- Email: [email protected]
Gene deletion: a new target for cancer chemotherapy
Gene deletion:
a new
target for cancer chemotherapy
clonal deletions are common in and have led to the discovery of tumour suppressor genes (TSGs). Such genetic loss is "wasteful" in the sense that a deletion includes from 50-1000, presumably normal, flanking genes. I propose that loss of flanking clonal genetic material may serve as a target for
Cytogenetically evident
human
cancer
chemotherapy.
bone marrow transplantation) makes the difference between palliation and cure for several tumours.12 50% reduction in an inactivating enzyme for a chemotherapeutic agent within a tumour may be the equivalent of a 2-fold increase in dose whereas a > 90% decrease might be the equivalent of a 5-fold increase. Thus, the above enzyme changes are potential targets for
possible by
chemotherapy. Gene deletion Restriction
fragment length polymorphism (RFLP) analyses at multiple chromosomal sites (the allelogram) confirm that clonal loss of heterozygosity (LOH) occurs in cancer in as many as 20% of probe sites. The loss of the first allele may be caused by a germ line mutation (eg, familial retinoblastoma),’ but in adult epithelial tumours is more often due to a somatic event. Loss of the second allele may be due to: (a) loss of all or part (interstitial deletion) of a chromosome;1.3 (b) smaller deletions as shown by LOH on allelograms;l (c) sister chromatid exchange;4 (d) duplication of the abnormal chromosome; (e) mitotic crossing over or recombination;4 and (/) "point" mutation, which is probably the most common event.5,6 All of these events, except point mutation, would lead to the loss of several genes.
Implications for chemotherapy Loss of one allele for a recessive gene often leads to a 50% reduction in gene product,7 a loss which has treatment implications in view of the steep dose response curve for chemotherapeutic agents. Loss of both alleles, and therefore close to 100% of the gene product, would have greater implications. However, for TSGs that are known and sequenced such as p53 and APC, gene deletions flanking both allelles have not been observed.3 Nevertheless, there is evidence that complete loss of gene product may occur. If the flanking allelic loss involved a dominant gene in a tumour, complete loss of gene product might occur in 50% of patients. Dominant negative mutations may alter the function of the product of the second allele.8 Exposure of cancer cells to pulsed mutagens may lead to highly drug-resistant cell lines. For recessive genes, this is consistent with loss of expression of the second allele. Allelic loss in cancer is common,9 and is presumably more likely to occur in or near genetic "hotspots": sites where TSG or other cancer-related mutation or deletions have occurred.4 Genes linked to TSG loss such as esterase for retinoblastoma and aminoacylase for lung cancer may show a 90-100% reduction in enzyme activitylo,l1 For most antitumour agents a 2-fold increase in dose will substantially improve response, and a 5-fold increase (made
Department of Medicine, Dana-Farber Cancer Institute, Harvard Medical School, 44 Binney Street, Boston, MA 02115, USA. (Dr E Frei MD).
662
Effect of allele loss
example is loss of the gene enzyme dihydrouracil dehydrogenase (DUD) which inactivates the antitumour agent fluorouracil (FU). In-vitro inhibition of this enzyme increases the cytotoxicity of FU (table 1).13 Several patients who had toxicity to a conservative dose of FU have been found to have almost no DUD in normal tissues.l4 Family An
studies indicate that DUD is transmitted as a recessive gene.13,14 More recently, heterozygotes have likewise been found to exhibit excessive toxicity to conservative doses of FU.15 Knowing about deletion of an enzyme, transporter, signal transducer, or other gene product could lead to the identification and exploitation of new "negative" therapeutic targets (figure). The first row is a hypothetical metabolic pathway with substrate to product (A D) by a sequence of enzymatic (E) steps. If the gene for the enzyme that converts substrate B to product C were partly (one allele, recessive) deleted, it could mean a 50% reduction in the enzyme. This, and the related pool size changes, would have treatment implications: the deleted enzyme could become rate-limiting and thus more susceptible to analogue inhibition; substrate B would have a longer half-life and might accumulate in tumour cells (so an analogue of A or B, rendered cytotoxic by an alkylating or other moiety, would also accumulate in the tumour cells and be selectively cytotoxic); and loss of the enyme for which B is a substrate should lead to a diminished pool of subsequent substrates (small c or d in figure), and lead to increased entry of a cytotoxic analogue of c or d. Biochemical Normal
A
E 10
B
E 0
E
A
0,
g
e* op
Cancer Substrate analoguef E
e*
pathway C
0
——— D E
EÞ C
E
D
10
Final E ----+ product E
Product analoguet
Figure:
New
"negative" targets for treatment
E=enzymes A B, etc=substrate product sequence. *e=decrease in enzyme activity due to gene loss. tSubstrate analogue concentrates due to decreased activity of e; may inhibit e and further decrease product formation. If analogue has toxic moiety there will be selected cytotoxicity. #Pool size of C decreased to c. Analogue will compete preferentially with increased incorporation into final product. If analogue has toxic moiety there will be increased selective cytotoxicity.
