This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Morel, A. Articles by Gamelin, E. Search for Related Content PubMed PubMed Citation Articles by Morel, A. Articles by Gamelin, E.

Research Articles: Therapeutics, Targets, and Development

Clinical relevance of different dihydropyrimidine dehydrogenase gene single nucleotide polymorphisms on 5-fluorouracil tolerance

¹ Oncopharmacology and Pharmacogenetic Laboratory and ² Medical Oncology Department, Institut National de la Sante et de la Recherche Medicale U564, Centre Paul Papin, Angers, France

Requests for reprints: Alain Morel, Oncopharmacology and Pharmacogenetic Laboratory, Institut National de la Sante et de la Recherche Medicale U564, Centre Paul Papin, 2 rue Moll, 49933 Angers cedex 9, France. Phone: 33-241-352-700. E-mail: a.morel{at}unimedia.fr

Abstract

Purpose: Although single nucleotide polymorphisms (SNP) of the dihydropyrimidine dehydrogenase gene (DPYD) have been reported, which affect enzyme activity and the severity of 5-fluorouracil (5-FU) toxicity, no pretherapeutic detection has thus far been developed. We investigated 22 DPYD gene SNPs, their respective incidence, their link with grade 3 to 4 toxic side effects, and their management in practice: 9 were looked for in 487 patients, whereas 13 others were investigated in 171 patients. Patients and Methods: SNPs were detected before 5-FU-based treatment in WBC using a Pyrosequencing method. Close clinical and biological follow-up was done. Results: Five different SNPs were found in 187 patients (IVS14 + 1G>A, 2846A>T, 1679T>G, 85T>C, –1590T>C). Three hundred patients had no SNP. Forty-four patients had grade 3 to 4 toxic side effects in either the first or second cycle. Sixty percent of patients with either IVS14 + 1G>A or 2846A>T SNPs and the only patient with 1679T>G SNP experienced early grade 3 to 4 toxicity, compared with 0%, 5.5%, and 15% of those with either –1590T>C, 85T>C SNP, or no SNP, respectively. In cases with grade 3 to 4 toxicity, treatment either had to be quickly stopped, or could be safely continued with an individual dose adjustment. Sensitivity, specificity, and positive and negative predictive values of the detection of these three major SNPs as toxicity predictive factors were 0.31, 0.98, and 0.62 and 0.94, respectively. Conclusion: Pretreatment detection of three DPYD SNPs could help to avoid severe toxic side effects. This approach is suitable for clinical practice and should be compared or combined with pharmacologic approaches. In the case of dihydropyrimidine dehydrogenase deficiency, 5-FU administration often can be safely continued with an individual dose adjustment. [Mol Cancer Ther 2006;5(11):2895–904]

Introduction

5-Fluorouracil (5-FU), a widely used drug, can result in severe toxic side effects, including death, (1–3). More than 80% of the administered drug is catabolized by dihydropyrimidine dehydrogenase (DPD) in dihydrofluorouracil (4). DPD activity is subject to a wide variability, mainly due to genetic polymorphism. This results in a broad range of enzymatic deficiency from partial (3–5 % of population) to complete loss (0.2% of the population) of enzyme activity (5–7) and consequently to severe polyvisceral 5-FU-induced toxicity (8).

Considering the frequent use of 5-FU and oral fluoropyrimidines, pretreatment detection of this deficiency could prevent severe toxicity. At least 50% of patients who experienced highly toxic 5-FU effects were genotypically heterozygous or homozygous for known single nucleotide polymorphisms (SNP) in the DPYD gene (9, 10). In addition to the splice site variant IVS14 + 1G >A considered the most common SNP (52%), >30 variant DPYD alleles have been published, and it has been suggested that more than half might have a deleterious effect on DPD enzyme activity (9–11). The prevalence of patients who are heterozygous or homozygous for these relevant SNPs and then at risk of 5-FU-induced severe toxicity is about 3%, but, to date, the respective frequencies of the different SNPs and their effects on DPD activity in patients have not been adequately investigated (11, 12).

The purposes of the present prospective study were (a) to set up an accurate minisequencing method suitable for the detection of several SNPs in clinical practice; (b) to determine the clinically relevant DPYD gene SNPs and their respective frequencies, their effect on 5-FU pharmacokinetics, their and clinical tolerance; (c) to evaluate DPYD genotyping with pyrosequencing for detection of patients at risk of toxicity with 5-FU administration in a population of 487 cancer patients (13).

Patients and Methods

Patients and Treatment
This prospective study was conducted on 487 French Caucasian patients treated for advanced carcinomas or in the adjuvant setting with 5-FU-based chemotherapy regimen. To be eligible, patients had to be naive of 5-FU, to have a WHO performance status <2, a life expectancy of at least 3 months, an age lower than 80 years, and adequate hematologic and cardiac status.

