A Metabolomics Approach for Predicting the Response to Neoadjuvant Chemotherapy in Cervical Cancer Patients

Yan Hou,Mingzhu Yin,Fengyu Sun,Tao Zhang,Xiaohua Zhou,Huiyan Li,Jian Zheng,Xiuwei Chen,Cong Li,Xiaoming Ning,Ge Lou,Kang Li
DOI: https://doi.org/10.1039/c4mb00054d
2014-01-01
Molecular BioSystems
Abstract:Cervical cancer is a clinical and pathological heterogeneity disease, which requires different types of treatments and leads to a variety of outcomes. In clinical practice, only some patients benefit from chemotherapy treatment. Identifying patients who will be responsive to chemotherapy could increase their survival time, which has important implications in personalized treatment and outcomes, while identifying non-responders may reduce the likelihood for these patients to receive ineffective treatment and thereby enable them to receive other potentially effective treatments. Plasma metabolite profiling was performed in this study to identify the potential biomarkers that could predict the response to neoadjuvant chemotherapy (NACT) for cervical cancer patients. The metabolic profiles of plasma from 38 cervical cancer patients with a complete, partial and non-response to NACT were studied using a combination of liquid chromatography coupled with mass spectrometry (LC/MS) and multivariate analysis methods. L-Valine and L-tryptophan were finally identified and verified as the potential biomarkers. A prediction model constructed with L-valine and L-tryptophan correctly identified approximately 80% of patients who were non-response to chemotherapy and 87% of patients who were had a pathologically complete response to chemotherapy. The model has an excellent discriminant performance with an AUC of 0.9407. These results show promise for larger studies that could produce more personalized treatment protocols for cervical cancer patients.
What problem does this paper attempt to address?