Lethal dose of Gy's (Grays) The unit Gray has the dimensions $\text{J/kg}$. I've read, that a dose of about 3-5 Gy's could kill a person within a few weeks - or at least that's usually the case.
But I'm not really understanding the concept I think. If the lethal dose is 5 Gy, that would mean about 400 J for a 80 kg person. But how is that dose then distributed ? Does the entire body take an average radiation, summing up to 400 J, or...?
Because in radiotherapy, some fractions can be around 2.5-3 Gy's (Of course in smaller areas) - and in the long run, some patients get over 70 Gy's. Does that mean, that they actually calculate the tumour mass, and then figure out what amount of Joule has to be distributed into the tumour ? And if so, why doesn't that kill people, when they get way over 5 Gy's. 
Hope you understand what I mean.
 A: The dose that kills a tumor is deliberately aimed at that tumor. If, instead of using a collimated beam, you put a person in a wide beam for radio "therapy", you would be treating their entire body as a tumor and kill them.
The dose in RT is computed locally - "this" part of the body (these grams of tissue) absorbed (were exposed to) "this many" Joules of ionizing radiation . 
You can't simply average local dose over the entire body - although blood perfusing the irradiated area will carry some damaged cells to other parts of the body, the majority of the tissue / cells in the body get a dose below the damage threshold - and the body is pretty good at repairing itself.
Note however that some organs are much more sensitive than others - there are tables of acceptable maximum dose (or 'dose sensitivity') that show that organs involved in reproduction (gonads) and blood production (bone marrow), and organs with a low ability to self-heal (spinal cord, brain, eyes) or that are acutely vital (heart), should be spared as much as possible during RT.
Having said that, RT does carry a risk - but when a patient is already diagnosed with cancer it is better to take away as much of the "known bad" tissue and take the risk of creating the potential of other mutated cells, rather than leave the cancer alone. At least that is the judgment that oncologists make on a per case basis when they refer for RT.
I will look up some references...
A: Radiographer working in radiation treatment here. (called MTRA in germany)
While I can't offer as much insight from a scientific standpoint I can offer a view from first hand treatment. If you speak about full body dosage it actually means full body dosage.  If you reduce the area you can go a good deal over 5 Gy. For example some kinds of lung and brain cancers that act very aggressive but are very small can be treated in a concept called stereotaxy where you apply a dosage of up to 8 Gy per day multiple times. This is only done with even more tightly controlled machines then usual and with a quasi-CT scan before the treatment. In case of lung cancer there is even an additional CT scan in the middle of the treatment. (Around 2 mSv per scan instead of the 20-30 mSv for a normal CT.)
It is very important for that that the area we treat is very small (a few mm in radius) and even then we are applying the dosage from 10 different arcs to minimize dosage at the skin and other unwanted places.
Another very important aspect of radiation damage is the different  limits of different organs and tissue types. The eye lenses for example are extremely sensitive to radiation.
But especially through VMAT, where we apply the radiation in a constant, continuous arc while also modulating the beam via a irregular area collimation we can strongly reduce the radiation on unwanted areas to a point where most of our patients end the therapy with almost minimal side effects.  The only real problems are the mucous membranes like in the mouth and esophagus as they are also very sensitive to radiation. They also heal fast, but in treatments in the neck area it is often the case that people are completely unable to eat or drink after 3 weeks of therapy. Thats why they normally get a PEG tube even before the start of the treatment to have an access for nutrients and fluid to the stomach.
A last very interesting aspect of radiation treatment is the application on non-cancerous cases. Especially on arthrosis in joints or heel spurs. By applying of a low dosage of 3 Gy over 5-6 daily treatments we can achieve strong pain reduction to complete pain freedom on any form of inflammation. Another application of radiation is a single 8 Gy application before the installation of artificial Joints for ossification prophylaxis.
I hope I could help with that insight and I apologize for possible things that are hard to understand.  English is not my first language.
A: That is the lethal does to stop human bodily function if radiated into a vital organ or the entire body. Cells die at a much lower level of grays, so a lethal does of gray's is not necessary to kill a tumor.
