|
During the period of this grant we have been able to show that mutations in the p53 gene can be readily analyzed in routinely available paraffin-embedded formalin-fixed pathology specimens, families with a predisposition to malignancy can be identified by this analysis, and a relationship exists between tobacco-associated mutations and certain types of malignancies. We have identified about 63 pediatric tumors with mutations in the p53 gene, identified two families with p53 gene mutations and a predisposition to malignancy, and shown that tobacco-associated mutations occur in patients with pulmonary blastomas. We have refined molecular, immunohistochemical and biologic assays for p53 mutations, and are evaluating these systems.
During the period of this grant we have been able to show that mutations in the p53 gene can be readily analyzed in routinely available paraffin-embedded formalin-fixed pathology specimens, families with a predisposition to malignancy can be identified by this analysis, and a relationship exists between tobacco-associated mutations and certain types of malignancies. We have identified about 63 pediatric tumors with mutations in the p53 gene, identified two families with p53 gene mutations and a predisposition to malignancy, and shown that tobacco-associated mutations occur in patients with pulmonary blastomas. We have refined molecular, immunohistochemical and biologic assays for p53 mutations, and are evaluating these systems.
We continue to find that immunohistochemical (IHC) positivity for p53 protein does not correlate with mutation, in fact in some tumors elevated p53 expression may not be associated with mutation. In some tumors IHC positivity for p53 protein may be present in areas of rapid cell division as demonstrated by PCNA or BRDU positivity, even if a mutation can not be identified in these areas by careful analysis. This suggests that although these issues of predisposition to malignancy are important medically traditional methods of analysis may not be useful for identifying p53 mutations clinically. We have developed our molecular methods (sequencing and SSCP) to a degree of sophistication and rapidity, and we are also evaluating a yeast complementation method for p53 analysis from frozen tissue (requiring isolation of RNA followed by complementation of an artificially developed yeast). We believe that in the future these methods will replace immunohistochemical evaluation of p53 mutations which appears to lack sufficient accuracy to be clinically useful. We also find that in our group pediatric patients with p53 mutations tend to do poorly, since more deaths have occurred in the group with p53 gene mutations.
The major correlations of the data accrued by molecular analysis of tumors and clinical information remains to be made, Although we have been able to make statistical correlations with smaller sets of patients (e.g. patients with pulmonary blastoma) we have not yet been able to completely finalize our molecular data on the large number of tumors in the main study, or obtain full clinical information on tobacco exposure and family history in the large number of patients represented. On the molecular level a number of cases remain to be sequenced, conversely, on the data accrual side of the study, more clinical data must be obtained. After this has been completed we can make the larger correlations on the main study including establishing two by two tables and applying Fisher's exact test. At that time we can also establish Kaplan-Meir survival curves to determine the validity or our earlier impression that the children with p53 gene mutations do poorly clinically.
We continue to find that immunohistochemical (IHC) positivity for p53 protein does not correlate with mutation, in fact in some tumors elevated p53 expression may not be associated with mutation. In some tumors IHC positivity for p53 protein may be present in areas of rapid cell division as demonstrated by PCNA or BRDU positivity, even if a mutation can not be identified in these areas by careful analysis. This suggests that although these issues of predisposition to malignancy are important medically traditional methods of analysis may not be useful for identifying p53 mutations clinically. We have developed our molecular methods (sequencing and SSCP) to a degree of sophistication and rapidity, and we are also evaluating a yeast complementation method for p53 analysis from frozen tissue (requiring isolation of RNA followed by complementation of an artificially developed yeast). We believe that in the future these methods will replace immunohistochemical evaluation of p53 mutations which appears to lack sufficient accuracy to be clinically useful. We also find that in our group pediatric patients with p53 mutations tend to do poorly, since more deaths have occurred in the group with p53 gene mutations.
The major correlations of the data accrued by molecular analysis of tumors and clinical information remains to be made, Although we have been able to make statistical correlations with smaller sets of patients (e.g. patients with pulmonary blastoma) we have not yet been able to completely finalize our molecular data on the large number of tumors in the main study, or obtain full clinical information on tobacco exposure and family history in the large number of patients represented. On the molecular level a number of cases remain to be sequenced, conversely, on the data accrual side of the study, more clinical data must be obtained. After this has been completed we can make the larger correlations on the main study including establishing two by two tables and applying Fisher's exact test. At that time we can also establish Kaplan-Meir survival curves to determine the validity or our earlier impression that the children with p53 gene mutations do poorly clinically. |