ara-C=cytarabine; AA=alkylating agent; BCNU=carmustine; Bleo=bleomycin ; CPA= cyclophosphamide; ClCdB=cleocytidine; DHFR=dihydrofol ate reductase; FU=fluorouracil;GST=g!utathione transferase; GSH=glutathione; LCV=leucovorin; 6MP=6 6 mercaptopurine; THFA = 5,10,methylene tetra hydrofolic acid; TdR=thymidine;Topo=topoisomerase.
Table 1:
"Negative" gene deletion targets for cancer chemotherapy
The
cytotoxicity of many established chemotherapeutic is agents affected by activating and inactivating enzymes. There are inactivating enzymes that account for up to 90% of the intratumour metabolism of the agent (table 1). Inhibition of such an enzyme will therefore increase cytotoxicity. Examples of agents that inhibit inactivating enzymes include tetrahydrouridine for cytidine deaminase and 5-benzyl-oxybenzyl uracil (BBU) for DUD, resulting in increased cytotoxicity for ara-C and FU, respectively, for both tumour and normal tissues.16 Other examples include: aldehyde dehydrogenase and cyclophosphamide ;17 glutathione transferases and alkylating agents.;18 06 methyl-transferase and the nitrosoureas;19 and transport of salvage pathway molecules such as nucleosides in the presence of de novo pathway inhibitors such as methotrexate (Mtx) and FU.2O Another negative target strategy involves "rescue." Leucovorin is converted in two enzymatic steps to 5-10 methylene tetrahydofolic acid, the product of dihydrofolate reductase (DHFR) which is inhibited by Mtx. Loss of one of the activating enzymes’s for leucovorin would prevent leucovorin rescue of the
compared to the host.21 The human genome contains about 100 000 genes, of which 5% have been cloned. Over 10% of the lung cancer genome is deleted in 3-54% of patients with lung cancer.6 There are about 60 enzymes which may effect the biotransformation of the 40 or 50 established cancer chemotherapeutic agents. Those enzyme genes which have been cloned and mapped to sites of deletion, and which may be deleted in patients with non-small-cell lung cancer are shown in table 2. The site of the deletion is in the first column. The frequency of deleted sites is in descending order in the third column. The size of the deletion area (second column) was determined for the more common deletions by multiplying the fraction of the chromosome deleted by the fraction of the genome which that chromosome represents. Loss of 1 % of the human genome consists of about 1000 genes. In the fourth column, enzymes of known importance to cancer chemotherapy whose genes have been mapped to the deletion site are listed. The suggested treatment strategy for the deletions is presented in the fifth column.
tumour as
*% of chromosome x96 of genome For codes, see footnote table 1.
