Eur opean Rev iew for Med ical and Pharmacol ogical Sci ences 2014; 18: 2772-2776 Correlation between dihydropyrimidine dehydrogenase and efficacy and toxicity of fluoropyrimidine drugs

X.-Q. LIU 1,2 , M. ZHUANG 1, Z. WANG 2, R.M. HUBER 2

1Department of Medical Oncology, No. 1 Renmin Hospital of Lianyungang, Lianyungang, P.R. China 2Division of Respiratory Medicine and Thoracic Oncology, Department of Internal Medicine V, Tho - racic Oncology Centre Munich, Ludwig-Maximilians Universität München, Germany

Abstract. At present, fluoropyrimidine, mor tissue. 5-FU in vivo is converted into 5-fluo - based on 5-fluorouracil (5-FU), remains one of ro-2-deoxyuridine nucleotides, which inhibit the the most frequently prescribed chemotherapeu - thymine nucleotide synthase and prevent trans - tics drugs for the treatment of cancer. Dihy - dropyrimidine dehydrogenase (DPD) is the rate- formation of the deoxyuridine nucleotide into the limiting enzyme in the catabolism of 5-FU, and deoxy-thymidine nucleotides, accordingly sup - DPD enzymatic activities are usually varied dra - pressing DNA synthesis. In addition, inhibiting matically from individual to individual, includ - RNA synthesis can be achieved by preventing ing both the intrapatient differences and the in - uracil and orotic acid incorporated into RNA. terpatient variability. There is a certain correla - These kinds of drugs have been widely used in tion between the DPD activity and efficacy and chemotherapy of a variety of malignant tumors toxicity following the administration of fluo - ropyrimidine drugs. Partial or complete loss of such as gastrointestinal tumors, head and neck DPD activity can lead to serious or even lethal tumors, and breast tumors, in which toxicity is toxicity. In this article, we review the relation - mainly displayed as gastrointestinal reaction, oral ship between DPD activity and efficacy and tox - mucositis , bone marrow suppression , nervous icity following the administration of fluoropy - system toxicity, and so on. Clinical trials, thus, rimidine drugs, and also the structure, function, far demonstrate that under the same treatment and characteristics of DPD. We report here that measurement of DPD ac - condition there is a great variability for different tivity may become a strategy and be paid much patients related to efficacy and toxicity of anti - attention to predict the efficacy and toxicity pri - cancer drugs. In recent years, scientists have re - or to starting a fluoropyrimidine-based therapy. ported that the DPD enzymes play an important role in determination of the efficacy and the toxi - Key Words: city of 5-FU drugs. Individual differences of Dihydropyrimidine dehydrogenase, Fluoropyrimidine DPD enzyme activity is one of the most impor - drugs, Efficacy, Toxicity. tant reasons resulting in individual variability of efficacy and toxicity treated by 5-FU drugs.

Introduction Structure, Function and The fluoropyrimidine drugs have been widely Mechanism of DPD applied to treatment of antitumor including 5-fluo - rouracil (5-FU), tegafur, compound tegafur Tegafur- Structure of DPD Uracil , Carmofur , floxuridine , deoxy-fluorouridine, DPD is an enzyme encoded by the dihydropy - chemotherapy-targeted capecitabine , S-1 (firstly rimidine dehydrogenase gene ( DPYD ), which is appearing on the market in Japan), and Tiji’ao located on the short (p) arm of chromosome 1 at capsule (made in China), among of which 5-FU position 22 (1p22) containing 23 exons (about has been firstly conducted in clinical trial and the 950 kb) 1,2 . The wild-type DPYD gene encodes a other drugs are derivatives from 5-FU. These protein composed of 1025 amino acid with two 3 drugs in vivo need to be converted into 5-FU and identical subunits and a molecule (Mw = 150 kD) play a role in the treatment of cancer through the containing binding sites with nicotinamide ade - natural hydrolysis and metabolism by the P450 nine dinucleotide phosphate, flavin adenine dinu - enzymes or by a specific enzyme system in tu - cleotide , and pyrimidine alkali .

2772 Corresponding Author: Rudolf M. Huber, MD; e-mail: [email protected] Correlation between dihydropyrimidine dehydrogenase and efficacy and toxicity of fluoropyrimidine drugs

Acting Mechanism of DPD slow onset or rapid emergence of epilepsy and The enzyme kinetic studies have shown that disturbance of consciousness, are performed with DPD takes effect by nonclassical two sites ping- the cerebellar ataxia and encephalopathy . pong kinetic mechanism: combination of two sep - arate sites with NADPH/NADP + and pyrimidine Distribution Feature of DPD alkali or 5-FU, respectively. After reduction of DPD is widely distributed in most tissue of FAD using NADPH as hydrogen donor, electrons people, especially exhibiting the highest activity are most likely transferred to FMN via [4Fe-4s] in the liver tissue and peripheral blood mononu - cluster, resulting in reduction of pyrimidine alka - clear cells (PBMC). In addition, it can also be li [4] . 5-FU is reduced to the dihydrofluorouracil found in the tumor and inflammatory tissue. The