Treatment consisted of five different types of regimens, according to the type of cancer:

• Biweekly 5-FU infusion plus folinic acid (FA), LV5FU2, or de Gramont regimen (14), for colorectal carcinomas, in adjuvant or metastatic settings;
  
• Biweekly 5-FU-FA + irinotecan (FOLFIRI regimen; ref. 15), for metastatic colorectal carcinomas;
  
• Biweekly 5-FU-FA + oxaliplatin (FOLFOX regimen; ref. 16), for metastatic colorectal carcinomas;
  
• Four-day continuous infusion 5-FU + cisplatin or carboplatin/3 weeks (17), for head and neck cancers;
  
• Bolus 5-FU + epirubicin + cyclophosphamide (FEC regimen), every 3 weeks (18), for breast cancers in adjuvant or metastatic settings.
  

Written informed consent was obtained from all patients as was relevant Ethical Committee approval before starting the trial. Results of pretherapeutic genotyping were not taken into account.

However, the 5-FU dose was individually adjusted in cases of prolonged 5-FU infusion, as it is usually done in some institutions in routine practice (19). It began with the second cycle and was based on the 5-FU plasma concentration measured at the end of the previous 5-FU infusion (i.e., the concentration at the steady state, C_ss). 5-FU plasma concentration should be within the optimal therapeutic plasma range, and 5-FU dose was adjusted based on a pharmacokinetic follow up, as previously published (19). Fluorouracil concentrations in plasma were measured in the Oncopharmacology Department. Precision and accuracy tests were carried out beforehand in the control quality test program of the French Clinical Pharmacology Group of the French Federation of Anticancer Centers.

Assessment of Tolerability
Patients were asked for treatment tolerance and underwent examination with hemograms, ionograms, and liver and kidney tests, before every chemotherapy cycle. All early 5-FU-related adverse events (i.e., at the first or the second 5-FU cycle) were taken in account and graded for severity according to the National Cancer Institute Common Toxicity Criteria scale: diarrhea, mucositis, neutropenia, thrombopenia, central neurotoxicity, cardiac toxicity, and hand-foot syndrome. Treatment had to be stopped in case of grade 3 to 4 toxicity until complete recovery, and in case of grade 4, the French Drug Committee had to be notified.

Methods
Detection of SNPs of DPYD and Its Promoter. We selected 22 relevant SNPs previously described in the international literature. These SNPs had either been previously found in patients with DPD activity deficiency, or by their nature and its location were clearly able to generate a deleterious effect on DPD enzyme structure and activity. The SNPs are displayed in Table 1 according to the international nomenclature (20–31). We systematically looked for nine SNPs that had been more frequently reported, in the entire population of 487 patients, and for an additional 13 SNPs, potentially implied in DPD deficiency but more infrequently reported, in 171 patients, with or without early severe 5-FU-induced toxicity.

WBC Isolation and DNA Extraction. Blood samples were obtained for DNA isolation and determination of genotypes. All procedures were reviewed and approved by accredited ethics review boards, and patients signed an informed consent form.

Peripheral blood mononuclear cells were isolated by centrifugation of the blood at 3,500 x g for 15 minutes. DNA was extracted using DNA Isolation kit for blood/bone marrow/tissue (Roche Molecular Diagnostics, Meylan, France). Each sample was controlled with respect to DNA isolation by UV transillumination of ethidium bromide–stained gels from subsequent electrophoretic separation in 1.2% agarose.

Pyrosequencing Analysis
PCR Conditions. PCRs were done with an initial denaturation for 5 minutes at 95°C followed by 50 cycles of denaturation for 30 seconds at 95°C, primer annealing for 30 seconds at 55°C, and extension for 1 minute at 72°C. All amplification reactions were done in a DNA Thermal Cycler 480 (Perkin-Elmer, Boston, MA) with 1 unit of Taq Polymerase (Euroblue Taq, Eurobio, Les Ulis, France).

PCR conditions were the same for the all mutations tested and produced in each case PCR products without significant background. Table 2 shows the different sets of primers used to amplify the sequences of interest.

Sequencing primers and dispensation order of the nucleotides are displayed in Table 3 .

5-FU Plasma Clearance
5-FU plasma clearance was calculated at the end of 5-FU infusion (i.e., at the steady state) because 5-FU half-life is about 8 to 10 minutes, according to the following formula: Cl = flow rate (mg m^–2 h^–1) / C_ss (mg L^–1), where C_ss is the 5-FU concentration at the steady state.

5-FU concentrations in plasma were determined with a previously described method (19, 34).

Statistical Analysis
Frequencies of SNPs and toxic side effects were measured and then compared. Correlations were looked for to assess whether or not statistical difference existed between two observations (i.e., between two percentages relative to toxic events occurrence). Pearson's ² test was used.

We determined the sensitivity, specificity, and positive and negative predictive values of SNPs as prediction factors of toxic side effects. Sensitivity (Se) was calculated as follows: number of true positive patients/number of true positive + false-negative patients.

Specificity (Sp) was calculated as follows: number of true negative patients/number of true negative + false-positive patients; positive predictive value was calculated as follows: Se x Pre/Se x Pre + (1 – Sp)(1 – Pre); negative predictive values was calculated as follows: Sp(1 – Pre) / Sp(1 – Pre) + Pre(1 – Se), where Se = sensitivity, Sp = specificity, and Pre = prevalence.

Analyses were done using SPSS software (SPSS, Paris, France).

