Summary Individuals with a deficiency in the enzyme dihydropyrimidine dehydrogenase (DPD) may experience severe life-threatening toxicity when treated with 5-fluorouracil (5-FU). As routine measurement of enzyme activity is not practical in many clinical centres, we have investigated the use of DNA mutation analysis to identify cancer patients with low enzyme levels. We have identified two new mutations at codons 534 and 543 in the DPD cDNA of a patient with low enzyme activity and screened the DNA from 75 colorectal cancer patients for these mutations and the previously reported splice site mutation (Vreken et al, 1996;Wei et al, 1996). In all cases, DPD enzyme activity was also measured. The splice site mutation was detected in a patient (1 out of 72) with low enzyme activity whereas mutations at codons 534 (2 out of 75) and 543 (11 out of 23) were not associated with low enzyme activity. These studies highlight the need to combine DPD genotype and phenotype analysis to identify mutations that result in reduced enzyme activity.Keywords: dihydropyrimidine dehydrogenase; 5-fluorouracil; polymorphism; colorectal cancer 5-Fluorouracil (5-FU) is widely used in the treatment of advanced solid tumours, including colorectal, breast and head/neck tumours. 5-FU is also frequently used in adjuvant chemotherapy for colorectal and breast cancers, in which its mild toxicity profile of mucositis and diarrhoea is well tolerated by patients who are at risk for tumour recurrence but have no current evidence of disease. 5-FU is a pyrimidine analogue and greater than 80% of a dose is degraded in a three-step pathway, initially catalysed by the enzyme dihydropyrimidine dehydrogenase (DPD; E.C. 1.3.1.2., Heggie et al, 1987). Deficiency in DPD enzyme activity is associated with a considerable delay in clearance of 5-FU from the plasma (Diasio et al, 1988;Fleming et al, 1992), leading to severe, life-threatening diarrhoea, neutropenia and in some cases neurotoxicity, incurring prolonged hospitalization . High concentrations of plasma and urine 5-FU, uracil and thymine may be detected along with low mononuclear cell DPD activity in these patients. The toxicity is thought to result from higher levels of 5-FU entering the anabolic pathway, resulting in an increased production of cytotoxic nucleotides. Although thymidine rescue has been attempted in one case (Takimoto et al, 1996), the majority of cases have been managed with supportive care after the cessation of 5-FU-based therapy. DPD activity is found in most human tissues, with the highest levels in the liver and lymphocytes. Population studies of peripheral blood mononuclear cell (PBMNC) DPD have shown that enzyme activity is variable with a seven-to 10-fold range observed (Lu et al, 1993;Etienne et al, 1994;McMurrough and McLeod, 1996). These studies suggest that although total deficiency is rare in adults, as many as 3% of the population may have low enzyme levels and thus be at One mutation in the DPYD gene has been reported to date in patients exhibiting severe toxicity after 5...

We have analyzed a series of nine infant leukemias that carry a t(11;19)(q23;p13). They had the morphologic features of acute lymphoblastic leukemia (ALL) and expressed markers typical of B-cell progenitor ALL or pre-B ALL; one coexpressed myeloid markers in addition to lymphoid markers (biphenotypic). Two probes (P/S4 and 98.40) subcloned from a yeast artificial chromosome (YAC) known to span the breakpoint in the t(4;11) were used to investigate DNA isolated from the leukemic cells of these patients. A total of approximately 15 kb of genomic DNA in the vicinity of the probes was examined by conventional Southern blot analysis using a series of restriction enzymes. In eight of the nine cases, the breakpoint could be mapped to an approximately 10-kb BamHI fragment disclosed by hybridization to the P/S4 probe.

Summary A mutation at codon 974 of the dihydropyrimidine dehydrogenase (DPD) gene was previously described in a cancer patient with undetectable DPD enzyme activity who experienced severe toxicity when treated with 5-fluorouracil. We have studied the frequency of this mutation in 29 Scottish subjects with low DPD enzyme activity and in 274 American subjects. We detected no mutations in the 606 alleles studied and conclude that mutations at codon 974 are a rare event.Keywords: dihydropyrimidine dehydrogenase; pyrimidine metabolism; 5-fluorouracil; thymine uraciluriaThe first and rate-limiting step in the catabolism of the pyrimidines uracil and thymine is carried out by the enzyme dihydropyrimidine dehydrogenase (DPD: EC 1.3.1.2). DPD enzyme activity has been detected in a number of human tissues, with the highest levels in the liver and lymphocytes (Naguib et al 1985). Population studies have shown that enzyme activity has a unimodel distribution over a 7-to 14-fold range, with some individuals having very low or even undetectable levels (Lu et al 1993;Etienne et al 1994;McMurrough and McLeod 1996). Low DPD activity is clinically important as such individuals can exhibit severe toxicity when treated with the anti-cancer agent 5-fluorouracil (5-FU) (Harris et al 1991;Houyau et al 1993). In addition, total DPD deficiency is associated with the congenital syndrome thymine uraciluria (Meinsma et al 1995). Population studies of DPD activity have suggested that the frequency of heterozygous and homozygous deficiency is 3% and 0.1% respectively . Two molecular alterations in the DPYD gene have been reported to date in patients with low enzyme activity. The deletion of a 165-bp exon has been demonstrated in a child with thymine uraciluria and in an adult who experienced near-fatal toxicity after receiving 5-FU therapy (Meinsma et al 1995; Wei et al 1996). A point mutation at codon 974 (aspartic acid to valine) has been identified in a patient who experienced severe 5-FU-related toxicity (Harris et al 1991;Albin et al 1995). The aspartic acid residue at codon 974 is not within the putative catalytic sites of the protein (Gonzalez and Fernandez-Salguero 1995) and the amino acid is conserved in the human, pig and cow sequences (GenBank/EMBL Accession numbers U09178, U09179 and U20981). The population frequency of these mutations is not known.A polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) assay was used to study the prevalence of point mutations at codon 974 in two population cohorts. InitialReceived 28 May 1996 Revised 14 August 1996 Accepted 21 August 1996 Correspondence to: SA Ridge studies were conducted in 29 Scottish blood donors with low DPD activity (19.1-69.3 pmol min-'mg-' protein), as measured using a high-performance liquid chromatography (HPLC)-based method for detecting radioactive metabolites (McMurrough and McLeod, 1996). This represented the lowest 10% of a larger population study. DNA from 274 American black and white blood donors (Memphis, TN;McLeod et al, 1994) was al...

We have analyzed a series of nine infant leukemias that carry a t(11;19)(q23;p13). They had the morphologic features of acute lymphoblastic leukemia (ALL) and expressed markers typical of B-cell progenitor ALL or pre-B ALL; one coexpressed myeloid markers in addition to lymphoid markers (biphenotypic). Two probes (P/S4 and 98.40) subcloned from a yeast artificial chromosome (YAC) known to span the breakpoint in the t(4;11) were used to investigate DNA isolated from the leukemic cells of these patients. A total of approximately 15 kb of genomic DNA in the vicinity of the probes was examined by conventional Southern blot analysis using a series of restriction enzymes. In eight of the nine cases, the breakpoint could be mapped to an approximately 10-kb BamHI fragment disclosed by hybridization to the P/S4 probe.