340 On Respiratory Impairment in Cancer Cells 



tissues, of certain tissues that had first been severely predamaged by treatment with 

 Cyanide and anaerobiosis ( !), thus lowering notably the calculated average Qq 2 for 



this group. 



Weinhouse's Statistical approach to the problem of possiblego, impairment in 

 cancer cells requires the use of the foregoing adjustments just as surely as Cancer 

 mortality and morbidity statistics require adjustment, in the absence of complete 

 and perfect data. The types of adjustments listed all tend, without exception, to 

 raise theQo 2 values of normal tissues relative to tumor tissues, with the end result 

 that, as one approaches truly physiological comparisons, one obtains average Q . 2 

 values for normal growing and nongrowing tissues that are "by and large" two 

 to three times higher than those for tumors, with reduction in spread of individual 



values. 



The first two types of adjustment listed are probably the most important 

 quantitatively and will be briefly discussed by way of illustration. It is well estab- 

 lished from the data extensively summarized and formalized, for example, by 

 Krebs* and Adolph 9 , that tissue respiratory rate, basal heat production, Venti- 

 lation rate, and a host of other vital properties are marked, linear log-log functions 

 of species body weight. The group of normal tissues selected by Weinhouse were 

 from the rat, rabbit, dog, and man, but none from mice; the tumor tissue types 

 selected were preponderantly from mice, the normal tissues of which have average 

 rates of respiration and basal heat production far above those of the larger species, 

 the basal heat productions being of the rough relative order 160, 100, 60, 35, and 

 25 in mice, rat, rabbit, dog, and man, respectively. Any reasonable (but impera- 

 tive !) adjustment of average respiratory rate for species difference alone would 

 place the tumor group well below that of the normal group, with an adjusted 

 average Qo* value of — 6 to —7 (instead of — 11. 8, 10 . This value agrees with the 

 grand average value that one obtains directly from all 30 widely varying types of 

 malignant, nonmouse tumors (rat, chicken, man) listed in "Burk's extensive 

 tables" VII and VIII 7 , many of which were excluded from, or improperly weighted 

 in, the selected and Condensed summary prepared by Weinhouse without due 

 regard to the effect of species. This single error in Weinhouse's statistics, by itself, 

 reverses the order of average Qa 2 values among his three groups, so that non- 

 growing growing tumor. 



The second factor, of equal quantitative importance in tending to widen further 

 the relative difference between normal and tumor tissues, is that in virtually all of 

 the experiments selected by Weinhouse from Burk's tables, simple saline media 

 were e mployed, instead of sera or similar body fluids that are much more physio- 

 logical and therefore much more pertinent for the Solution of the problem at hand. 

 Extensive modern studies show, even better than earlier studies, that the use of 

 sera instead of saline media increases the average Qo 2 values of normal tissues much 

 more than that of tumors— by a relative factor up to 2 and more. This differential 

 response is due partly, though not wholly, to the greater response of normal tissue 

 to respiratory Substrates occurring naturally in sera but not added to the early 

 saline media employed in the experiments selected by Weinhouse. This differential 

 response finds equivalent and additional expression in the relatively small succin- 



