DISCUSSION 183 



passage the tumours were very much less specific. We were not sure 

 whether this was immunoselection or not. I am rather incHned to 

 think it was not. The change was permanent. 



Feldman: This alteration which you have obtained without selection 

 is extremely interesting, particularly because we ourselves suspected in 

 our cases even if there is any selective pressure it is not immuno- 

 selection. 



Medawar: Could I make a suggestion arising out of McKhann's 

 work, and to some extent out of Lejeune's ? You might find that these 

 changed tumours would sensitize if you trypsinized the cells. This 

 seems to uncover antigenic potencies apparently not present before. 



G. Klein: I have a comment which relates to the acquisition of 

 homotransplantabihty after passage through newborn mice and perhaps 

 also passage through chimeras. We have been working with some F^ 

 hybrid tumours, and both our group and Mitchison have found that if 

 you have an F^ hybrid tumour heterozygous for H-2, which can be 

 symbolized, for the sake of simplicity by AS, then out of these tumours 

 one can select variants compatible with A and variants compatible with 

 S. If one takes one of these variants, for instance the one compatible 

 with A, and tries to make it also compatible with S by passaging it 

 through newborn hosts of the S type, sometimes it becomes homo- 

 transplantable and non-specific. But while the original variant form- 

 ation and the development of the compatibility with the A parent 

 involved the complete loss of the specific antigens of the S origin, to 

 such an extent that they were no longer detectable by haemagglutin- 

 ation or by cytotoxic methods, the development of homotransplant- 

 abihty after passage through the newborn did not involve any further 

 detectable loss of antigens and the H-2A system was still fully main- 

 tained. Therefore this second change is something entirely different 

 from the first ; it is not a further loss of antigenicity. I am wondering 

 if there isn't a phenomenon here that still eludes us, and that cannot be 

 explained simply as loss of antigenicity. 



Feldman: You are right. 



G. Klein: You mentioned in your paper the effect of cortisone on the 

 formation of metastases. I recently had occasion to review the hterature 

 on this point, and it appears that all authors who have shown that 

 cortisone treatment has a promoting effect on metastasis formation 



