CARCINOSTATIC AND CARCINOGENIC ACTIONS 773 



just tolerated dose, and usually a fraction of the animals die, so that the 

 effect is not specific. Various arsenicals have been tested for the inhibition 

 of tumors during the past 100 years, usually without remarkable results. 

 Beck (1953) found trivalent arsenicals to damage mouse lymphomas and 

 the Harding-Passey melanoma, but again at doses causing toxic effects. 

 The survival time of mice with Ehrlich ascites tumor is increased some 25% 

 by the administration of arsenite at 10 mg/kg/day, and the cell count in 

 the ascites fluid is reduced after 10 days (Creech et al., 1955); repetition of 

 this work led to a 13% extension of survival (Di Paolo, 1963). Atoxyl 

 seems to be much more effective, inasmuch as it reduces the ascites mass 

 from 4.3 to 0.94 ml in 11 days at a dose of 0.2 mg/mouse/day, which is 

 far below the LD50 (Osswald, 1956). Mitotic reduction in ascites smears 

 was noted, but at tlys dose there was no chromosomal fragmentation. The 

 organic arsenicals have not been adequately tested for general carcinostatic 

 activity. It is interesting that oxophenarsine has a wider range of action 

 than arsenite in HeLa cell cultures (Toplin, 1959). The ratio of the lethal 

 end-point concentration (to cause complete degeneration) to the cytotoxic 

 end-point concentration (to cause some degeneration in 5 days) for arsenite 

 is around 2.3, whereas for oxophenarsine it is 5. 



Whether the carcinostasis is due to an action on the nucleus or to some 

 metabolic inhibition is not known; however, some results point to the latter. 

 For example, the cytochrome oxidase of Sarcoma 37 is reduced by oxo- 

 phenarsine and can be almost completely eliminated at high doses (Leiter 

 et al., 1953). There was evidence of cellular damage and possibly the loss 

 of the enzyme is only secondary, especially as cytochrome oxidase itself 

 is not sensitive to the arsenicals. The respiration of tumor tissue is depressed 

 by the arsenicals but to no greater degree than that of normal tissues 

 (Dresel, 1926; Warren, 1943). It seems that the carcinostatic activity, as 

 far as we know, is not due to enzyme or metabolic inhibition but, inasmuch 

 as no serious attempts to correlate these actions have been made, one should 

 suspend judgment until the appropriate data are available. One cannot 

 avoid the conclusion that arsenicals are often specifically antimitotic, 

 although we do not understand why, and that carcinostasis is not the result 

 of a selective action on neoplastic tissue simply because it is neoplastic. 

 Arsenicals are not taken up to a greater extent by tumors than by corre- 

 sponding normal tissues. It is true that meningiomas may accumulate 

 somewhat more arsenite than brain, but this is not a valid comparison, and 

 carcinomas show no difference (Mealey et al., 1959). Ependymoblastomas 

 take up less atoxyl than liver or kidney, but more than muscle (Mego and 

 McQueen, 1963). In all probability the effects are more subtle than could 

 be observed on the systems studied. 



It is definitely established that the prolonged administration of arsen- 

 ite to man will lead to characteristic and sometimes dangerous changes in 



