Purpose: To determine whether the immunohistochemical expression of proliferation-associated antigens (proliferating cell nuclear antigen, MIB1) and the nuclear p53 reactivity in addition to total tumor volume, nodal CT density and T and N category are predictive for overall survival and locoregional tumor control in patients with squamous cell carcinoma of the head and neck region.
Materials and methods: Between October 1989 and September 1993, 87 patients with biopsy proven head and neck cancer were randomly allocated to receive radiation alone or simultaneous radiation and chemotherapy as part of a multicenter trial with a total of 298 randomized patients. There were only inoperable lesions in UICC (1992) stage III (8%) and IV (92%). Radiotherapy was delivered with 180 cGy twice daily up to a total dose of 7020 cGy in 51 days. Three cycles of 2340 cGy each were separated by a rest period of 11 days. Chemotherapy consisted of cis-DDP, 5-fluorouracil and leucovorin and was repeated on days 22 and 44. Routinely-processed paraffin-embedded sections were stained using monoclonal antibodies for detection of proliferation-associated antigens (MIB1 and PCNA) and p53 oncoprotein to determine the labeling index (LI). In addition, the total tumor volume and the percentage of necrosis were measured using CT data. The median follow-up was 3.9 years (range 1.9-5.0 years).
Results: The overall survival and locoregional control for all 87 patients were 34 and 39% at 3 years, respectively. The addition of chemotherapy resulted in a better overall survival (27 versus 47%, P = 0.03) but did not influence locoregional control (31 versus 47%, P = 0.08). In univariate analysis, nodal CT density (P < 0.0001), total tumor volume (P < 0.0001), age (P = 0.001) and the MIB1-LI (P = 0.04) had a significant impact on overall survival. However, in the final Cox model only the nodal CT density (P = 0.0003) and age (P = 0.05) were independent prognostic factors for survival and only the nodal CT density (P = 0.0006) was an independent prognostic factor for locoregional control. The expression of the p53 oncoprotein was not found to have a clear predictive value.
Conclusion: Nodal CT density, total tumor volume and age will remain the relevant prognostic factors in stage III/IV head and neck cancer.