Results

Technical Results
Conditions for the Pyrosequencing Analysis. To sequence the 22 variants, the sequencing primers designed were suitable for our method, and no interfering peaks were generated. The detection of the 22 variants used standards variables and provided satisfactory results both for the wild type detected in patients as for the mutant sequence from the constructs, engineered to harbor a variant sequence, as illustrated by pyrograms of IVS14 + G>A, 2846A>T, and 1679T>G (Fig. 1 ).

View larger version (26K): [in this window] [in a new window]   Figure 1. Pyrograms of IVS14 + G>A (A), A2846T (B), and T1679G (C) SNPs from patients.

5-FU-related toxicities are displayed in Table 4 . Of the 487 patients, 20 had grade 4 toxic side effects, in either the first or the second cycle, consisting principally of diarrhea and neutropenia. In these cases, toxic effects comprised a polyvisceral syndrome. Two of the 20 patients experienced grade 4 toxicity despite an initial 5-FU dose reduction of 20% and 40%. Fourteen of the 20 patients who had grade 4 toxicity had to be transferred to an intensive care medicine department for 15 to 28 days because of a polyvisceral toxic syndrome, consisting of diarrhea, febrile neutropenia and mucositis. Four of these 14 patients had neurologic symptoms with coma. Treatment had to be stopped for 8 of the 20 patients for grade 4 toxicity but was continued in 11 patients, following a variable 5-FU dose reduction from 15% to 75% of the initial dosage, followed by an individual 5-FU dose adjustment with pharmacokinetic follow-up.

Fourteen and nine patients had grade 2 or 1 toxic side effects, respectively, as early as the first cycle, and these had an individual dose adjustment, with 15% to 45% and 20% to 35% dose reductions, respectively.

We assessed whether or not the distribution of early toxicity depended on the chemotherapy regimen and concluded that toxicity in the first and second cycles was independent of the regimen used. The frequencies of grade 3 to 4 toxic side effects were equivalent with the 5-FU + FA, FOLFIRI, and FOLFOX regimens (8.3%, 6%, and 5.5%, respectively). Although the toxicities were somewhat little higher in patients receiving the 5-FU-CDDP regimen (16%), it was felt that this was a result of the combination, and the resultant increased incidence of hematotoxicity.

Concerning DPYD gene polymorphism, nine SNPs were systematically searched for in the whole population of patients, whatever the tolerance of the treatment (Table 5 ). Ten patients had IVS14 + 1G>A polymorphism. In nine of these, sequencing showed a heterozygous result. These nine included three women and five men, as well as one man whose DNA was shown to be heterozygous both for the IVS14 + 1G>A and the –1590T>C SNPs. In one woman, sequencing showed a homozygous result. In 10 patients (4 women and 6 men), a heterozygous result was observed for the 2846A>T SNP. In 150 and 15 patients, heterozygous and homozygous results, respectively, were found for the 85T>C SNP. DNA in seven patients was heterozygous for the –1590T>C SNP. One patient's DNA was heterozygous for 1679T>G.

We looked for a correlation between the DPYD SNPs and the 5-FU-related toxic side effects (Table 3). Of 10 patients with the IVS14 + 1G>A SNP, six experienced grade 3 to 4 toxic side effect (60%; one FEC, one LV5FU2, two FOLFIRI, one FOLFOX, and one 5-FU-CDDP), presenting principally as a combined polyvisceral syndrome, which was lethal in one patient. One patient, a woman with breast cancer whose sequence analysis was homozygous for IVS14 + 1G>A, experienced severe toxicity after receiving low-dose 5-FU as party of the FEC protocol given in an adjuvant setting. Within a few days of the first cycle, she was hospitalized in intensive care, with grade 4 diarrhea, neutropenia, and mucositis, remaining there for 15 days. Two additional patients with the IVS14 + 1G>A SNP had no toxicity but had an their initial 5-FU dose reduced. Among the six patients with severe toxicity in the first or second cycle, treatment was stopped as early as the second cycle in three and was continued for the other three with a good tolerance following a 25% to 50% dose reduction with pharmacokinetic follow-up.

Similarly, six patients of the 10 with a 2846A>T SNP experienced grade 3 to 4 toxic side effects in the first two cycles (60%; four LV5FU2, one FOLFOX, and one 5-FU-CDDP). Among the three patients without toxicity in the first two cycles, two had a 20% and 45% 5-FU dose reduction. Among the seven patients with initial toxicity, three had their treatment stopped, whereas four received additional 5-FU with 25% to 50% dose reductions and pharmacokinetic follow up.

One patient with a 1679T>G SNP experienced grade 4 toxicity (LV5FU2). After a 6-week treatment interruption, 5-FU was safely reintroduced with individual pharmacokinetic adjustment.

One hundred sixty-five patients presented with the 85T>C SNP (150 heterozygous and 15 homozygous). Of these, 11 had grade 3 to 4 toxicity (6.7%); none of them was homozygous. Seven patients had the –1590T>C SNP. Only one of these seven experienced grade 1 toxicity in the first cycle, and this was a patient in whom a IVS14 + 1G>A SNPs was also noted.

For 300 patients, a SNP was not detected (62%). Thirty-three experienced early toxic side effects of any grade in the first or second cycle (11%), whereas 20 had severe grade 3 to 4 toxicities (6.7%). Of the seven patients who had grade 4 toxic side effects, treatment was stopped for three and continued for four, with dose reductions from –20% to –40%. The therapy tolerance of those whose dose was reduced was excellent. Thirteen patients experienced grade 3 toxicity. Among them, treatment was stopped for two patients and continued for the others with –15% to –40% dose reductions and individual pharmacokinetic follow-up.

We compared the frequency of grade 3 to 4 toxic side effects according to the presence of a SNP and the type of SNP. We found a highly significant difference in toxicity when patients with either a IVS14 + 1G>A, 2846A>T, or 1679T>G SNP were compared with those without a SNP (P = 10^–6). We pooled these three SNPs even if we could not draw any statistical conclusion about the statistical correlation between 1679T>G and severe toxicity because this SNP was shown to modify the FAD site, to interference with cofactor binding, and to suppress DPD activity. We found no difference between –1590T>C or 85T>C patients and patients without a SNP (P = 0.9). The presence of one of the three relevant SNPs (IVS14 + 1G>A, 2846A>T, and 1679T>G) was highly correlated with early severe toxicity (P = 2.8 10^–10).

Sensitivity, specificity, and positive and negative predictive values of the detection of these three major SNPs (IVS14 + 1G>A, 2846A>T, or 1679T>G) as predictive factors for toxic side effects were 0.31, 0.98, and 0.62 and 0.94, respectively.

5-FU plasma clearance at the first cycle was evaluated in 457 patients. The clearance was not evaluated in the FEC regimen, where 5-FU was given in bolus infusion. We compared the clearances of the patients with IVS14 + 1G>A, 2846A>T, or 1679T>G SNP with those of the patients with no SNP found or with 85T>C or –1590T>C (Table 6 ). Mean clearances were 74.9 ± 38.3 L h^–1 m^–2 (range = 21.2–183.5) and 132.3 ± 46.6 L h^–1 m^–2 (range = 36.8–369.7), respectively. Median values were 66.2 and 123.9 L h^–1 m^–2, respectively. They were statistically different (P < 0.001). Because some patients (seven with measurable clearances) with IVS14 + 1G>A, 2846A>T, or 1679T>G SNP had no or light or mild toxic side effects (0 to II), and because some patients with no SNP found or with 85T>C or –1590T>C (26 with measurable clearances) experienced severe toxic side effects, we compared the clearances in these two populations of patients. Mean clearances were 95.5 ± 47.5 L h^–1 m^–2 (range = 21.2–183.5) and 77.5 ± 21.1 L h^–1 m^–2 (range = 36.8–119.5), respectively. Median values were 90.7 and 76.6 L h^–1 m^–2, respectively. Interestingly, we found no statistical difference (P = 0.156).

In the present prospective study, we determined frequencies and effects of certain DPYD SNPs on 5-FU tolerance and evaluated a high-throughput genotyping method for detection of patients at risk of severe 5-FU-induced toxic side effects. Twenty percent to 30% of grade 3 to 4 toxic side effects and 0.5% mortality rate have been reported in patients treated for advanced diseases or in adjuvant therapy with 5-FU bolus administration or continuous infusion (14–16, 34, 35). They have been investigated for their relationship with 5-FU metabolism rate and/or DPD activity deficiency, but no prevention strategy has been widely proposed and accepted (2, 3, 35). Certain DPYD SNPs can have a deleterious effect on DPD activity, if they are located in some hotspots of the gene, for example, interfering with substrate or cofactor binding, with electron transport by altering [4Fe-4S] function, or resulting in truncated nonfunctional protein (10, 24, 30, 36, 37). According to the literature, the prevalence of DPYD SNP heterozygous patients at risk of severe fluoropyrimidine's toxicity is about 3% to 5%, the splice site mutation IVS14 + 1G>A being by far considered the most common one, with 52% frequency compared with the other ones (12). However, until now, the exact determination of relevant DPYD alleles, their respective frequencies, and their effects on DPD activity and patients' tolerance to 5-FU, remained to be determined (34).

Different pharmacologic approaches have been tested for detecting DPD deficiency, such as the DPD activity measurement in peripheral mononuclear cells by radioenzymatic assay (38, 39), the measurement of natural pyrimidines concentrations in plasma (40), the determination of the ratio dihydrouracil/uracil (UH₂/U) in plasma (41), and the 2-¹³C-uracil breath test (42). The detection of relevant DPYD gene SNPs by genotyping technique is promising, but many questions remain unresolved, such as which adequate method and which relevant SNPs (43).

In the present prospective study, we set up a high-throughput minisequencing method by pyrosequencing technique for genotyping DPYD gene on a large population of patients, suitable to a clinical practice. We were able to detect 22 SNPs. First, in our whole population of 487 Caucasian patients, we looked for nine SNPs, eight of the gene itself in the coding regions and one in its promoter, because of both their frequency and their potential deleterious effect on DPD activity. Then, we systematically searched for the 13 other ones: first, in 44 patients with early 5-FU-related toxicity but no SNP found and, because we found none of them, we continued for a total of 171 patients, and at last we stopped because the minimum frequency of a SNP is usually estimated at 1%.

