July 28, 19 10] 



NATURE 



127 



so great and of such constancy that strains propagated 

 separately from the same mother-material would not be 

 suspected to have any relation to one another if submitted 

 1:0 one ignorant of their life-history. 



In former years we have pointed out that an increase 

 in the rate of growth, or in the percentage of successful 

 inoculations, does not necessarily imply a fundamental 

 biological alteration finding expression in an accelerated 

 rate of proliferation of the tumour cells, but may be ex- 

 plained by the selection of particular cells adapted to the 

 conditions of growth, and, consequently, the survival and 

 proliferation of a larger number of such cells. That is to 

 say, these two phenomena may be explained by an increase 

 in the dose of the cells able to grow. The evidence for 

 the acquirement of new properties by tumour cells is very 

 much stronger when one observes the occurrence of morpho- 

 iogical alterations which become of relative constancy, such 

 as the disappearance — or latency — of their typical character- 

 istics in the case of squamous-celled carcinoma, the dis- 

 appearance of acinous structure in the case of glandular 

 carcinoma, the derivation from cubical epithelium of 

 epithelial cells which, if their previous history had not been 

 known, could not have been distinguished from those of a 

 spindle-celled sarcoma. In other cases, the change is made 

 manifest by the alterations taking place in the supporting 

 connective tissue and blood-vessels, so that tumours which 

 at one time exhibited dilated blood-vessels lose this 

 character. Biological alterations occur without evident 

 morphological expression, e.g. some tumours at the com- 

 mencement of propagation, after an initial exuberant 

 growth, disappear in a large proportion of cases, whereas 

 after the propagation is prolonged, a large percentage of 

 the implantations grow progressively.' The opposite pheno- 

 menon may also be observed, and tumour strains which 

 grew progressively at the outset of propagation may later 

 be found very liable to disappearance. A tumour which 

 grows well only by the implantation of intact grafts, i.e. if 

 the tissue structure is preserved, can be adapted to trans- 

 plantation as a cell emulsion, and again brought back to 

 its original condition. 



Of the twenty-nine tumours of the mamma that have 

 been propagated in the laboratory for more than two years, 

 as many as sixteen have shown departures from the 

 features they exhibited at the outset, these departures 

 affecting the degree and nature of the histological differ- 

 entiation, the percentage of successful inoculations, the 

 rate of growth of the resulting tumours, the relative pro- 

 portions of progressivelv growing tumours and of tumours 

 which undergo spontaneous absorption after transitory 

 growth, the susceptibility of the tumour to method of 

 transplantation, to dose, to race, to age, and to the in- 

 fluence of indued immunity. 



Thirteen tumours have shown a relative constancy of 

 their structural and biological characters. 



Of the sixteen variable tumours, nine have varied from 

 the primary condition in both respects. Two have shown 

 biological variations witliout histological change, and five 

 have altered in microscopical characters without noticeable 

 modification of their biological behaviour. On the whole, 

 therefore, histological character is less constant than bio- 

 logical behaviour. 



The relative constancy, but still more the variability 

 which the tumour cells exhibit during propagation, throws 

 indirect light of the most suggestive kind upon the nature 

 and the manner of the development of cancer. The 

 variabilitv in a constant environment, during propagation, 

 allows one to infer that corresponding variations may take 

 place while the cells are under the influence of the par- 

 ticular environment provided by the animal in which the 

 tumour developed spontaneously. The environment of the 

 cell will depend on the individuality of the animal, and. 

 with the progress of life, distinctions between one animal 

 and another may become more and more marked. This 

 inference accords with what h.as been said above on the 

 ease with which auto-transplant.ation is effected and the 

 difficulty with which transplantation can be effected to 

 another individual, and therefore also with the fact that 

 all cancerous mice do not exhibit an equally suitable soil 

 for tumours in general. * 



These soontaneous variations of the parenchyma cells of 

 tumours during nropagation suggest that we have here a 

 NO. 2126. VOL. 84] 



repetition, in a minor degree, of the cellular processes 

 responsible for the primary transformation of non- 

 cancerous into cancerous tissue ; just as cellular changes 

 occurring during propagation may transform within a brief 

 space of time an acinous growth into a solid one, or a 

 slow-growing tumour into one rapidly proliferating, so in 

 the tissues prior to the development of a malignant new 

 growth the responsive proliferation of cells may pass into 

 the progressive, independent proliferation of cancer. 



