A method of predicting the likelihood of a patient with a solid tumor cancer or a metastasis thereof responding to cancer therapy comprises the steps of assaying a sample of the primary solid tumor or a local or distant metastasis thereof obtained from the patient to determine a level of expression of CerS5, and if an elevated CerS5 expression level is detected in the sample from the patient, then predicting a likelihood that the patient will have a positive clinical response to cancer therapy, or if a non-elevated CerS5 expression level is detected in the sample from the patient, then predicting a likelihood that the patient will have a negative clinical response to cancer therapy.

The invention relates to a method of screening a patient to identify and quantify risk of colorectal cancer, and thereby identify patients suitable for further invasive investigation such as a colonoscopy. The method employs auto-antibodies that are shown to correlate with colorectal cancer risk method and involves of assaying a biological sample from the individual for a combination of a plurality of biomarkers selected from SEQUENCE ID NO's: 1 to 12, where the combination of biomarkers is chosen such that detection of all biomarkers in the patient correlates to at least a 50% risk of the patient being positive for colorectal cancer. Detection of all of the combination of biomarkers indicates that the patient should undergo a colonoscopy. Kits for performing the method of the invention are also provided.

Related: