December t, i , 



1908J 



NA TURE 



203 



progress to a uniform histological structure nor a gradual 

 advarucc lo the exhibition of uniform biological behaviour, 

 nor acquisition of a uniforiii rate of growth. The trans- 

 formation of normal into cancer cells really covers a 

 scale of changes which do not pass into one another. 

 I'erinanent features arc stamped upon cancer cells at the 

 outset. There is no tr.msition from one degree of the 

 cancerous change lo another. 



In the transplantation of a tumour into a new host 

 success or failure is determined primarily by two factors. 

 These are the qualities of the tumour cells and the nature 

 of the " soil " the new animal offers. During continued 

 propagation the cells of the tumours of a single organ, 

 e.g. the mamma, exhibit other differences corresponding 

 tu those mentioned above with reference to the " support- 

 ing " scaffolding, and logi'ther with them pointing still 

 more strongly to primary qualitative differences in the 

 (•■lis of different rumours, .\lthough cancer occurs spon- 

 i.uK'ously mostly in old .ininiaU, young animals are more 

 suitable for growth. The introduction of a minute particle 

 of cancerous tissue into a normal animal leads to all the 

 ciijisequences which accompany the growth of a spon- 

 taneous tumour. Thus the adequacy of the assumption 

 with regard to man, that the origin of cancer is primarily 

 circumscribed, is demonstrated. .A consideration of all the 

 nsults proves that the genesis of a tumour and the growth 

 of a tumour are two different things. 



'I'he " soil " which different races of mice offer, as it 

 were, for the growth of cancer varies naturally in suita- 

 I'ility ; but tumours can gradually or rapidly adapt them- 

 -ilves to a soil which was unsuitable, e.g. when a Danish 

 iiimour was first tr.insiil.inied in England it grew in only 

 5 per cent, of the mice inoculated, but later the success 

 rose to 90 per cent. There are natural constitutional 

 conditions which are favourable, and others which arc un- 

 favourable, to the growth of a tumour. The unfavourable 

 (onditions act as sieves, permitting certain kinds of cells 

 lo pass, and once they have passed they can multiply 

 beyond our powers of measurement. 



The '■ soil " can, however, also be modified experi- 

 Tuentally. It can be made absolutely unsuitable for growth 

 or rendered more suitable than normal. Mice and rats 

 I an be rendered unsuiiable for growth only by vaccinating 

 I hem with malignant new growths of their own species 

 md by vaccinating with normal tissues of their own 

 ^()ecies. In the latter case the degree of " resistance " 

 iiDrmal tissues produce directly corresponds to the close- 

 ness of the relationship between the normal tissue 

 vaccinated and the tumour subsequently inoculated, e.g. 

 ^Uin protects best against sUin cancer. These facts refer 

 us back .again to the limitations to the transplantation of 

 tumours, and together with them demonstrate the reten- 

 lion by malignant new growths, not only of the tissue 

 1 liaracters of a species, but also of the biochemical as 

 well as of the histological characters distinctive of the 

 v.veral species. A sarcoma of a rat or cat, vaccinated into 

 .1 mouse, lacks the power of protecting it against sub- 

 ^'•quent inoculation of a mouse sarcoma; this fact shows, 

 as clearly as the method permits, the absence of any 

 ' straneous agent common to the growths of these different 

 species. The growths of different species of animal 

 resemble one another just as much, and differ just as 

 much, as their respective organs and tissues do. -As 

 differences exist in certain properties of tumours already 

 .:iluded to above, so corresponding other differences arc 

 I' vealed by the extent to which tumours, when vaccinated, 

 i iduce protection against one another. A tumour does not 

 .iccinate so well against other tumours as it does against 

 I -elf or against those of its own kind. \ lesser degree 

 • >( protection which one kind of inouse-tumour induces 

 against other kinds is due, prob.ibly, not to cancer-tissue 

 a- such, but to its properties quit mouse-tissue. 



.\nimals which are absolutelv protected against inocula- 

 tion do not yield a serum which, when introduced into 

 new animals, has a power of protecting them against in- 

 oculation, still less is there any evidence of immune sera 

 having a power to cure aniiials of tumours already grow- 

 ing. Highly immune mothers do not transfer immunity to 

 their offspring as do animals immune to diphtheria or other 

 poison of infective disease. Indeed, the mechanism of the 

 protection which can be induced against cancer is of a 



NO. 2044, VOL 



kind quite unknown before. Most painstaking observations 

 have been necessary to penetrate somewhat into its nature. 

