Last year, one of my professors made the statement that "You cannot inherit cancer". This seems very contradictory to what I grew up with; most cancer patients I've met have a family history of cancer, which is also why I am pretty sure I'll get it one day, between the prevalence on either side of my family and the frequent exposure to carcinogens.
The truth of the statement tends to be easier to grasp once you go down into developmental biology. Cancer at its root is the unregulated growth of cells. This is a stark contrast to development, which is the process of a highly specific differentiation of cells. Starting from a relatively spherical egg, a human being begins to form "front-sides" and "back-sides", ending up with different organs, which contin highly specialized cells types: nerve cells, skin cells, muscle cells, etc. Without regulation, not only will there be too many cells, the functions that is needed for survival of a human would no arise. If an individual can proceed through development without trouble, then there must have been changes that occurred after the germ cells left the two parents.
So why are descendent of people with cancer are more likely get cancer themselves? It is because of the high amount of genetic mutations that need to occur before a cell becomes cancerous. Without a whole slew of these mutations, cells are not cancerous. So someone can pass on a lot of these mutations, but until the individuals obtain more mutations, s/he will not have cancer.
This entry, and I'm suspecting the next, will be describing some of the mutations necessary.
First off, comes telomeres. Telomeres are a bunch of repeating nucleotides, the building blocks of DNA, at the end of strands of DNA. At first it seems that they are relatively useless, as they do not code for any functions - TTAGGG, TTAGGG, TTAGGG, so on and so on. They can't be read to make proteins, which provide most of the functions of a cell. It turns out, when DNA is replicated, in one direction, the cell machinery can read all the way through, but in the other, the machinery needs to start and stop all the time. Everytime it starts, it needs some preformed short fragments of nucleic acids, RNAs, from which the additional DNAs can attach to. After replication, the machinery goes along and replace all the RNA fragments with DNA. One downfall of this technology is that it needs DNA pieces of the other side of the fragment in order to replace the RNA. The last bits of RNA will be by themselves, and will not be able to replace to DNA, before they are just chewed up and destroyed. So everytime new cells are formed, their DNA becomes just a bit shorter. At some point, the "useless" DNAs are gone, and every time a cell replicates, it starts losing bits of genetic information that are necessary for the functions of the cell. These changes can be to imoprtant for the survivals of the cell, or in some cases, gives just enough damage that it'll push the cell towards being cancerous. One thing that a potential cancer cell will need to do is somehow stop this limit (See Hayflick limit) on the amount of replications a cell can undergo before it stops being able to survive. To accomplish this, they need to borrow technology from one of the few cell types that are constantly dividing - germ line cells.
With the Hayflick limit, each cell can usually undergo 40~60 divisions before stopping division, but humans definitely had more than 60 generations. It is because germ lines, embryonic stem cells, immune cells, and a few rare others have an enzyme called telomerase that would reverse the effect of the shortening DNA by adding more TTAGGG sequences back into the code. The gene that encodes for this enzyme is in all cells, but it is only active in a few rare ones. Without being able to reactivate this gene, even the most potent mutations, the mutated cells will eventually die off.