Aims: A recent study of SELDI-TOF MS analysis of serum has demonstrated accurate multimodal therapy response prediction ability early into treatment for rectal cancer. This study aims to assess response prediction of serial serum sampled at early time-points in a larger cohort of patients.
Methods: Serum from 40 consecutive patients was included in the study at the following time-points pre-treatment, 24 hrs and 48 hrs into treatment. The response to treatment was measured by Mandard Tumour Regression Grade (TRG), a 5-point scale based on the degree of residual tumour to fibrosis. All serum samples were denatured prior to analysis by SELDI-TOF mass spectrometry using a weak cation exchange ProteinChip® array.
Results: Spectra from 17 good responders (TRG 1+2) and 23 poor responders (TRG 3+4) were generated. Technical replicates were performed. Data was analysed using Ciphergen ExpressTM software. Using a high level of stringency six statistically significant protein peaks were differentially expressed between the good and poor response categories. These peaks were present in the low molecular weight region of the human proteome.
Conclusions: The results of our study indicate that analysis of low molecular weight serum proteins using SELDI-TOF may provide an accurate and non-invasive means of response prediction to RCT early into treatment. We are currently analysing these samples on two other ProteinChip® arrays types (strong anion and metal ion binding surfaces). Fractionation of these samples will then be performed, by Albumin and IgG depletion of the serum. This will enable peaks suppressed by these high abundance proteins to be detected. Our main focus will be on the identification of the putative biomarkers of response prediction.