Model of extreme hypoglycemia in dogs made ketotic with (R,S)-1,3-butanediol acetoacetate esters.

The rationale behind this study is that controlled starvation of poorly differentiated (anaplastic) fast-growing tumor cells, but not host cells, might be possible in vivo. The energy metabolism of anaplastic tumor cells, but not host cells, is largely dependent on carbohydrate metabolism at all times. Therefore depleting plasma of carbohydrate fuels could place these tumor cells at a significant metabolic disadvantage. Hence an animal model was developed in which all cells would be required to oxidize fatty acids, ketoacids, and/or 1,3-butanediol to satisfy their energy needs. To achieve this aim, one would need ketosis, severe hypoglycemia, and low lactatemia. Anesthetized normal dogs were infused with somatostatin and a mixture of (R,S)-1,3-butanediol monoacetoacetate and (R,S)-1,3-butanediol diacetoacetate; these latter compounds are nonionized precursors of ketoacids. They were infused at 90% of the dog's caloric requirement. After establishment of a moderate ketosis (2-3 mM) over < 100 min, a severe degree of hypoglycemia (close to 0.5 mM) without rebound and without hyperlactatemia was induced by infusing insulin and dichloroacetate. Tracer kinetic measurements showed 1) a 20% decrease in the rate of appearance of glucose, 2) 50 and 62% increases in glycerol and nonesterified fatty acid rates of appearance, reflecting stimulation of lipolysis, and 3) no change in the rate of glutamine appearance. We suggest that this model may prove useful for selectively starving those cancer cells that are unable to utilize fat-derived fuels while preserving nutrient supply to vital organs.