Grantee Research Project Results

The size of the primary tumor in these cases was:

Whether or not the regional lymph nodes were involved at the time of initial surgery also was determined from the pathology report. Regional nodes were positive for metastatic malignancy in 26 percent of cases. Only three cases (1.5 percent) had evidence for distant metastases at the time of their original surgery. The majority of the cases (64 percent) were incident (defined as diagnosed within two years of study enrollment) while only 36 percent had been diagnosed more than 2 years prior to enrollment.

The demographics of the patient population both cases and controls are as follows:

The racial distribution of cases and controls is essentially that of the study catchment area in the Portland Metropolitan area. The number of non-Caucasians (predominantly Black and Asian) in the study was too small to allow any conclusions to be drawn for any non-Caucasian subgroup. The median and mean age of cases is significantly older than that of controls consistent with age as a known risk factor for the development of HNSCC.

Each of the cases and controls in the study completed a questionnaire regarding cancer history, family history of cancer and exposure to tobacco products and ethanol. The Smoking and Alcohol Use reported was as follows:

Although the cases show a markedly increased history of tobacco exposure with approximately twice the average pack years as controls, this underestimates the effect of smoking as a risk factor in this population. In order to isolate the effect of genetic polymorphisms as a risk for HNSCC, the control population was purposely recruited from a population of smokers. Thus, smokers are over-represented in the control subjects as compared to the overall population of those without HNSCC. The cases also used alcohol substantially more than the control group.

Analysis of Genetic Polymorphisms. Five genetic polymorphisms common to the population in this study were selected for analysis. These included three glutathione transferase genes that encode detoxification enzymes, GSTM1, GSTT1 and GSTP1, a cytochrome p450, the tumor suppressor p53 and a gene encoding a DNA repair enzyme HOGG. The polymorphisms studied included GSTM1 and GSTT1 gene deletions, the GSTP1A polymorphism, the common CYP1A1 polymorphism, the codon 7 polymorphism of HOGG and the codon 72 polymorphism of p53. In each case DNA was prepared from the peripheral blood of subjects and subjected to PCR using appropriate paired primers for the region of the gene of interest. In the case of the GSTM1 and GSTT1 deletion genotypes, the absence of an appropriately sized PCR product on agarose gel electrophoresis indicated the homozygous deletion of the gene as is well established. In the case of the GSTP1A, HOGG7 and p53(codon72) polymorphisms, the single nucleotide polymorphisms (SNPs) produced a Restriction Fragment Length Polymorphism (RFLP). The presence or absence of the RFLP was assessed by gel electrophoresis following digestion of the PCR products with the appropriate restriction enzyme. This allows assessment of all three genotypes wt/wt, wt/mut and mut/mut for each SNP.

Odds Ratios. Chi Square Analysis. The results of the genotyping for each case and control were recorded and the data analyzed using the Chi-Square test. The results are as follows:

Based on these results, it appears that of the polymorphisms tested, only the homozygous deletion of the GSTT1 appears to be significantly associated with an altered risk of HNSCC. In the case of GSTT1, it is the subjects that bear the homozygous deletion (~30 percent of control population) who have the lowest risk of developing HNSCC. This makes biological and biochemical sense as the GSTT1 enzyme is known to be capable of activating a number of carcinogens to an alkylating moiety. The protective effect of GSTT1 deletion was most marked in women (OR .33, p <.005) and heavy ( > 45 pack years) smokers (OR .41. p < .01).

Logistic Regression. The effect of SNP genotype on risk of HNSCC also was examined by logistic regression. In logistic regression, the initial model included the known risk factors for HNSCC including age, gender, smoking and alcohol use. Each of the polymorphisms was then added singly and in pairs to the regression. Again, the only polymorphism that was associated with an altered risk of HNSCC was the GSTT1 deletion. In the logistic regression model, GSTT1 deletion was again associated with a significant ( p =.02) protective effect (OR .59). When only incident cases were examined the effect was unchanged.

In summary, squamous cell carcinoma of the head and neck is clearly a disease strongly associated with chronic heavy exposure to tobacco and alcohol. Our study revealed no significant increase in a family history of HNSCC in the subjects as compared to controls. In addition, several of the common genetic polymorphisms that have been linked to altered risk of smoking related cancers at other sites such as lung, esophagus and bladder were not associated with the risk of HNSCC whether or not the data were adjusted for smoking and alcohol exposure. The observed association of GSTT1 deletion with a reduced risk of HNSCC may be useful in identifying a portion of the population at lower risk for smoking related cancers at this and other sites. Further studies using the DNA samples obtained in this study are planned to examine other genetic polymorphisms as risk factors for HNSCC.

Supplemental Keywords:

Progress and Final Reports: