While no expert in the field, I will just try to provide some general reflections regarding the use of tumor cell lysates when studying tumor-directed T cell responses. Hopefully, someone with more direct experience of working with tumor antigens can give further advice and insight.
Although most studies of tumor-directed T cells have employed purified protein or peptides, some people have also used more crude preparations of antigen such as tumor cell lysates of the type you describe. These have been used both for vaccination and for monitoring the resulting immune response often in the context of preloaded autologous dendritic cells.
A problem with using cell lysates rather than more characterized proteins or peptides is that T cells require a certain concentration of antigen to be efficiently triggered. However, the concentration of individual antigens in a cell lysate is rarely known and may also differ between preparations. Due to this, titration of the tumor cell lysate with one or more positive donors may be recommended in order to define a suitable concentration. Typically, concentrations of a few hundred µg/ml have been used but in some studies the amount of added lysate has instead been related to the number of cells from which the lysate was prepared using, for instance, a 1 to 1 mixture of PBMC and tumor cells.
With regard to the incubation time in the IFN-g ELISpot assay, it might be recommended to extend this from the overnight incubation often used with peptides to two nights in order to accommodate for the more complex antigen processing and presentation. The use of separately generated autologous dendritic cells may also help to optimize conditions.
Best regards / Staffan