The Null Genotype of Glutathione S-Transferase M1 and T1 Locus Increases the Risk for Thyroid Cancer¹

The majority of human tumors is considered to be a result of the interaction between environmental factors and personal genetic susceptibility (1, 2). However, people vary greatly in their likelihood of developing cancer in response to natural hazards. Individual differences in susceptibility to carcinogens play an essential role in the development of sporadic cancer. The biochemical basis for this susceptibility is related to genetic polymorphisms that normally occur in the general population regarding genes involved in predisposition to a specific cancer, in the metabolic activation or detoxification of environmental genotoxins, and in controlling DNA repair or cellular damage (3, 4, 5).

The etiology of thyroid cancer is markedly uncertain. Exposure to ionizing radiation, especially in childhood, remains the only factor clearly associated with benign and malignant thyroid tumors in humans (6). However, there is strong epidemiological evidence pointing toward the involvement of geographic, ethnic, and dietary factors in the risk of sporadic thyroid cancer (7). A variety of drugs, pesticides, goitrogenic xenobiotics, and chemicals have been shown to increase the incidence of thyroid tumors in rodents (8, 9, 10). However, chemicals have seldom been associated with human thyroid cancer, in contrast to lung, bladder, and many other cancers. No increase in the risk of human thyroid cancer has ever been consistently observed with any drug (6, 7, 11). On the contrary, a number of studies have reported a reduced risk of thyroid cancer in women who smoke cigarettes (12, 13).

The GST³ system consists of a large multigenic group of detoxifying enzymes, the activity of which, catalyzing the conjugation of toxic and mutagenic compounds with glutathione, is essential for cell protection (14). Conversely, this important service may be disadvantageous during chemotherapy where GSTs have been associated with multidrug resistance of tumor cells (15, 16). At present, five classes of isoenzymes have been identified: alpha, mu, pi, sigma, and theta. The genes that encode the GST enzyme system are polymorphic in the general population (14, 15, 16, 17). Both the GSTM1 and GSTT1 genes have a null variant allele in which the entire gene is absent. Persons with homozygous deletions of either the GSTM1 or the GSTT1 locus have no functional activity of the respective enzyme. Epidemiological studies suggest that individuals who are homozygous null have an increased risk of developing cancer at a number of sites like lung, bladder, colon, and breast (18).

The primary objective of this study was to test the hypothesis that individuals with an inherited homozygous deletion of the GSTT1 and/or the GSTM1 genes are at an increased risk of thyroid cancer. We conducted a prospective case control study in which we compared the proportion of GSTT1 and GSTM1 null genotypes between a group of patients with benign and malignant thyroid tumors and a control group. Heterogeneity of risk according to clinical and morphological subtypes of thyroid tumors and their correspondent genotype was also explored.

The study was approved by the Ethics Committee of the HC-FCM/UNICAMP, and informed written consent was obtained from a total of 416 individuals from our region, considered to have a normal iodine intake. Because of the highly heterogeneous ethnic composition of the Brazilian population, we included a large control group of 300 individuals (99 males and 201 females, 16–78 years old, 35 ± 23 years old) selected from the general population. To obtain a comparable control group with respect to gender proportion and the range of ages, we selected two to three women for every man who presented himself to donate blood, because thyroid cancer occurs more frequently in women than in men. Data on lifetime occupational history, smoking history, general health conditions, previous diseases, and other anamnestic data were obtained through interviews. Individuals with history of previous thyroid disease, exposure to radiation, and antecedents of malignancy were excluded. There were 252 healthy blood donors (78 males and 174 females, 18–60 years old, 31 ± 21 years old), who provided a representative group of the general population that seeks medical assistance in this region, and 48 volunteers (21 males and 27 females, 16–78 years old, 36 ± 25 years old) recruited among students and co-workers from the State University of Campinas. One hundred sixteen patients consecutively referred to the outpatient clinic of the University Hospital (HC-FCM/UNICAMP) for thyroid disease evaluation, who agreed to participate, were enrolled in the study. The study population included 49 cases of benign thyroid lesions [multinodular goiter (38 cases, 3 males and 35 females, 12–71 years old, 45 ± 15 years old) and follicular adenomas (11 cases, 2 males and 9 females, 22–77 years old, 43 ± 17 years old)] and 67 cases of thyroid tumors [50 papillary carcinomas (16–72 years old) and 17 follicular carcinomas (45–79 years old)]. Stage and grade of differentiation of the tumors were obtained from surgical and pathological records. Diagnoses were confirmed by experienced pathologists of the University Hospital (HC-FCM/UNICAMP). Patients were classified into whites and nonwhites. Clinical features of the patients with thyroid carcinomas are detailed in Table 1.

Data on general health conditions and medical history with emphasis on previous and/or current thyroid diseases were obtained through interviews. The use of drugs was also carefully assessed, in particular nutritional goitrogens, drugs that could interfere with thyroid function, as well as medicines for other concomitant diseases. Cigarette smoking habit was recorded but—because of the few reliable data obtained on the duration of smoking, age started smoking, quantity smoked, and years since stopped smoking—the patients were grouped in never-smokers and ever-smokers categories. This last group included individuals who consumed at least 20 packages (20 cigarettes each pack) for 1 year in the last 5 years. None of the patients had a history of accidental or medical radiation exposure. All data, including pathological diagnoses, were confirmed in the patients’ records. A peripheral blood sample was collected from all 116 patients. Also, thyroid tissue samples were obtained at surgery from 83 of these patients, snap frozen in liquid nitrogen immediately after surgery, and kept at −80°C until processed. We were able to obtain both cancer tissue samples and autologous blood specimens and/or normal thyroid tissue from the contralateral lobe in 35 cases of malignant tumors.

Patients were followed with periodic whole body scans, serum thyrotropin and thyroglobulin measurements according to a routine follow-up protocol that includes X-ray, ultrasonography, computer tomography scan, and other eventual procedures to detect distant metastasis for a period of 12–214 months (23 ± 58 months). Patients with high serum thyroglobulin levels (>2 mg/dl) and/or suspicious whole body scans were submitted to a thorough image search. We defined tumors as recurrent and/or presenting long distance metastasis according to the above parameters.

Genomic DNA was extracted from frozen specimens and from leukocytes separated from whole blood using a standard proteinase K-phenol-chloroform protocol. A tissue sample was collected from the central portion of the tumor to minimize the possibility of contamination of normal tissue.

A multiplex-PCR assay was used to simultaneously amplify the GSTT1 and GSTM1 genes. β-Globin gene coamplified as an internal positive control in a total volume of 50 μl containing 25 mm KCl; 10 mm Tris-HCl (pH 8.4); 1.5 mm MgCl₂; 0.1 mm each of dATP, dCTP, dGTP and dTTP; 500 ng of genomic DNA; and 2 units of Taq DNA polymerase recombinant (Life Technologies, Inc.). A fragment of 273 bp was obtained using 120 ng of each primer (sense and antisense) for GSTM1 gene amplification, 150 ng of each primer for the GSTT1 gene that amplified a fragment of 480 bp, and 95 ng of each primer for the β-globin gene that amplified a fragment of 630 bp (19). The reaction involved 35 cycles of incubation with denaturing at 94°C for 1 min, primer annealing at 62°C for 1 min, and extension at 72°C for 1 min. The PCR fragments were visualized into an ethidium bromide-stained 2% agarose gel as 273-bp and 480-bp products, corresponding to the normal presence of the allele of GSTM1 and GSTT1 genes, respectively. A 630-bp PCR fragment was obtained from the β-globin gene as shown in Fig. 1.

The analysis was conducted using statistical software (Statistical Analysis System, version 8.1, 1999–2000; SAS Institute, Inc., Cary, NC). χ² or Fisher’s exact tests were used to examine homogeneity between cases and controls regarding gender, color, previous thyroid disease, use of medicines, and cigarette smoking. Also, extent of disease was compared between PCs and FCs using Fisher’s exact test. The Kruskal-Wallis test was used to compare age among groups. The Mann-Whitney or Wilcoxon tests were used to compare age among different genotype groups. The odds ratio and 95% confidence interval provide a measure of the strength of association (e.g., indicating the increase in odds of a given benign or malignant thyroid tumor demonstrating a particular genotype compared with the control population). All tests were conducted at the P = 0.05 level of significance.

Table 1 summarizes clinical characteristics and parameters of aggressiveness at diagnosis and during follow-up of the thyroid cancer patients. Patients with FCs were older than the patients with PCs (Kruskal-Wallis, P = 0.0012). Nevertheless, there were no differences among the patients regarding color, gender distribution, smoking habit, the use of drugs, or preexisting benign thyroid diseases. The occurrence of lymph node involvement at the time of the diagnosis was more frequent among PC (39%) than FC (6%; χ², P < 0.05). However, these last tumors already presented long distance metastasis at the time of the diagnosis (53%) more frequently than PCs (16%; χ², P < 0.02). Follow-up data did not evidence differences between PC and FC concerning distant metastasis and/or recurrence of the tumor, although FC patients presented evidence of recurrence and/or metastasis in 50% of the cases against 30% of the PC cases (χ², P = 0.13). No patients died during the period of observation. Table 2 summarizes data of the overall proportions of the GSTT1 and GSTM1 genotypes in the control population and in the benign and malignant thyroid disease patients.

