Like all mammalian cells, normal or malignant, the primary source of their energy (and, thereby their continuing existence) is glucose, a monosaccharide derived from the higher sugars, the disaccharides. These comprise sucrose (the sweet stuff you put in your food or drinks), lactose ("milk sugar") and maltose (the result of the digestion of starches and other carbohydrates).
In a PET (Positron Emission Tomography) scan the image is produced by the measurement and localisation of positrons (short-lived sub-atomic particle akin to the electron but with a positive charge instead of the negative charge that the electron has). These are spontaneously emitted from a "radioactive tracer" given to the patient as the first step in the PET scan. The tracer is fludeoxyglucose F18 ("FDG") and it is taken up (just like glucose) by any cell which is metabolising (ie that is alive). Malignant cells are more active metabolically than normal cells because their metabolic control mechanisms have been lost, resulting in runaway activity. The higher the cells' metabolic rate, the more glucose (and FDG) it takes up so that when the glucose source is predominantly FDG the greater the numbers of positrons that will be emitted from within malignant cells. Concentrations of malignant cells can thus be identified, localised and quantified.
If glucose is not restricted before the PET scan all cells will be relatively replete with glucose with the result that insufficient FDG is taken up by the cells. This results in low levels of positron emission so that the scanner is unable to identify "hot spots" indicating the presence of overactive malignant cells. To avoid this, sugars are restricted before the PET scan to "slightly starve" the cells of glucose so that when the FDG is given a good amount of the FDG is taken up by the malignant cells. The concentration of positron emission can then be picked up in the PET scan.