Table 2:
Cytogenetic deletions and gene enzyme assignment in lung cancer 663
Questions and approach Multiple clonal deletions or mutations are common in other epithelial tumors (breast, colon, head and neck, bladder) but differ in that different flanking gene targets for chemotherapy will be exposed. Do the gene enzymes that match on paper with common clonal deletion sites show biochemical evidence of deletion? The relevant cDNA probe could be used in Southern and northern blots to determine whether the gene or its transcript was present. The gene product-the enzyme-could be determined by western immunoblotting, with normal tissue studied in parallel. With fluorescent in-situ hybridisation (FISH), normal and tumour cells can be identified and compared in histopathology sections. In view of evidence for specific gene enzyme loss, what are the effects on the related biochemical pathway? Is the substrate pool expanded? Is there evidence of substrate or substrate analogue concentration? Is the product pool size decreased? Is there evidence for increased uptake of product or product analogue? Is there evidence that the deleted enzyme has become rate limiting? What are the biological effects of the deletion? Is there direct evidence for selective cytotoxicity? This could readily be determined for gene targets which relate to the biochemical pharmacology of known chemotherapeutic agents. Thus, if an inactivating enzyme is deleted in tumour as compared with control cells, the cytotoxicity of the relevant agent for the tumour should be increased. Such studies would provide the most direct and compelling information necessary for moving to in-vivo studies. For enzyme loss not known to be associated with a chemotherapeutic agent, the biochemical pharmacology should be reviewed to see if analogues can be developed. Some of the initial in-vitro work might involve established cell lines to avoid contamination with normal cells. It would be necessary, however, to confirm observations in fresh tissue culture, and when possible in biopsy specimens. Preclinical in-vivo treatment studies could use human tumour cells, with the deletion(s) of interest, in nude or SCID mice. The Human Genome Project should be monitored for progress in DNA sequencing, gene cloning, and mapping. Emphasis should be given to DNA sequencing in areas of common clonal deletions especially close to the site of the TSG loss/mutation. Computer matching for gene enzymes at such sites would provide important information about potential treatment targets as defined in this project .23 In addition, genes relating to potential target enzymes-eg, deoxycytidine deaminase and dehydrouracil dehydrogenase-should be cloned and mapped with the expectation that some of them will map to clonal deletion sites. I thank Dr M. Brown, Dr D. and Dr R. Tantravahi.
664
Kufe, Dr A. Pardee, Dr R. Sager, Dr H. Smith,
References 1
2 3
4 5
6 7 8 9
10
11
12
13
Whang-Peng J, Knutsen T, Gazzdar A, et al. Non-random structural and numerical changes in non-small cell lung cancer. Genes Chromosomes Cancer 1991; 3: 168-88. Knudsen AG Jr. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA 1971; 68: 820. Le Beau MM, Albain KS, Larson RA, et al. Clinical and cytogenetic correlations in 63 patients with therapy related myelodysplastic syndromes and acute non-lymphocytic leukemia: further, evidence for cytogenetic abnormalities of chromosomes numbers 5 and 7. JClin Oncol 1986; 4: 325-45. Gelehrter TD, Collins FS. Principlers of medical genetics. Baltimore: Williams and Wilkins, 1990: 255-289. Sager R. Genetic instability, suppression and cancer. Cancer Surv 1984; 3: 321-334.
