An Epidemiologic Approach to Outcome Prediction
Margaret R. Spitz,Xifeng Wu
IF: 13.801
2006-01-01
Clinical Cancer Research
Abstract:PL-12 Overview: There is substantial interindividual variation in drug disposition, efficacy, and toxicity. This is of particular relevance for drugs with low therapeutic indices, such as chemotherapeutic agents. Much of this variation can be explained by specific genes that influence the absorption, metabolism, or disposition of drugs or are in the target cellular pathway. Therefore, by analyzing the genetic make-up of both the tumor and the patient, we can generate a "personalized molecular medicine" profile, or pharmacogenetic profile that can be used to individualize therapy and help understand the functional consequences of chemoprevention, chemotherapy, or radiotherapy response and toxicity. This approach will also enable clinicians to select drugs likely to be most effective and least toxic for each patient. This is the promise and potential of pharmacogenetics.We now have available a new series of high-throughput, low-cost technologies to comprehensively evaluate changes in tumor DNA (genomic analysis), RNA (transcriptional profiling), and protein (proteomic analysis) and determine the consequences of those changes (functional genomics). We can also evaluate the patient9s own genetic make-up (SNP analysis) to identify molecular markers of susceptibility to toxic effects of particular drugs. By selecting patients who are most likely to respond to a drug9s mechanism of action, it should be possible to obtain indications of efficacy with a smaller number of patients, and make best use of the precious resource of patients participating in clinical trials. In addition, it would decrease the costs of clinical trials and correspondingly increase the number of trials that can be undertaken with fixed financial resources.This session will focus on germline determinants of therapeutic efficacy and toxicity. Examples will be provided of how changes in the function of a single gene in the target cellular pathway genes, metabolizing enzymes or major histocompatibility complex can contribute to patient prognosis and prediction of response. Label changes to include pharmacogenetic testing are on the horizon. However, it must be remembered that response to a therapy is the result of a complex interaction between genetic and nongenetic factors, and prediction is not perfect.Molecular epidemiology will assume an increasingly prominent role in pharmacogenetic research, for construction of genetic profiles. A unifying premise to applying a molecular epidemiologic approach to outcome prediction is that the same genes that are implicated in cancer risk, are also involved in the prediction of patient outcome. In this way, we will attempt to illustrate that risk prediction and prediction of outcome are part of a continuum. We will select two representative examples relative to lung cancer to illustrate our theme.Treatments for lung cancer have improved survival only marginally over the past few decades. Furthermore, these treatments have a narrow therapeutic index, and response rates are unpredictable, even controlling for the known predictors of stage, histology, grade and performance status. Drug resistance is a major problem since even if there is an initial favorable response rate, most patients relapse. Radiation toxic effects are similarly unpredictable and are dose limiting. In this presentation we employ a pathway based genotyping approach, which assesses the combined effects of a panel of polymorphisms that interact in the same pathway, rather than focus on the candidate gene approach. In this way, we hope to amplify the effects of individual polymorphisms and enhance the predictive power. DNA repair: DNA repair may be considered a double-edged sword. Nucleotide excision repair (NER) removes bulky adducts induced both by benzo[a]pyrene in tobacco and by platinum chemotherapy. Suboptimal DNA repair capacity, as assessed by the in-vitro host cell reactivation assay, is an independent risk factor for cancer, but has also been shown to be associated with improved response to chemotherapy (due to decreased removal of tissue platinum adducts). We have applied a pathway-based approach to examine the associations of a comprehensive panel of polymorphisms in the NER pathway with both risk and patient outcome. We found a significant gene dosage effect for increasing risk associated with increasing numbers of potential high-risk alleles in DNA repair pathways separately and combined. These variant genotypes also appeared to predict outcome in patients with advanced stage non small cell lung cancer treated with platinum-based chemotherapy.Detoxification of carcinogens and of platinum compounds by conjugation with glutathione or GSH is catalyzed by the multi-gene family of glutathione S -transferases (GSTs) yielding glutathione S -conjugates that are more water-soluble and often less reactive than their parent compounds. Such inactivation of tobacco carcinogens is protective against lung cancer risk, but a major mechanism of tumor resistance to platinum compounds. Thus variant alleles associated with reduced GST activity would be expected to be associated with increased cancer risk, but improved response to chemotherapy. Again, we used a pathway based approach to evaluate variants in three genes in this pathway, GSTM1, T1, and Pi and both risk and outcome in lung cancer patients undergoing chemoradiation.This multivariate approach could have long-term prognostic potential by allowing greater individualization of therapy, based on the genetic makeup of the patient, thereby both increasing efficacy and reducing morbidity. In summary, molecular epidemiology can provide a major integrative force both for cancer risk assessment and for prediction of outcome.This project was supported by NCI grants CA55769 (Spitz), DOD (Hong), Flight Attendant Medical Research Institute (Spitz), CA70907 (Minna, Roth)