 Artificially protected animals do not supply the cancer 

 cell with the peculiar scaflolding of supporting tissues it 

 requires in order to grow into a tumour. It dies .because 

 it cannot grow into an organised tissue, and hence cannot 

 nourish itself ; being damaged, it falls a prey to the natural 

 guardians — the phagocytes — of the body. The process is 

 tlie same whether vaccination has been made with cancer 

 or with normal tissue. The way in which this protection 

 becomes general in the body fluids or tissues has not yet 

 been fully ascertained; nevertheless, so far as it is known, 

 it helps to elucidate the spontaneous healing of primary 

 and .secondary growths in man, and its further study gives 

 promise of our being able ultimately to enhance the powcis 

 of resistance of the body to a degree which will prevent 

 the dissemination of a primary growth. 



Before so much can be attained there are many difficul- 

 ties to be overcome, not the least of which is the dis- 

 covery of the fact mentioned above, that the soil may be 

 rendered more than normally suitable for the growth of 

 cancer. Hypersensitivencss can be induced by many 

 different agencies ; indeed, as contrasted with the induc- 

 tion of protection, it is not specifically induced.' The 

 growth of one. tumour does at times I'lake the "soil" 

 of an animal more favourable for the growth of a second 

 tumour, and therefore, presumably, for dissemination. It 

 is much more difficult to protect an animal already bear- 

 ing a tumour against the transplantation of a second 

 tumour than it is to protect an animal which has not 

 already got one. 



.Aniiiials spontaneously attacked with cancer make 

 efforts, which are sometimes successful, to cure themselves 

 both of prfmarv and of disseminated growths, e.g. in the 

 vessels of the' lungs. There is no longer room for 

 scepticism regarding the statements which have been made 

 from time to time of similar occurrences in man. The 

 process of spontaneous healing is much more common in 

 animals bearing transplanted tumours. In their case it 

 can be studied in great detail, and it has been found to 

 follow the same course as in man. A weighty factor 

 contributing lo its occurrence resides in the properties of 

 the cancer cells themselves, for it has been discovered 

 that thev multiply with unequal rapidity at different times. 

 They alternate i-egularlv between positive and negative 

 phases of growth. Thev are much more vulnerable to 

 attack in the negative phase, e.g. through the heightened 

 unsuitabilitv or resistance which can be induced in the soil 

 as described above. The further study of the relations 

 obtaining hero will ultimately assist us to prevent a primary 

 tumour from disseminating and establishing offshoots in 

 remote parts of the body. . 



.A startling phenomenon has been stu?iibled upon during 

 the artificial propagation of epithelial malignant new- 

 growths (carcinomata). In the course of time some of 

 these tumours have been replaced by connective tissue new 

 growths (sarcomata). There is no question of the con- 

 version of epithelial into connective tissue cells. All the 

 facts point to the acquisition of canceroiis properties by 

 what were previouslv normal connective tissues, viz. cells 

 of the supporting scaffolding or "stroma." It appears 

 probable that in this wav malignant new growths have 

 been produced for the first time experimentally. The 

 development of sarcoma in this w-ay occurs in circumstances 

 throwing much light upon whv cancer in man is so fre- 

 quently associated w-ith chronic irritation, as referred to 

 above,' and resulting continuous or intermittent attempts 

 at regeneration and repair in man. Together with other 

 facts, notablv the differences in incidence of cancer_ in 

 different race's of mankind as determined by the applica- 

 tion of irritants to different parts of the body, it gives 

 the colli) dc grace to the gener.alisation of the idea that 

 cancer is of congenital origin. 



Many new facts recorded above are of fundamental 

 importance in enabling us better to comprehend the nature 

 of cancer. Two factors have been proved to be of pnme 

 importance in its development; one is the alteration within 



t Ttie variety of the agents which render an animal hypersenMtive for the 

 erawlh of cancer acquires added interest when regarded in assori'»tion wit*! 

 the vaiety of causes of chronic irritation related to the development of 

 cancer in mankind, as referred to above. 