In our population of patients, we found five SNPs. IVS14 + 1G>A was the most frequently reported in the literature (11, 12). Its frequency in the present study, both heterozygous and homozygous status, was 1.8% and 0.2%, respectively, quite similar to that previously published (11, 12). Concerning 2846A>T, previous data about its incidence and relevance were very scarce, but we found a high frequency of heterozygosity (2%). Besides, 85T>C frequency was high: 31% heterozygous and 3% homozygous. One patient presented with 1679T>G (0.2%). At last, concerning DPYD promoter, seven patients were heterozygous for –1590T>C SNP (1.4%). We found no other SNP.

Of the five DPYD SNPs, two were highly correlated with early 5-FU toxicity because 60% of the patients with IVS14 + 1G>A and 2846A>T experienced very early severe toxic side effects. Compared with the risk of toxicity for patients with no SNP found (i.e., 6.6%), the difference was highly significant. For IVS14 + 1G>A, our results are in agreement with previous ones (11, 12). On the contrary, 2846A>T SNP, much less reported in the literature than IVS14 + 1G>A, had a high prevalence in our population of French, White Caucasian patients and was as deleterious as IVS14 + 1G>A on 5-FU tolerance. One patient heterozygous for 1679T>G had grade 4 febrile neutropenia, and his treatment quickly stopped. The frequency in our population of patients is then 0.2%. We are unable to draw any statistical conclusion, but this SNP is known to suppress DPD activity.

We evaluated the effect of these DPYD gene variants (IVS14 + 1G>A, 2846A>T, and 1679T>G) on 5-FU pharmacokinetics. Clearly, patients with these SNPs had a significantly reduced 5-FU plasma clearance and were brutally exposed to too high 5-FU plasma concentrations, when its dose was adjusted to body surface area. A close relationship between high 5-FU plasma concentrations and clinical toxicity has previously been shown (34). Thirty percent of patients with either the IVS14 + 1G>A or the 2846A>T SNP had no adverse events. Interestingly, these patients had 5-FU plasma clearance significantly lower than patients with no SNP found and no toxicity but a little higher than patients with SNPs and toxicity. In these cases, we can hypothesize that toxicity did not occur because of either an initial 5-FU dose reduction and/or a pharmacokinetic follow-up. On the other hand, patients with no SNP found, but who experienced severe initial toxic side effects, had a significantly reduced 5-FU plasma clearance.

We would point out the following. Even after severe toxicity, treatment could be safely resumed, in 50% of the patients, provided the dose was reduced, and close pharmacokinetic monitoring was instituted to avoid too high 5-FU plasma levels. Thus, the identification of genetic heterozygosity at either IVS14 + 1G>A or 2846A>T is not an absolute contraindication to 5-FU administration, provided the physician is aware and appropriate precautions are instituted.

We would also note that although the 85T>C SNP was found at a high frequency in our population (31% heterozygous and 3% homozygous), this SNP seems to have had little to no effect on 5-FU tolerance because only 11 of 165 (6.6%) patients had severe toxic side effects. Although previous studies have reached contradictory conclusions about its possible role in 5-FU toxicity (23, 32), our results agree with those of Collie-Duguid et al., in that we would characterize it as a common and pharmacokinetically inconsequential polymorphism (32). Similarly, although the –1590T>C SNP, located on the IFN-binding site of the DPYD promoter, has been previously suspected of lowering DPYD gene expression, our results suggest it has little to no clinical effect. In our population, only one of the seven patients experienced toxicity, and this patient was found to have SNPs at both T-1590C and IVS14 + 1G>A.

Finally, we would point out that in 300 patients representing 62% of the total population, a mutation could not be found at any of the 22 SNPs screened. Twenty of these 300 patients or 6.6% had severe toxicity in either the first or the second cycle. In three of these 20 patients, treatment was stopped, whereas in the other 17 patients, treatment was resumed after they had recovered from the toxic side effects with secondary dose reductions of 15% to 40%, according to a pharmacokinetic follow-up protocol. Although the incidence of toxicity in these patients was substantially less than in patients with some of the SNPs, these result raises the question of the cause of 5-FU toxicity in these patients. Possible explanations include (a) additional DPYD mutations (SNPs) that can affect pharmacokinetics may yet remain to be discovered, (b) Mutations or SNPs in other enzymes, involved in 5-FU catabolism, may be causative. For one such enzyme, dyhydropyrimidinase, a deficiency has been rarely reported (44). (c) Polymorphism in thymidylate synthetase, a target of 5-FU, may be etiologic. A polymorphism in the thymidylate synthetase gene promoter (2R/2R and 3R/3R) has been correlated with toxicity, although usually delayed and chronic in its presentation (45). Finally, (d) epigenetic factors not yet identified may be responsible.

We would argue that given the large number of patients treated each year with 5-FU or other fluoropyrimidines, such as capecitabine and UFT, and the human and economical cost of grade 3 and 4 toxic side effects, pretherapeutic detection of DPD deficiency should be considered. DPYD genotyping to detect the IVS14 + 1G>A, 2846A>T, and 1679T>G SNPs is a promising approach. We would argue that detection of only the IVS14 + 1G>A SNP is insufficient. As for the other SNPs that seem to be much rarer but clinically relevant, we would suggest that evaluation may depend on the country one is dealing with because their prevalence may vary among different populations. In practice, one may ask whether all known relevant SNPs, or merely those that are found more frequent, should be evaluated. We would note that because pyrosequencing can give results given within a day, a more comprehensive screen may be suitable for clinical practice. However, we would caution that DPYD SNP detection cannot prevent all severe toxic side effects that occur in the first or second cycle. In our experience, DPYD SNP detection would seem to be able to avoid at least 40% of 5-FU early severe toxicities. Although the sensitivity of this test at 0.31 is low, its positive predictive value at 0.62 is good, and its specificity (0.98) and its negative predictive value (0.94) are very high.

The pretherapeutic detection of DPD deficiency could reap benefits in terms of reduced costs and more importantly reduced toxicity with improved quality of life. We feel that this test will have to be compared, in terms of suitability to clinical practice, sensitivity, specificity, and positive and negative predictive values, as well as cost, with other approaches, such as the pretherapy uracil plasma measurement, UH₂/U plasma ratio determination, or more recently reported, the 2-¹³C-uracil breath test, although 2-¹³C-uracil is not currently available in every country (42).

We would also point out that one should not simply rely a detection of a SNP or a DPD deficiency. The management of a patient with DPD deficiency depends on his homozygous or heterozygous variant status and on other personal variables, such as age, comorbidity factors, and disease staging. It requires a therapeutic and pharmacologic advice and a close collaboration among physicians, pharmacologists, and molecular biologists. In case of homozygosity, an alternative treatment is recommended, 5-FU metabolism being quite totally impaired. In case of heterozygosity, 5-FU can be safely and efficiently given for several months, with added strict safety precautions, such as an initial 5-FU dosage reduction and an individual dose adjustment based on a close clinical and pharmacokinetic follow-up (19, 20).

Footnotes

Grant support: Comité Départemental du Maine et Loire de la Ligue Contre Le Cancer.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 6/ 1/06; revised 8/30/06; accepted 9/25/06.

References

1. Tuchman M, Stoeckeler JS, Kiang DT, et al. Familial pyrimidinemia and pyrimidinuria associated with severe fluorouracil toxicity. N Engl J Med 1985;313:245.[Medline]
2. Diasio RB, Beavers TL, Carpenter JT. Familial deficiency of dihydropyrimidine dehydrogenase (biochemical basis for familial pyrimidinemia and severe 5-fluorouracil-induced toxicity). J Clin Invest 1988;81:47–51.[Medline]
3. Diasio RB, Johnson MR. Dihydropyrimidine dehydrogenase: its role in 5-fluorouracil clinical toxicity and tumor resistance. Clin Cancer Res 1999;5:2672–3.[Free Full Text]
4. Pinedo HM, Peters GFJ. Fluorouracil: biochemistry and pharmacology. J Clin Oncol 1988;6:1653–64.[Abstract/Free Full Text]
5. Etienne MC, Lagrange JL, Dassonville O, et al. Population study of dihydropyrimidine dehydrogenase in cancer patients. J Clin Oncol 1994;12:2248–53.[Abstract/Free Full Text]
6. Lu Z, Zhang R, Diasio RB. Dihydropyrimidine dehydrogenase activity in human peripheral blood mononuclear cells and liver: population characteristics, newly identified deficient patients, and clinical implication in -fluorouracil chemotherapy. Cancer Res 1993;53:5433–8.[Abstract/Free Full Text]
7. Van Kuilenburg ABP, Abreu R, Van Gennip A. Pharmacogenetic and clinical aspects of dihydropyrimidine dehydrogenase deficiency. Ann Clin Biochem 2003;40:41–5.[CrossRef][Medline]
8. Johnson MR, Hageboutros A, Wang K. Life-threatening toxicity in a dihydropyrimidine dehydrogenase deficient patient after treatment with topical 5-Fluorouracil. Clin Cancer Res 1999;5:2006–11.[Abstract/Free Full Text]
9. Mc Leod HL, Collie-Duguid ESR, Vrecken P, et al. Nomenclature for human DPYD alleles. Pharmacogenetics 1998;8:455–9.[Medline]
10. Van Kuilenburg ABP, Haasjes J, Van Gennip H, et al. Clinical implication of dihydropyrimidine dehydrogenase (DPD) deficiency in patients with severe 5-fluorouracil-associated toxicity: identification of new mutations in the DPD gene. Clin Cancer Res 2000;6:4705–12.[Abstract/Free Full Text]
11. Van Kuilenburg ABP, Meinsma R, Zoetekouw L, et al. Increased risk of grade IV neutropenia after administration of 5-fluorouracil due to dihydropyrimidine dehydrogenase deficiency: high prevalence of the IVS14 + 1G>A mutation. Int J Cancer 2002;101:253–8.[CrossRef][Medline]
12. Van Kuilenburg ABP, Muller EW, Haasjes J, et al. Lethal outcome of a patient with a complete dihydropyrimidine dehydrogenase (DPD) deficiency after administration of 5-fluorouracil: frequency of the common IVS14 + 1G>A mutation causing DPD deficiency. Clin Cancer Res 2001;7:1149–53.[Abstract/Free Full Text]
13. Rouits E, Boisdron-Celle M, Dumont A, et al. E. Relevance of different UGT1A1 polymorphisms in irinotecan-induced toxicity. A molecular and clinical study of 75 patients. Clin Cancer Res 2004;10:5151–9.[Abstract/Free Full Text]
14. de Gramont A, Bosset JF, Milan C, et al. Randomized trial comparing monthly low-dose leucovorin and fluorouracil bolus with bimonthly high-dose leucovorin and fluorouracil bolus plus continuous infusion for advanced colorectal cancer: a French intergroup study. J Clin Oncol 1997;15:808–15.[Abstract/Free Full Text]
15. Douillard JY, Cunningham D, Roth AD, et al. Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer: a multicentre randomised trial. Lancet 2000;355:1041–7.[CrossRef][Medline]
16. de Gramont A, Figer A, Seymour M, et al. Leucovorin and fluorouracil with or without oxaliplatin as first-line treatment in advanced colorectal cancer. J Clin Oncol 2000;18:2938–47.[Abstract/Free Full Text]
17. Vokes EE, Weichselbaum RR, Lippman S, et al. Head and neck cancer. N Engl J Med 1993;328:184–94.[Free Full Text]
18. French Epirubicin Study Group: a prospective randomized trial comparing chemotherapy with cyclophosphamide, fluorouracil, and either doxorubicin or epirubicin. J Clin Oncol 1988;6:679–88.[Abstract]
19. Gamelin E, Boisdron-Celle M, Delva R, et al. Long-term weekly treatment of colorectal metastatic cancer with fluorouracil and leucovorin: results of a multicentric prospective trial of fluorouracil dosage optimization by pharmacokinetic monitoring in 152 patients. J Clin Oncol 1998;16:1470–8.[Abstract/Free Full Text]
20. Den Dunnen JT, Antonarakis SE. Mutation nomenclature extensions and suggestions to describe complex mutations: a discussion. Hum Mutat 2000;15:7–12.[CrossRef][Medline]
21. Collie-Duguid ERS, Johnston SJ, Powrie RH, et al. Cloning and initial characterization of the human DPYD gene promoter. Biochem biophys com 2000;271:28–35.
22. Vrecken P, Van Kuilenburg ABP, Meinsma R, et al. Dihydropyrimidine dehydrogenase (DPD) deficiency: identification and expression of missense mutations C29R, R886H and R235W. Hum Genet 1997;101:333–8.[CrossRef][Medline]
23. Van Kuilenburg ABP, Vrecken P, Riva D. Clinical and biochemical abnormalities in a patient with dihydropyrimidine dehydrogenase deficiency due to homosigosity for the C29R mutation. J Inherit Metab Dis 1999;22:191–2.[CrossRef][Medline]
24. Collie-Duguid ESR, Etienne MC, Milano G, et al. Known variant DPYD alleles do not explain DPD deficiency in cancer patients. Pharmacogenetics 2000;10:217–23.[CrossRef][Medline]
25. Vrecken P, Van Kuilenburg ABP, Meinsma R, et al. Identification of a four-base deletion (delTCAT296–299) in the dihydropyrimidine dehydrogenase gene with variable clinical expression. Hum Genet 1997;100:263–5.[CrossRef][Medline]
26. Kouwaki M, Hamajima N, Sumi S, et al. Identification of novel mutations in the dihydropyrimidine dehydrogenase gene in a Japanese patient with 5-fluorouracil toxicity. Clin Cancer Res 1998;4:2999–3004.[Abstract]
27. Dobritzsh D, Schneider G, Schnakerz K, et al. Crystal structure of dihydropyrimidine dehydrogenase, a major determinant of the metabolism of the anticancer drug, 5-FU. EMBO J 2001;20:650–60.[CrossRef][Medline]
28. Van Kuilenburg ABP, Baars JW, Meinsma R, et al. Lethal 5-fluorouracil toxicity associated with a novel mutation in the dihydropyrimidine dehydrogenase gene. Ann Oncol 2003;14:341–2.[Free Full Text]
29. Mattison L, Johnson MR, Diasio R. A comparative analysis of translated dihydropyrimidine dehydrogenase cDNA; conservation of functional domains and relevance to genetic polymorphisms. Pharmacogenetics 2002;12:133–44.[CrossRef][Medline]
30. Van Kuilenburg ABP, Dobritzsch D, Meinsma R, et al. Novel disease-causing mutations in the dihydropyrimidine dehydrogenase gene interpreted by analysis of the three-dimensional protein structure. Biochem J 2002;364:157–63.[Medline]
31. Yamaguchi K, Arai Y, Kanda Y, et al. Germline mutation of dihydropyrimidine dehydrogenase gene among a Japanese population in relation to toxicity to 5-fluorouracil. Jpn J Cancer Res 2001;92:337–42.[CrossRef]
32. Alderbon A, Kristofferson A, Hammerling U. Determination of single-nucleotide polymorphism by real-time pyrophosphate DNA sequencing. Genome Res 2000;10:1249–58.[Abstract/Free Full Text]
33. Ronaghi M. Pyrosequencing sheds light on DNA sequencing. Genome Res 2001;11:3–11.[Abstract/Free Full Text]
34. Gamelin EC, Danquechin-Dorval EM, Dumesnil Y, et al. Relationship between 5-fluorouracil dose-intensity and therapeutic response in patients with advanced colorectal cancer receiving 5-FU containing infusional therapy. Cancer 1996;77:441–51.[CrossRef][Medline]
35. Van Kuilenburg ABP. Dihydropyrimidine dehydrogenase and the efficacy and toxicity of 5-fluorouracil. Eur J Cancer 2004;40:939–50.[CrossRef][Medline]
36. Hagen WR, Vanoni MA, Rosenbaum K, et al. On the iron-sulfur clusters in the complex redox enzyme dihydropyrimidine dehydrogenase. Eur J Biochem 2000;267:3640–6.[Medline]
37. Gross E, Ullrich T, Seck K, et al. Detailed analysis of five mutations in dihydropyrimidine dehydrogenase detected in cancer patients with 5-fluorouracil-related side effects. Human Mut 2003;22:801–7.
38. Fleming R, Milano G, Thyss A, et al. Correlation between dihydropyrimidine dehydrogenase activity in peripheral mononuclear cells and systemic clearance of fluorouracil in cancer patients. Cancer Res 1992;52:2899–902.[Abstract/Free Full Text]
39. Johnson MR, Yan J, Shao L, et al. Semi-automated radioessay for determination of dihydropyrimidine dehydrogenase (DPD) activity screening cancer patients for DPD deficiency, a condition associated with 5-fluorouracil toxicity. J Chromatogr 1997;696:183–91.[Medline]
40. Van Gennip AH, Van Bree-Blom EJ, Wadman SK, et al. HPLC of urinary pyrimidines for the evaluation of primary and secondary abnormalities of pyrimidine metabolism. In: Hawk GL, editor. Biological biomedical applications of liquid chromatography, III. New York and Basel: Marcel Dekker, Inc.; 1982. p. 285–96.
41. Gamelin E, Boisdron-Celle M, Guerin Meyer V, et al. Correlation between uracil and dihydrouracil plasma ratio, fluorouracil (5-FU) pharmacokinetic parameters, and tolerance in patients with advanced colorectal cancer: a potential interest for predicting 5-FU toxicity and determining optimal 5-FU dosage. J Clin Oncol 1999;17:1105–10.[Abstract/Free Full Text]
42. Mattison LK, Ezzeldin H, Carpenter M, et al. Rapid identification of dihydropyrimidine dehydrogenase deficiency by using a novel 2–13C-uracil breath test. Clin Cancer Res 2004;10:2652–8.[Abstract/Free Full Text]
43. Ahluwalia R, Freimuth R, Mc Leod H, et al. Use of Pyrosequencing to detect clinically relevant polymorphisms in dihydropyrimidine dehydrogenase. Clin Chem 2003;49:1661–4.[Free Full Text]
44. Van Kuilenburg AB, Meinsma R, van Gennip AH. Pyrimidine degradation defects and severe 5-fluorouracil toxicity. Nucleosides Nucleotides Nucleic Acids 2004;23:1371–5.[CrossRef][Medline]
45. Santini D, Vincenzi B, Perrone G, et al. Thymidylate synthase expression in normal colonic mucosa: a predictive marker of toxicity in colorectal cancer patients receiving 5-fluorouracil-based adjuvant chemotherapy. Oncol 2004;67:135–42.[CrossRef]