Experimental Sarcoma. 



In this connection it may also be well to refer again to 

 the production of sarcoma under experimental conditions 

 from what have been the non-malignant connective tissues 

 of carcinoma. Not the least significant aspect of the 

 origin of sarcoma by the transformation of the stroma of 

 transplantable carcinomata is the rarity of its occurrence. 

 Two only of our strains have exhibited it, and the con- 

 clusion seems warranted that in these cases the parenchyma 

 is possessed of peculiar properties. In one of our strains 

 the change occurred only in a small number of animals, 

 and the whole process, from the first indications of sarco- 

 matous changes in the stroma to the substitution of the 

 carcinomatous elements by pure sarcoma, took place slowly, 

 and was onlv completed after several successive trans- 

 plantations, in the other strain, the transformation was 

 much more frequent, took place more rapidly, and the 

 disappearance of the carcinomatous element may be com- 

 plete in one transference. In spite of these differences, the 

 parallelism between the histological pictures in the two 

 strains is extremely close, and leaves no doubt of the 

 essential similarity of the processes involved. The stimulus 

 exerted by the carcinoma cells on the stroma must be 

 different in these two strains from that exemplified by 

 the other transplantable tumours, otherwise every _ trans- 

 plantable carcinoma should end in sarcoma, as it has, 

 indeed, been asserted they might do. A fairly long dura- 

 tion of the stimulus exerted by the carcinoma cells without 

 cessation of their proliferation seems to be necessary, and 

 the first steps of the process are always localised in an 

 extremelv minute area of what are often large tumours. 

 The parallel to the circumscribed origin of squamous 

 epitheliomata arising in areas subjected to chronic irrita- 

 tion in man (chimney-sweeps' cancer, paraffin cancer, 

 Kangri cancer) does not require to be insisted on further, 

 since it has been emphasised in previous years. 



Immunisation. 



It is now possible, under given experimental conditions, 

 to prevent a secondary transplantation, i.e. _ artificial 

 metastasis, taking place for certain tumour-strains. _ This 

 result has been obtained by inserting between the primary 

 and secondary transplantations an inoculation of a very 

 rapidly growing tumour showing only transitory growth ,_ as 

 the following sample experi.nent shows. Of twelve mice, 

 already bearing progressively growing tumours and treated 

 in .the manner described, the secondary inoculation was 

 successful in three only, and then the tumours were very 

 much smaller than in the control consisting of thirteen 

 mice, of which ten developed new progressively growing 

 tumours on secondary inoculation. A similar result can 

 be obtained by the implantation of tumours growing much 

 more slowly and liable to spontaneous absorption, as well 

 as by an inoculation of normal mouse-tissue. Bv similar 

 methods the growth of the primary transplanted tumour 

 may be greatly hindered, can be brought to a standstill 

 and the animal cured, in circumstances under which the 

 disease would certainly have progressed, and where the 

 possibilitv of the occurrence of spontaneous cure can almost 

 certainly" be excluded. Thus t'he control of transplanted 

 cancer has been brought within the region of probability. 



These achievements must not be confounded with 

 successful vaccination against snontaneous cancer arising, 

 or against infectious disease, .\nimals perfectly protected 

 against the repeated inoculation of cancer may develoo 

 tumours of their own — an observation often confirmed. 

 Still more emphatically do we warn against applying to 

 the human subject the methods which, after long perse- 

 verance, have enabled us to arrest the growth, and even 

 to cure, animals of transplanted tumours that were well 