Bishop JM. Molecular themes in oncogenesis. Cell 1991; 64: 235-48. Meyer UA, Zanger UM, Skoda RC, Grant D, Blum M. Genetic polymorphisms in drug metabolism. Prog Liver Dis 1990; 9: 307-23. Hershkowitz I. Functional inactivation of genes by dominant negative mutations. Nature 1987; 329: 219. Owens TK. Resistance to &bgr; arabinoturanosyl-cystosine in human lymphoblasts mediated by mutations within the deoxycytidine kinase gene. Cancer Res 1992; 82: 2389-93. Miller YE. Lack of expression of aminoacylase in small cell lung cancer: evidence of inactivation of genes encoded by chromosome 3p. J Clin Invest 1989; 83: 2120-42. Lee WH. A null allele of esterase D is a marker for genetic events in retinoblastoma formation. Human Genet 1987; 76: 33-36. Frei E III, Antman K. Combination chemotherapy, dose, and schedule. In: Holland JF, Frei E, Bast RC, Kufe DW, Morton DL, Weichselbaum VRR, eds. Cancer medicine. 3rd ed. Philadelphia: Lea and Febiger, 1993: 631-40. Harris BE, Carpenter JT, Diasio RB. Severe fluorouracil toxicity
secondary to dihydropyramidine dehydrogenase deficiency: a potentially more common pharmacogenetic syndrome. Cancer 1991; 68: 499-501. 14 Tuchman M, Schoeckeler JS, Kiang DT, O’Dea RF, Rammaraine ML, Mirkin BL. Familial pyrimidinemia and pyrimidinuria associated with severe fluorouracil toxicity. N Engl J Med 1985; 313: 245-49. 15 Lilenbaum RC, Harris BE, Diasio RB, Naes J, Lyss AP. Heterozygosity for dihydropyrimidine dehydrogenase deficiency may result in life-threatening toxicity from fluorouracil. Proc Am Soc Clin Oncol 1992; 33: 11. 16 Wolfenden R, Wentworth DF. On the interaction of 3, 4, 5, 6-tetrahydrouridine with human liver cytidine deaminase. Biochemistry 1975; 14: 599. 17 Diasio RB, Harris BE. Clinical pharmacology of fluorouracil. Clin Pharmacokinet 1989; 16: 215. 18 Waxman DJ. Glutathione S-transferase: Role of alkylating agent resistance and possible target for modulation chemotherapy-a review. Cancer Res 1990; 50: 6449-54. 19 Brent TP, Houghton PJ, Houghton J. O6-alkylguanine-DNA alkyltransferase activity correlates with the therapeutic response of human rhabdomyosarcoma xenografts to chloethyl-trans-4methylcyclohexyl nitrosourea. Proc Natl Acad Sci USA 1985; 82: 2985. 20 Remick SC, Grem JL, Fischer PH, et al. Phase I study of fluorouracil and dipyridamole administered by 72 hour concurrent continuous infusion. Cancer Res 1990; 540: 2667-72. 21 Bertino JR, Romanini A. Folate antagonists. In: Holland, Frei, Bast, Kufe, Morton, Weichselbaum, eds. Cancer medicine. 3rd ed. Philadelphia: Lea and Febiger, 1993: 698-71. 22 Pinkeld, Landegent JE, Collins C, Fuscoe J, Segraves R, lucas J, Gray J. Fluorescense in situ hybridization with human chromosome specific libraries: detection of trisomy 12 and translocation of chromosome 4. Proc Natl Acad Sci USA 1988; 85: 9138-42. 23 Human Gene Mapping. Ninth International Workshop. Cytogenet Cell Genet 1987; 46: 1-762.
a new
target for cancer chemotherapy
clonal deletions are common in and have led to the discovery of tumour suppressor genes (TSGs). Such genetic loss is "wasteful" in the sense that a deletion includes from 50-1000, presumably normal, flanking genes. I propose that loss of flanking clonal genetic material may serve as a target for
Cytogenetically evident
human
cancer
chemotherapy.
bone marrow transplantation) makes the difference between palliation and cure for several tumours.12 50% reduction in an inactivating enzyme for a chemotherapeutic agent within a tumour may be the equivalent of a 2-fold increase in dose whereas a > 90% decrease might be the equivalent of a 5-fold increase. Thus, the above enzyme changes are potential targets for
possible by
chemotherapy. Gene deletion Restriction
fragment length polymorphism (RFLP) analyses at multiple chromosomal sites (the allelogram) confirm that clonal loss of heterozygosity (LOH) occurs in cancer in as many as 20% of probe sites. The loss of the first allele may be caused by a germ line mutation (eg, familial retinoblastoma),’ but in adult epithelial tumours is more often due to a somatic event. Loss of the second allele may be due to: (a) loss of all or part (interstitial deletion) of a chromosome;1.3 (b) smaller deletions as shown by LOH on allelograms;l (c) sister chromatid exchange;4 (d) duplication of the abnormal chromosome; (e) mitotic crossing over or recombination;4 and (/) "point" mutation, which is probably the most common event.5,6 All of these events, except point mutation, would lead to the loss of several genes.
Implications for chemotherapy Loss of one allele for a recessive gene often leads to a 50% reduction in gene product,7 a loss which has treatment implications in view of the steep dose response curve for chemotherapeutic agents. Loss of both alleles, and therefore close to 100% of the gene product, would have greater implications. However, for TSGs that are known and sequenced such as p53 and APC, gene deletions flanking both allelles have not been observed.3 Nevertheless, there is evidence that complete loss of gene product may occur. If the flanking allelic loss involved a dominant gene in a tumour, complete loss of gene product might occur in 50% of patients. Dominant negative mutations may alter the function of the product of the second allele.8 Exposure of cancer cells to pulsed mutagens may lead to highly drug-resistant cell lines. For recessive genes, this is consistent with loss of expression of the second allele. Allelic loss in cancer is common,9 and is presumably more likely to occur in or near genetic "hotspots": sites where TSG or other cancer-related mutation or deletions have occurred.4 Genes linked to TSG loss such as esterase for retinoblastoma and aminoacylase for lung cancer may show a 90-100% reduction in enzyme activitylo,l1 For most antitumour agents a 2-fold increase in dose will substantially improve response, and a 5-fold increase (made
Department of Medicine, Dana-Farber Cancer Institute, Harvard Medical School, 44 Binney Street, Boston, MA 02115, USA. (Dr E Frei MD).
662
Effect of allele loss
example is loss of the gene enzyme dihydrouracil dehydrogenase (DUD) which inactivates the antitumour agent fluorouracil (FU). In-vitro inhibition of this enzyme increases the cytotoxicity of FU (table 1).13 Several patients who had toxicity to a conservative dose of FU have been found to have almost no DUD in normal tissues.l4 Family An
studies indicate that DUD is transmitted as a recessive gene.13,14 More recently, heterozygotes have likewise been found to exhibit excessive toxicity to conservative doses of FU.15 Knowing about deletion of an enzyme, transporter, signal transducer, or other gene product could lead to the identification and exploitation of new "negative" therapeutic targets (figure). The first row is a hypothetical metabolic pathway with substrate to product (A D) by a sequence of enzymatic (E) steps. If the gene for the enzyme that converts substrate B to product C were partly (one allele, recessive) deleted, it could mean a 50% reduction in the enzyme. This, and the related pool size changes, would have treatment implications: the deleted enzyme could become rate-limiting and thus more susceptible to analogue inhibition; substrate B would have a longer half-life and might accumulate in tumour cells (so an analogue of A or B, rendered cytotoxic by an alkylating or other moiety, would also accumulate in the tumour cells and be selectively cytotoxic); and loss of the enyme for which B is a substrate should lead to a diminished pool of subsequent substrates (small c or d in figure), and lead to increased entry of a cytotoxic analogue of c or d. Biochemical Normal
A
E 10
B
E 0
E
A
0,
g
e* op
Cancer Substrate analoguef E
e*
pathway C
0
——— D E
EÞ C
E
D
10
Final E ----+ product E
Product analoguet
Figure:
New
"negative" targets for treatment
E=enzymes A B, etc=substrate product sequence. *e=decrease in enzyme activity due to gene loss. tSubstrate analogue concentrates due to decreased activity of e; may inhibit e and further decrease product formation. If analogue has toxic moiety there will be selected cytotoxicity. #Pool size of C decreased to c. Analogue will compete preferentially with increased incorporation into final product. If analogue has toxic moiety there will be increased selective cytotoxicity.
ara-C=cytarabine; AA=alkylating agent; BCNU=carmustine; Bleo=bleomycin ; CPA= cyclophosphamide; ClCdB=cleocytidine; DHFR=dihydrofol ate reductase; FU=fluorouracil;GST=g!utathione transferase; GSH=glutathione; LCV=leucovorin; 6MP=6 6 mercaptopurine; THFA = 5,10,methylene tetra hydrofolic acid; TdR=thymidine;Topo=topoisomerase.
Table 1:
"Negative" gene deletion targets for cancer chemotherapy
The
cytotoxicity of many established chemotherapeutic is agents affected by activating and inactivating enzymes. There are inactivating enzymes that account for up to 90% of the intratumour metabolism of the agent (table 1). Inhibition of such an enzyme will therefore increase cytotoxicity. Examples of agents that inhibit inactivating enzymes include tetrahydrouridine for cytidine deaminase and 5-benzyl-oxybenzyl uracil (BBU) for DUD, resulting in increased cytotoxicity for ara-C and FU, respectively, for both tumour and normal tissues.16 Other examples include: aldehyde dehydrogenase and cyclophosphamide ;17 glutathione transferases and alkylating agents.;18 06 methyl-transferase and the nitrosoureas;19 and transport of salvage pathway molecules such as nucleosides in the presence of de novo pathway inhibitors such as methotrexate (Mtx) and FU.2O Another negative target strategy involves "rescue." Leucovorin is converted in two enzymatic steps to 5-10 methylene tetrahydofolic acid, the product of dihydrofolate reductase (DHFR) which is inhibited by Mtx. Loss of one of the activating enzymes’s for leucovorin would prevent leucovorin rescue of the
compared to the host.21 The human genome contains about 100 000 genes, of which 5% have been cloned. Over 10% of the lung cancer genome is deleted in 3-54% of patients with lung cancer.6 There are about 60 enzymes which may effect the biotransformation of the 40 or 50 established cancer chemotherapeutic agents. Those enzyme genes which have been cloned and mapped to sites of deletion, and which may be deleted in patients with non-small-cell lung cancer are shown in table 2. The site of the deletion is in the first column. The frequency of deleted sites is in descending order in the third column. The size of the deletion area (second column) was determined for the more common deletions by multiplying the fraction of the chromosome deleted by the fraction of the genome which that chromosome represents. Loss of 1 % of the human genome consists of about 1000 genes. In the fourth column, enzymes of known importance to cancer chemotherapy whose genes have been mapped to the deletion site are listed. The suggested treatment strategy for the deletions is presented in the fifth column.
tumour as
*% of chromosome x96 of genome For codes, see footnote table 1.
Table 2:
Cytogenetic deletions and gene enzyme assignment in lung cancer 663
Questions and approach Multiple clonal deletions or mutations are common in other epithelial tumors (breast, colon, head and neck, bladder) but differ in that different flanking gene targets for chemotherapy will be exposed. Do the gene enzymes that match on paper with common clonal deletion sites show biochemical evidence of deletion? The relevant cDNA probe could be used in Southern and northern blots to determine whether the gene or its transcript was present. The gene product-the enzyme-could be determined by western immunoblotting, with normal tissue studied in parallel. With fluorescent in-situ hybridisation (FISH), normal and tumour cells can be identified and compared in histopathology sections. In view of evidence for specific gene enzyme loss, what are the effects on the related biochemical pathway? Is the substrate pool expanded? Is there evidence of substrate or substrate analogue concentration? Is the product pool size decreased? Is there evidence for increased uptake of product or product analogue? Is there evidence that the deleted enzyme has become rate limiting? What are the biological effects of the deletion? Is there direct evidence for selective cytotoxicity? This could readily be determined for gene targets which relate to the biochemical pharmacology of known chemotherapeutic agents. Thus, if an inactivating enzyme is deleted in tumour as compared with control cells, the cytotoxicity of the relevant agent for the tumour should be increased. Such studies would provide the most direct and compelling information necessary for moving to in-vivo studies. For enzyme loss not known to be associated with a chemotherapeutic agent, the biochemical pharmacology should be reviewed to see if analogues can be developed. Some of the initial in-vitro work might involve established cell lines to avoid contamination with normal cells. It would be necessary, however, to confirm observations in fresh tissue culture, and when possible in biopsy specimens. Preclinical in-vivo treatment studies could use human tumour cells, with the deletion(s) of interest, in nude or SCID mice. The Human Genome Project should be monitored for progress in DNA sequencing, gene cloning, and mapping. Emphasis should be given to DNA sequencing in areas of common clonal deletions especially close to the site of the TSG loss/mutation. Computer matching for gene enzymes at such sites would provide important information about potential treatment targets as defined in this project .23 In addition, genes relating to potential target enzymes-eg, deoxycytidine deaminase and dehydrouracil dehydrogenase-should be cloned and mapped with the expectation that some of them will map to clonal deletion sites. I thank Dr M. Brown, Dr D. and Dr R. Tantravahi.
