Introduction: Epidemiologically, thyroid gland tumors are lesions of the highest importance among endocrine tumors in humans. Although the results of surgical treatment of the highly differentiated (follicular and papillary) tumors seem to be satisfactory, treatment of the poorly differentiated (medullary and anaplastic) tumor still demands clinical and basic investigations. In this study the authors sought to evaluate clinical and molecular factors that could contribute to preoperative detection of more advanced thyroid cancers (i.e., those that exhibit extrathyroid spread and lymph node invasion).
Methods: A total of 27 patients operated on for thyroid cancer were evaluated according to age, sex, time from the onset of the disease, cytogenetic changes, and loss of heterozygosity (LOH) in 14 microsatellite markers. The output variables were defined according to postoperative findings and the TNM 2002 score. The T1-2 N0 M0 cases were defined as local malignancy (LM); and T3-4 any N any M, any T N1 any M, or any T any N M1 were considered advanced malignancy (AM). The control groups consisted of 25 patients with multinodular goiter (MNG) and 32 patients with follicular adenoma (FA). In all cases, clinical and molecular data similar to those listed above were collected, excluding staging and follow-up information.
Results: There was no predominant specific type of chromosomal aberration observed and no marker lost in more than five patients (18%). The logistic regression identified three input variables as contributing significantly to the dichotomized outcome measure (LM vs. AM): LOH in any of the examined loci, age of the patient at the presentation, and the sex of the patient. Furthermore, discriminant analysis revealed four input variables differentiating among TC, FA, and MNG patients. Based on the multivariate analysis results, two numeric prognostic scales were fashioned: LAST-1, a scale applicable to differentiation of thyroid cancers at different degrees of clinical advancement; and LAST-2, a scale applicable to differentiation of any thyroid lumps.
Conclusions: It was concluded that LOH and the age and sex of the patients can provide sufficient data to predict thyroid cancer with a high degree of clinical advancement. LAST-1 scale is a reliable tool for identifying these patients. The LAST-2 scale gives supportive information about the character of thyroid lumps, distinguishing TC from MNG and FA.