Surely this would only be of benefit to the recipient if it could be ensured that only malignant epitopes / peptides from the tumour would be used to sensitise the dendrocytes. The patients own dendrocytes automatically recognise what is self (and do not react) and what is non-self (malignantly transformed cells), becoming sensitised to that foreign tissue.
It is inevitable that in any biopsy/excision of malignant tissue, normal cells and tissue is included with the abnormal. If dendrocytes from a non-identical "donor" are used the dendrocytes would be sensitised not only to malignant (therefore "foreign") epitopes / peptides but also to the patient's normal tissue which will be recognise as being non-self (ie foreign to the dendrocytes). Injecting those sensitised dendrocytes into the recipient will result in what is essentially a graft-vs-host reaction, potentially causing multi-organ damage should the injected clones of dendrocytes become established in the recipients immune system. .