This article has been cited by other articles:

				E. Gamelin, R. Delva, J. Jacob, Y. Merrouche, J. L. Raoul, D. Pezet, E. Dorval, G. Piot, A. Morel, and M. Boisdron-Celle Individual Fluorouracil Dose Adjustment Based on Pharmacokinetic Follow-Up Compared With Conventional Dosage: Results of a Multicenter Randomized Trial of Patients With Metastatic Colorectal Cancer J. Clin. Oncol., May 1, 2008; 26(13): 2099 - 2105. [Abstract] [Full Text] [PDF]

				M. Schwab, U. M. Zanger, C. Marx, E. Schaeffeler, K. Klein, J. Dippon, R. Kerb, J. Blievernicht, J. Fischer, U. Hofmann, et al. Role of Genetic and Nongenetic Factors for Fluorouracil Treatment-Related Severe Toxicity: A Prospective Clinical Trial by the German 5-FU Toxicity Study Group J. Clin. Oncol., May 1, 2008; 26(13): 2131 - 2138. [Abstract] [Full Text] [PDF]

				L. Gamelin, O. Capitain, A. Morel, A. Dumont, S. Traore, L. B. Anne, S. Gilles, M. Boisdron-Celle, and E. Gamelin Predictive Factors of Oxaliplatin Neurotoxicity: The Involvement of the Oxalate Outcome Pathway Clin. Cancer Res., November 1, 2007; 13(21): 6359 - 6368. [Abstract] [Full Text] [PDF]

				H. Gurney and R. Shaw Obesity in Dose Calculation: A Mouse or an Elephant? J. Clin. Oncol., October 20, 2007; 25(30): 4703 - 4704. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Morel, A. Articles by Gamelin, E. Search for Related Content PubMed PubMed Citation Articles by Morel, A. Articles by Gamelin, E.