Predictive and prognostic factors in the complex treatment of patients with colorectal cancer.

Colon cancer is the second most prevalent lethal cancer. The main cause for high mortality rate is that the prognosis for progressed metastatic colon cancer is most unfavorable. Recent data suggest that disease outcome can be further improved by the addition of targeted biological agents to the first- or second-line treatment. As a result of molecularly targeted anti-EGFR therapies (cetuximab and panitumumab) complementing chemotherapy, liver metastases can reduce in size and become operable in certain patients, which can contribute to the complete recovery of the patient. The main problem, however, is the fact that a positive response only occurs in one third of the patients, even in the case of chemotherapy combined protocol, and the side effects are considerable. For the application of individually tailored treatments, it is an urgent need to develop a system of biomarkers that can predict the effect of treatment and provide information about the optimal selection of both chemotherapy and biological treatment. It should be clarified what the most important requirements of a good and reliable biomarker are. As currently there is no precise predictive molecular diagnostics at our disposal, oncologists have to make one of two choices: they treat a large number of patients with anti-EGFR agents which has negative effects on the quality of life and also reduces the patient's chances of getting appropriate treatment or, if the oncologists refuse to take risks, they omit the use of anti-EGFR treatment in which case those patients for whom this would have been the appropriate treatment are also denied the chance of short-term survival or recovery. Clinical data (response rate, time to progression (TTP) and overall survival (OS)) of 130 colorectal cancer patients have been retrospectively analyzed. Patients have received different chemotherapy protocols in combination with anti-VEGF antibody or with anti-EGFR antibody therapies. EGFR expression was evaluated with immunohistochemistry, KRAS, BRAF and PIK3CA mutations were evaluated by direct sequencing and high resolution melting analysis in the archived formalin-fixed, paraffin-embedded tissue samples. The study found similar efficacy of first-line therapeutic protocols. Protocols combining chemotherapy with biological therapies achieved better overall survival but this difference was not significant (OS: 35.9 versus 36.7 months). The frequency of KRAS mutations was 44% (n=100). None of the KRAS mutant tumors responded to the anti-EGFR monotherapy. TTP in the case of cetuximab monotherapy was twice longer (208 months) than in the KRAS mutant tumors (104 months). One BRAF mutant tumor was also identified (4%) This tumor was also resistant to cetuximab monotherapy. The KRAS and BRAF mutations excluded each other. Except one case, the KRAS status was identical in both the primary tumor and the metastasis. In contrast, PIK3CA mutations were heterogeneous in different tumor samples. In 5 out of 6 cases the mutation status of PIK3CA was different in the primary tumor and the metastasis. New biological therapies provide an additional clinical benefit only for a subset of patients. We need biomarkers to identify these patients. KRAS and most probably BRAF testing can double the efficacy of the anti-EGFR therapies, but we need additional molecular diagnostic tests. PIK3CA is an important candidate but we might need to take biopsy directly from the metastasis or we have to evaluate the circulating tumor cells to judge the molecular status of distant metastasis.