METHODS FOR PREDICTING THE OUTCOME OF A CANCER IN A PATIENT BY ANALYSING GENE EXPRESSION

1 . (canceled)

2 . A method for predicting the outcome of a cancer in a patient comprising the steps of:

i) determining the expression level of a gene cluster in a sample obtained from said patient, wherein said gene cluster consists of at least 7 genes selected from the group consisting of GZMH, IFNG, CXCL13, GNLY, LAG3, ITGAE, CCL5, CXCL9, PF4, IL17A, TSLP, REN, IHH, PROM1, and VEGFA by

extracting mRNAs from one or more biopsies of a tumor tissue of the patient,

measuring the expression level of the genes of said gene cluster in said extracted mRNAs;

ii) calculating a score by i) determining for every gene of the cluster their expression level in the sample ii) allocating for every gene a positive coefficient when the gene is associated with a good prognosis and a negative coefficient when the gene is associated with a bad prognosis and iii) multiplying the expression level of one gene with the coefficient allocated in step ii) and iv) adding up all values determining at step iii) for obtaining said score wherein a coefficient of +1 is allocated to the genes associated with a good prognosis, namely GNLY, CXCL13, CXCL9, LAG3, CCL5, GZMH, ITGAE and IFNG, and a coefficient of −1 is allocated to the genes associates with a bad prognosis, namely PF4, REN, TSLP, IL17A, PROM1, IHH and VEGFA;

iii) comparing the score calculated at step ii) with a predetermined reference value, determined by the expression levels of said gene cluster in reference samples, and

iv) concluding that the patient has a good or a bad prognosis depending on whether the score is higher or lower than the reference value.

3 . The method according to claim 2 , wherein said gene cluster consists of 8, 9, 10, 11, 12, 13, 14 or 15 genes selected from the group consisting of GZMH, IFNG, CXCL13, GNLY, LAG3, ITGAE, CCL5, CXCL9, PF4, IL17A, TSLP, REN, IHH, PROM1, and VEGFA.

4 . The method according to claim 2 , wherein the expression level of the genes of said gene cluster in said extracted mRNAs is measured by performing a real-time polymerase chain reaction (RT-PCR).