Every cancer phenotype starts as a single mutation event in one of its ancestor cells. During the subsequent cell division cycles, the daughter cells develop additional mutations to gain oncogenicity.

There are over 570 genes that are frequently mutated in cancer, and the majority of them are somatic mutations. That means that such mutations are not inherited but developed in the body's somatic cells during an organism's lifetime.

However, a single mutation is not sufficient to turn a healthy cell into a cancerous cell. At least five to six independent and rare genetic alterations in succession over a period of time can result in malignancy. This multi-step process of cancer development is explained by multi-hit theory. 

Consider a group of healthy somatic cells. In rare instances, one of the cells may acquire a random mutation in one of its cell cycle regulator genes - gene X, allowing it to divide slightly faster than the neighboring cells.

Suppose the mutation is left unrepaired by the DNA repair enzymes. In that case, the progenies of this cell will carry the same mutated gene X. Over time, one of these clonal cells may acquire another random mutation in a different gene, Y.

This new mutation might allow the daughter cells to accumulate random mutations at a rate faster than their healthy neighbors and other clonal cells.

Subsequently, a third mutation in gene Z, another cancer-critical  gene, may allow cells to escape terminal differentiation and apoptosis. Such mutant cell types continue to grow to form a mass of abnormal cells.

Further, a random mutation in metabolic pathway genes may allow cells to increase metabolism to fuel rapid growth and tumor formation.

For instance, mutations in the cancer-critical genes, such as APC, c-Myc , K-Ras, and p53, are frequently observed in patients with colon carcinoma.