664
Kufe, Dr A. Pardee, Dr R. Sager, Dr H. Smith,
References 1
2 3
4 5
6 7 8 9
10
11
12
13
Whang-Peng J, Knutsen T, Gazzdar A, et al. Non-random structural and numerical changes in non-small cell lung cancer. Genes Chromosomes Cancer 1991; 3: 168-88. Knudsen AG Jr. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA 1971; 68: 820. Le Beau MM, Albain KS, Larson RA, et al. Clinical and cytogenetic correlations in 63 patients with therapy related myelodysplastic syndromes and acute non-lymphocytic leukemia: further, evidence for cytogenetic abnormalities of chromosomes numbers 5 and 7. JClin Oncol 1986; 4: 325-45. Gelehrter TD, Collins FS. Principlers of medical genetics. Baltimore: Williams and Wilkins, 1990: 255-289. Sager R. Genetic instability, suppression and cancer. Cancer Surv 1984; 3: 321-334.
Bishop JM. Molecular themes in oncogenesis. Cell 1991; 64: 235-48. Meyer UA, Zanger UM, Skoda RC, Grant D, Blum M. Genetic polymorphisms in drug metabolism. Prog Liver Dis 1990; 9: 307-23. Hershkowitz I. Functional inactivation of genes by dominant negative mutations. Nature 1987; 329: 219. Owens TK. Resistance to &bgr; arabinoturanosyl-cystosine in human lymphoblasts mediated by mutations within the deoxycytidine kinase gene. Cancer Res 1992; 82: 2389-93. Miller YE. Lack of expression of aminoacylase in small cell lung cancer: evidence of inactivation of genes encoded by chromosome 3p. J Clin Invest 1989; 83: 2120-42. Lee WH. A null allele of esterase D is a marker for genetic events in retinoblastoma formation. Human Genet 1987; 76: 33-36. Frei E III, Antman K. Combination chemotherapy, dose, and schedule. In: Holland JF, Frei E, Bast RC, Kufe DW, Morton DL, Weichselbaum VRR, eds. Cancer medicine. 3rd ed. Philadelphia: Lea and Febiger, 1993: 631-40. Harris BE, Carpenter JT, Diasio RB. Severe fluorouracil toxicity
secondary to dihydropyramidine dehydrogenase deficiency: a potentially more common pharmacogenetic syndrome. Cancer 1991; 68: 499-501. 14 Tuchman M, Schoeckeler JS, Kiang DT, O’Dea RF, Rammaraine ML, Mirkin BL. Familial pyrimidinemia and pyrimidinuria associated with severe fluorouracil toxicity. N Engl J Med 1985; 313: 245-49. 15 Lilenbaum RC, Harris BE, Diasio RB, Naes J, Lyss AP. Heterozygosity for dihydropyrimidine dehydrogenase deficiency may result in life-threatening toxicity from fluorouracil. Proc Am Soc Clin Oncol 1992; 33: 11. 16 Wolfenden R, Wentworth DF. On the interaction of 3, 4, 5, 6-tetrahydrouridine with human liver cytidine deaminase. Biochemistry 1975; 14: 599. 17 Diasio RB, Harris BE. Clinical pharmacology of fluorouracil. Clin Pharmacokinet 1989; 16: 215. 18 Waxman DJ. Glutathione S-transferase: Role of alkylating agent resistance and possible target for modulation chemotherapy-a review. Cancer Res 1990; 50: 6449-54. 19 Brent TP, Houghton PJ, Houghton J. O6-alkylguanine-DNA alkyltransferase activity correlates with the therapeutic response of human rhabdomyosarcoma xenografts to chloethyl-trans-4methylcyclohexyl nitrosourea. Proc Natl Acad Sci USA 1985; 82: 2985. 20 Remick SC, Grem JL, Fischer PH, et al. Phase I study of fluorouracil and dipyridamole administered by 72 hour concurrent continuous infusion. Cancer Res 1990; 540: 2667-72. 21 Bertino JR, Romanini A. Folate antagonists. In: Holland, Frei, Bast, Kufe, Morton, Weichselbaum, eds. Cancer medicine. 3rd ed. Philadelphia: Lea and Febiger, 1993: 698-71. 22 Pinkeld, Landegent JE, Collins C, Fuscoe J, Segraves R, lucas J, Gray J. Fluorescense in situ hybridization with human chromosome specific libraries: detection of trisomy 12 and translocation of chromosome 4. Proc Natl Acad Sci USA 1988; 85: 9138-42. 23 Human Gene Mapping. Ninth International Workshop. Cytogenet Cell Genet 1987; 46: 1-762.