Despite major advances in the diagnosis and treatment of human lymphomas, there remains considerable heterogeneity in clinical behavior within traditional subgroups. Application of newer diagnostic modalities, including monoclonal antibodies and nucleic acid probes, has not yet had a major impact on improving the predictability of clinical behavior for many patients. This project will address three major aspects of the tumor host reaction in vivo: (1) expression of target cell recognition structures LFA-1, HLA-Class I, and HLA-Class II in lymphoma subgroups, (2) enumeration of effector cell subtypes in lymphoma subgroups, (3) level of expression of proliferation and "activation" associated markers in subtypes of lymphoma. Ongoing studies will rely heavily on frozen section immunohistochemical staining combined with estimation and quantitation methods for enumerating types and numbers of immune cells. Automated cell analysis systems will be employed as they are developed. A better understanding of important tumor and host features may aid in the selection of current treatment modalities, provide the rational for delay treatment and may prove useful in devising experimental treatment strategies. Long term goals are to devise a more reproducible, clinically predictive classification of lymphomas. Such an in vivo evaluation of tumor host relationship should provide valuable insights in lymphoma immunobiology.