Immunobiologic staging of patients with carcinoma of the head and neck.

For decades immunologic deficiencies have been documented in patients with carcinoma, and many investigators have attempted to utilize this information in determining prognosis more accurately. Determining prognosis more accurately would be most helpful with Stage III patients, since at present there are no useful guidelines as to which Stage III patients would benefit from aggressive surgery and/or radiation therapy and which would not. Assays of cellular immunity--such as skin testing, determining peripheral T-lymphocyte counts, and assessing lymphocyte reactivity--have achieved extremely limited clinical application. These assays are too expensive and/or too difficult to perform in the routine clinical setting. Total lymphocyte counts, however, are readily available in the clinical setting and can be helpful in determining prognosis. Other readily available clinical information, such as age and pretreatment serum immunoglobulin levels, can also be of value in determining prognosis. In this study, logistic regression was utilized to determine the prognostic implications of stage, age, and pretreatment total lymphocyte counts in 132 patients with carcinoma of the head and neck. Although each variable determined prognosis accurately less than 70% of the time, the stage, age, lymphocyte prognostic index (SALPI) accurately predicted outcome at 3 year follow-up 80% of the time (Pr. less than .0001). Discriminant function analysis was utilized to determine the prognostic implications of pretreatment serum immunoglobulins in 243 patients with head and neck carcinoma. A serum immunoglobulin prognostic index (SIPI) was derived which was based on the observations that elevated levels of immunoglobulin A (IgA) indicated a poor prognosis, while elevated levels of immunoglobulin E (IgE) and immunoglobulin D (IgD) indicated a favorable prognosis (Pr. less than .009). The SIPI was not as accurate as the SALPI in determining prognosis; however, when the indices were used simultaneously, concurring indices predicted the outcome of Stage III patients correctly 86% of the time.