Abstract
ObjectivesLittle research has been done in pharmacoepidemiology on the use of machine learning for exploring medicinal treatment effectiveness in oncology. Therefore, the aim of this study was to explore the added value of machine learning methods to investigate individual treatment responses for glioblastoma patients treated with temozolomide.MethodsBased on a retrospective observational registry covering 3090 patients with glioblastoma treated with temozolomide, we proposed the use of a two-step iterative exploratory learning process consisting of an initialization phase and a machine learning phase. For initialization, we defined a binary response variable as the target label using one-by-one nearest neighbor propensity score matching. Secondly, a classification tree algorithm was trained and validated for dividing individual patients into treatment response and non-response groups. Theorizing about treatment response was then done by evaluating the tree performance.ResultsThe classification tree model has an area under the curve (AUC) classification performance of 67% corresponding to a sensitivity of 0.69 and a specificity of 0.51. This result in predicting patient-level response was slightly better than the logistic regression model featuring an AUC of 64% (0.63 sensitivity and 0.54 specificity). The tree confirms confounding by age and discovers further age-related stratification with chemotherapy-treatment dependency, both not revealed in preceding clinical studies. The model lacked genetic information confounding treatment response.ConclusionsA classification tree was found to be suitable for understanding patient-level effectiveness for this glioblastoma–temozolomide case because of its high interpretability and capability to deal with covariate interdependencies, essential in a real-world environment. Possible improvements in the model’s classification can be achieved by including genetic information and collecting primary data on treatment response. The model can be valuable in clinical practice for predicting personal treatment pathways.
Highlights
Glioblastoma is one of the most common and aggressive brain tumors in adults, with a median survival of less than one year from the time of diagnosis
Associated with this area under the curve (AUC) was a sensitivity of 0.6850, meaning that 31% of patients who would benefit from the treatment were not recognized by the model, and a specificity of 0.5114, meaning that 49% of patients who would not benefit from the treatment were predicted to benefit by the model
We showed an increased understanding of patient-level treatment responses and specification of individual treatment paths that were not be identified using cohort-oriented methods used in previous Randomized Controlled Trials (RCTs) studies
Summary
Glioblastoma is one of the most common and aggressive brain tumors in adults, with a median survival of less than one year from the time of diagnosis. Specific details on the mechanisms that drive individual response to temozolomide treatment in clinical practice, or on the drivers of real-world patient-level treatment effectiveness, are unknown (van Genugten et al, 2010; Eichler et al, 2011; Liu et al, 2016) To study these personal responses, traditional cohort-oriented methods, such as the Kaplan-Meier survival techniques currently used in pharmacoepidemiology (Strom and Kimmel, 2006) for investigating real-world evidence (RWE) data, have shown to be inadequate because of their difficulties to cope with heterogeneous patient populations; their restrictive assumptions regarding linear relationships among variables; their inability to provide patient-level predictions; and their inability to infer causality (Ankarfeldt et al, 2017; Arora et al, 2019). Other statistical methods commonly used in the domain of medicine, such as logistic regression models, have hitherto focused mainly on investigating survival probability and their associated confounding factors when used in pharmacoepidemiology, as opposed to treatment effectiveness (Burke et al, 1995)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
