly used to treat childhood ALL are also used to treat many other non-malignant disorders. Genetic variants associated with drug effects, particularly adverse events, may have relevance for these other conditions as well. Knowledge gleaned from pharmacogenomics of these anti-leukemic agents can aid physicians in using the same class of drugs more effectively and appropriately for other conditions [1].

In cancer, both the somatically acquired genomic features of the malignancy and the inherited genomic variation of the host have the potential to influence efficacy of therapy. In this review article, we will focus primarily on germline variants which influence outcome of ALL. These variants affect the efficacy of antileukemic treatment through effects on response to treatment (as evidenced by MRD), drug sensitivity and relapse, as well as adverse drug effects [1].