Biomarkers for Chemotherapy Associated Neurotoxicity

Neurotoxicity during treatment for childhood ALL/LBL remains a significant problem. It can be acute as in the "stroke-like syndrome" seen with methotrexate, or may result in chronic neurocognitive defects. Neurocognitive impairment is estimated to occur in between 20-50% of long-term survivors. Although mean effects on global measures such as intelligence quotient (IQ) are moderate, this hides a significant number of severely affected individuals and the risk factors for severe impairment are poorly understood. Recent studies have highlighted that neurocognitive impairment may continue to evolve throughout adulthood, even with chemotherapy-only regimes. Worryingly, reported imaging and neurocognitive results are similar to those seen in early-onset dementia and the burden of neurological late effects may worsen over the next few decades. Thus, there is an urgent need to better understand the pathophysiology of neurotoxicity in childhood ALL/LBL, develop ways of identifying children at risk and to devise new treatments and/or protective measures.

For those without overt neurotoxicity symptoms, identification of patients at risk for poor neurocognitive outcomes might be achieved using early sensitive measures of neurocognition. Long-term follow up studies indicate that children treated for ALL/LBL have particular problems with memory, executive functions, attention and processing speed. Whether these can be picked up early enough to act as a biomarker for long-term adverse outcome is unknown and will be tested in this study. Alternatively, biomarkers in cerebrospinal fluid may indicate high levels of toxic metabolites or biochemical markers of neurological damage that precede neurocognitive effects. The most highly implicated neurotoxic drug is methotrexate. Methotrexate pharmacokinetics do not correlate with neurotoxicity but secondary metabolites such as folate, homocysteine, adenosine and glutamate analogues have been implicated in both acute and chronic toxicity. An animal model of methotrexate neurotoxicity suggests that excitotoxic glutamate analogues may cause neurotoxicity by binding to N-methyl D-aspartate (NMDA) receptors. Importantly, in this pre-clinical model, NMDA antagonists can reverse and/or prevent neurotoxicity when administered concurrently with methotrexate. This raises the possibility that cerebrospinal fluid (CSF) biomarkers will not only identify children at risk of neurotoxicity but also inform therapeutic strategies.

Finally, in the era of personalised medicine an alternative approach is to develop a panel of genetic predictors of neurotoxicity, which can prospectively identify high-risk children prior to any damage. Several candidate polymorphisms for chemotherapy-induced neurotoxicity have been proposed using small cohorts of patients with diverse cancer types. The investigators will test these prospectively in a large standardised patient cohort. In addition, patients exhibiting overt chemotherapy related neurotoxicity such as those with SPS are likely to exhibit the highest hazard ratios for genetic predisposition and therefore their inclusion in this study will strengthen the likelihood of finding clinically relevant and actionable polymorphisms.

Overall, the investigators predict that this study will provide significant insights into potential neurocognitive, biochemical and genetic biomarkers for neurotoxicity and provide strong underpinning science to determine whether NMDA antagonists are suitable for prevention and/or treatment of these potentially devastating neurotoxic effects