An immunological outcome predictive score for head and neck carcinoma patients

To address the increasing demand for individualization of tumor therapy, a panel of immunological parameters was evaluated as potential early prognosticators for the outcome of treatment.

Thirty-one patients with advanced squamous cell carcinomas of the head and neck were treated either with a 2-course radiation treatment (60 to 70 Gy total dose) in combination with and subsequent to the administration of mitomycin C and 5-fluorouracil (radiochemotherapy) or with radiotherapy (2 patients) only. In 8 patients radio(chemo)therapy was preceded by surgical removal of the tumor. Before, during and after therapy, patients were immunophenotyped (in absolute numbers) and the respiratory burst function of granulocytes (polymorphonuclear [PMN] cells) was evaluated flow cytometrically.

Before treatment a reduction of T and B lymphocytes to 64% to 81% of the means of 101 controls (healthy volunteers and hematologic normal patients) was observed, absolute PMN counts were increased by 31%, whereas monocytes and natural killer cells were not influenced. The helper (TH)/suppressor-cytotoxic (Ts/c) T cell ratio was significantly elevated. The respiratory burst reaction of the majority (74%) of patients was normal. During therapy all lymphocyte populations declined further as did the PMN counts. Natural killer cells were not significantly influenced while absolute monocytes increased significantly beyond normal levels after initial depletions during each course of treatment. The helper/suppressor ratio was reduced to normal levels. Overall, treatment resulted in systemic effects at the level of leukocyte subpopulations and appeared to cause a shift of immunocompetence to a predominantly monocytic system. The deficiency in the humoral immune system could be correlated with the short survival time of most patients. Based on early effects of the treatment (after 10 Gy), an immunological outcome predictive score could be defined. A simple mathematical combination of the changes of B and Ts/c cells (after 10 Gy vs. 0 Gy) together with the respiratory burst reaction of PMN prior to treatment allowed retrospectively to classify correctly 90% (17/19) of patients as survivors (> 120 weeks) or early deaths (< 96 weeks; p < 0.01).

Single individual immunological data were not able to function as prognosticators for longer survival after therapy, but a combination of 3 parameters measured early during radio(chemo)therapy seems to allow the identification of "sensitive" patients. In how far the tumor disease per se and/or the immunological "sensitivity" are causes of death in these patients requires further clarification.