386 RETROGRESSIVE CHANGES 



trypsin, will not produce fat necrosis. The possibility remains that 

 pancreatic lipase is different from liver or serum lipase, and can 

 by itself cause fat necrosis; more probably, however, the production 

 of fat necrosis depends upon a double action, tr\'psin causing the 

 death of the cells, and lipase splitting the fats."^ The fatty 

 acids alone will not cause necrosis of fat-cells, and it was shown 

 that the first steps in the process consist of a necrosis of the surface 

 endothelium extending into the connective and fat tissue; this may 

 occur in a few minutes, while evidence of fat-splitting can be ob- 

 tained onlj^ after about three hours, and the splitting occurs only in 

 cells that have already become necrotic ; hence the fat-splitting is 

 not the cause of the necrosis, but occurs subsequent to the necrosis. 

 After about four hours a substance appears in the decomposed fat 

 that stains with hematoxylin, which is probably calcium. 



Fat necrosis may be produced by any means that will cause the 

 escape of pancreatic juice from the natural channels within the 

 gland. In human pathology it has followed trauma and acute in- 

 fection of the gland, and the blocking of the ampulla of Vater b^^ gall- 

 stones which permits the bile to back up into the pancreatic duct, 

 where it produces an acute inflammation of the pancreas (Opie)."* 

 Flexner °^ has shown that it is the bile salts that cause the inflamma- 

 tion, and also that this effect is decreased or prevented by the presence 

 of large amounts of colloids. ]\Iuch emphasis is laid by some au- 

 thors ^** upon the necessity of enterokinase passing up the ducts to 

 activate the trypsinogen (an idea first advanced by Starling and Ba}^- 

 Hss in 1902), but it should be remembered that there are kinases pres- 

 ent in leucocytes, and that kinases can develop in the pancreas itself 

 during autolysis, which can activate the trypsinogen ; hence the pres- 

 ence of entero-kmase is not essential for sufficient activation of tryp- 

 sinogen to account for pancreatitis and fat necrosis. Lattes °'^ believes 

 that fresh pancreatic juice, which digests tissues very' slowly, can pro- 

 duce typical fat necrosis but not the characteristic intoxication; this 

 results from the action of juice which has been activated by entero- 

 kinase, or by products of pancreatic autolysis M'hicli have a similar 

 effect. The kinases of leucocytes he found unable to activate pan- 



!i3 Wlion fat tissue dies in the liody from other causes, tlie lipase normally eon- 

 tain<'d within the fat tissue does not cause the changes seen in fat necrosis. It is 

 possible, therefore, that the coml)inin<r of newly split fatty acids hy the alkali of 

 tlie j)ancreatic juice is resjjonsihle for tlie formation of the larpe amount of 

 soajjs found in fat necrosis. Otheiw isc we mipht exjx'ct the lijjase to ])roduee only 

 an ('(piilihiium, and tliat, in tlie case of fat, s(^eins to exist wlien most of the 

 substance is neutral fat. In support of this idea I foiuid that stronix alkalies 

 injected into fat tissue sometimes caused changes very closely resembling areas 

 of fat necrosis in the early st,ages. 



»4Bull. dohns Hopkins 'llosp., 1001 (12), 182. 



o'-. Jour. Exp. Med., lOOG (8), 107. 



0" I'olva, ]\1itt. Grenz. Med. u. Chir., 1911 (24), 1; Rosenbach. Arch. klin. (hir., 

 1910 (93), 278. 



07 Virchow'a Arch., 1913 (211), 1. 



