This is all an issue of semantics. Oral cancers comprise more than 70% of all head and neck cancers. Broadly speaking, within oral cancers you have those in the anterior of the mouth called oral, those in the back of the mouth and the top of the throat and soft palate referred top as oropharyngeal cancers, those in the back of the mouth, soft palate and involving structures above the back of the throat as well are nasopharyngeal cancers, and those of the throat itself. Laryngeal cancer is usually referred to as a cancer unto itself, though the risk factors for it are the same as oral cancers. There are about 12,000 laryngeal cancers diagnosed each year, so if you added those to the categories already mentioned which have about 30,000 diagnoses, you would get about 42,000 per year. Then to further muddy the name-based issue, this cancer produces second primaries, as well as metastasis from the original primary cancer. Alan King was originally a maxillary oral cancer patient. When he died just a few months ago, 7 years out from his original oral cancer (which he had been declared disease free from) it was from lung cancer. However, this pulmonary cancer was a second primary from his oral cancer... so the question could be raised from which did he die? Semantics. To the best of my information Mr. Harrison originally had a tobacco induced oral cancer for which he was treated, and declared cancer free. It involved his throat like many of us as well as an oropharyngeal primary. Unfortunately he returned to smoking after his treatment, a testimony to the addiction of tobacco more than a weakness of the man, and developed a lung cancer, which then metastasized to a brain cancer. This would be considered a second primary even though the continued smoking was no doubt a factor. Oral cancers kill people not from the disease in their mouths, but from the metastasis of that disease to vital organs. Reporters have a poor understanding of this mechanism, so they will report that George Harrison died of a brain tumor, which is true but does not reflect the entire etiology of the process that created the disease.