GSMT1 and GSTT1 gene profiles proved to be exactly the same in the tumor and normal tissue samples, as well as in the corresponding peripheral blood samples in all tested samples.

The combined absence of both the GSTT1 and GSTM1 genes was more frequent in thyroid carcinomas (12%) than in the control population (5%; Fisher’s exact test, P < 0.05) but did not distinguish benign from malignant tumors (Fisher’s exact test, P = 0.55) or patients with benign thyroid tumors from the control population (Fisher’s exact test, P = 0.32). There was no association between genotype and the patients’ clinical features, tumor parameters of aggressiveness at diagnosis, or behavior during follow-up. Also, no relationship was found between the GSTM1 and the GSTT1 allele null genotypes and benign or malignant thyroid tumors. The combined GSTT1 and GSTM1 null genotype presented an odds ratio of 2.576 (95% confidence interval, 1.044–6.355) in thyroid cancer patients.

The recognition of factors designed to identify people at risk of developing cancer is essential for good medical practice. Five to 10% of the population presents detectable nodules during their life span, mainly in iodine-deficient communities (6, 7, 20). However, most of these nodules will prove to be benign because thyroid cancer is responsible for only 0.6% to 1.6%, respectively, of all kinds of cancers that occur in men and women in the United States (6, 7). The environment has the principal role in causing sporadic cancer (2). Acting in concert with individual susceptibility, environmental factors such as smoking, diet, and pollutants play a role in most human cancers (1).

Variations of thyroid cancer incidence in different geographic and ethnic groups suggest that environmental factors may influence the thyroid tumorigenesis process, but available data regarding carcinogenic products are conflicting. Several chemicals produce thyroid neoplasia in rodents, generally acting through two basic mechanisms. The first involves a direct carcinogenic effect activating oncogenes, inactivating tumor suppressor genes, and producing specific alterations in the expression and function of genes involved in cell growth, differentiation, and life span. The second involves chemicals that, through a variety of mechanisms, disrupt thyroid function and produce neoplasia secondary to hormone imbalance (21). However, there are important species-specific differences in thyroid gland physiology between humans and rodents. Broad screening of more than 200 drugs for carcinogenicity revealed that just two, griseofulvin and senna, were associated with increased risk of thyroid carcinoma in humans (22). Spironolactone and vitamin D, but not calcium supplements, were found to be significantly associated with thyroid cancer, mostly medullary carcinoma (23). Nutritional goitrogens intake, like vegetables containing cyanogenic glucosides (most forms of cabbage, cauliflower, broccoli, and other members of the cruciferous family), was not associated with increased risk of thyroid carcinoma in humans and may even exert a protective effect (23). There is no epidemiological evidence of increased risk of thyroid cancer in smokers. On the contrary, recent data suggest a protective effect of smoking, perhaps involving an effect on thyroid-stimulating hormone and estrogen metabolism (12, 13).

Individuals with a homozygous deletion of the GSTT1 and GSTM1 genes lack enzymatic conjugation of foreign compounds with glutathione. This results in diminished ability to detoxify many environmental carcinogens, including 1.3-butadiene, ethylene oxide, epoxybutanes, and monohalomethanes. Absence of GSTT1 and of GSTM1 activity in blood, corresponding to the GSTT1 and GSTM1 null genotypes, have been associated with carcinogen-induced and background chromosomal changes in some case-control studies in lung, bladder, and colon cancers, particularly mediating the risk of smoke-related cancers (24, 25, 26). We were not able to find any literature data regarding the role of detoxifying enzymes in thyroid tumors. Therefore, we carried out a study of benign and malignant thyroid lesions involving the GSTT1 and GSTM1 genes. We studied a large control group that presented a genotype profile similar to that previously described in our population, confirming its high heterogeneity (27, 28). Our data on thyroid nodules indicate a high prevalence of the combined null genotype in malignant lesions in FC (17%) and PC (10%), significantly higher than in the control population (P < 0.05). We were not able to find any association between clinical features, histology, parameters of aggressiveness at diagnosis or during follow-up, and the genotype. Moreover, genotypes of tumor and normal autologous cells were always identical, indicating that GST gene inheritance does not play any role in the follicular cell transformation process. However, an estimated 2.6-fold greater risk of malignant thyroid nodules was observed in individuals with combined null genotypes. These data suggest that individuals with GSTT1 and GSTM1 combined null inheritance may be genetically predisposed for an increased risk of developing thyroid cancer.

This study would not have been possible without the cooperation of Dr. Alfio José Tincani and colleagues from the Department of Surgery, HC-FCM/UNICAMP, in particular Drs. José Geraldo dos Santos, Carlos Frazatto, Jr., and Gustavo Magaldi. We also thank Luís Francisco Cintra Baccaro for valuable suggestions regarding statistical evaluation and clinical data.