Methods are provided for determining whether a cancer patient is likely or not likely to experience HFSR (hand foot skin reaction) from a therapy comprising administration of an effective amount of regorafenib or an equivalent thereof, the methods including screening a biological sample isolated from the patient for the rs2280789 polymorphism, and/or the rs3817655 polymorphism. Methods are further provided for identifying the clinical outcome of cancer patients following a therapy comprising regorafenib or an equivalent thereof, the methods entailing screening a biological sample isolated from the patient for the rs7527092 polymorphism, and/or rs2071559 polymorphism. After determining if a patient is likely to be successfully treated, the disclosure also provides methods for treating the patient.