(FUH 2). After ring cleavage through dihydropy - DPD activity distribution in PBMCs shows a rimidinase, 5-fluoro- β-ureide propionic acid (FU - Gaussian distribution, but a large degree of in - PA) is produced, which is finally synthesized into terindividual variation is observed, even up to 5- 5-fluoro- β-alanine (FBAL) under catalysis of β- 26 times. A meta-analysis of over 1200 cases of alanine synthase and excreted by the kidney. patients showed that more than 30% of patients suffered severe drug-related toxicity after chemotherapy using 5-FU 8. It was estimated that Clinical Manifestations and the incidence of low DPD activity among the to - Distribution Feature of DPD Deficiency tal population was about 3-5% 9,10 . It was also re - ported that differences of DPD activity among Clinical DPD Deficiency different ethnic groups have been noted. Pheno - Clinical DPD deficiency is observed with both typic and genotypic analysis showed that, among the phenotypes and genotypes: (1) phenotype is East Asia (Japanese, Taiwanese, South Korean), the apparent abnormal birth after onset, manifest - South Asia (Indian, Pakistani, Sri Lankan), ed as spasms, bradykinesia , and mental retarda - Africa (Egyptian, Kenyan, Ghanaian), Europe tion; (2) genotype is the usually asymptomatic, (whites), and the United States (whites and only performing severe toxicity during African-American) , the incidence of low DPD chemotherapy treated by fluorinated pyrimidine activity remains variable, for example, the South drugs 5. These two kinds of manifestations are Koreans have a higher level of DPD enzyme ac - based on different molecular genetics . In 1985, tivity while a higher incidence rate of low DPD Tuchman et al [6] first reported that a breast cancer activity in African-American women can be ob - patient, having induced by severe diarrhea, served [2,11-16] . At present, the definition of low myelosuppression, neurotoxicity, and disturbance DPD activity threshold value has not yet reached of consciousness , was treated by regular 5-FU a consensus. Prior to using the fluoropyrimidine chemotherapy. Detectable pyrimidine concentra - drugs, as a treatment in cancer, more attention tions in the blood and urine samples of patient should be paid on how much of lowest DPD ac - increased abnormally, strongly suggesting the tivity it can reach, which is not known and needs DPD activity deficiency . Thereafter, there have to be further studied . been reports on toxicity due to partial or com - plete lacking of DPD activity during chemothera - Methodology of DPD Assay py treated by 5-FU. Its toxicity is mainly mani - There are several techniques for detecting the fested as the gastrointestinal tract and bone mar - DPD in blood and tissue at a gene or protein lev - row suppression, but severer (mostly grade IV el. The most common approach is to measure than patients with normal enzyme activity, even both the dihydrouracil (UH 2) and the uracil (U) it is hard to avoid toxicity by dose reduction. In concentrations in plasma by a high-performance addition, neurotoxicity in patients with DPD ac - liquid chromatography (HPLC) and to evaluate tivity deficiency is relatively common. In 1999, the DPD activity in vivo by UH 2/U ratio. This Milano et al 7 reported that, among the whole method is simpler and accurate for DPD activity group of 11 patients with DPD activity deficien - assay comparing to the radio-enzymatic analysis cy, seven had significant neurotoxicity, and lead - requiring large volume of radio substrates and ing to patient death in two cases, moreover point - time. Although the highest DPD enzyme activity ing out that women are particularly prone to is known in the liver, it is difficult to obtain liver DPD enzyme deficiency. Typical neurotoxic pa - sample from patient and is seldom used. A new tients, having some neurological symptoms of technique was reported to speculate the DPD en -

2773 X.-q. Liu, M. Zhuang, Z. Wang, R.M. Huber zyme activity by quantifying peripheral DPYD Correlation Between DPD mRNA according to a real-time quantitative RT- and Efficacy and Toxicity PCR Reverse transcription polymerase chain re - of Fluoropyrimidine Drugs action. However, there is a conflict among the re - cent several studies and conclusions 17-21 , and Antitumor Mechanism whether replacing DPD enzyme activity assay by of Fluoropyrimidine Drugs quantification of peripheral DPYD mRNA still Currently, in addition to the traditional 5-FU, needs to be further explored. It was also report - capecitabine and S-1 (also called Tiji’ao capsule, ed 22 that the final metabolite of 5-FU, FBAL, made in China) have been widely used in the may be used to predict the DPD activity in vivo , treatment of cancer. Capecitabine (namely, n-4- showing that low FBAL level indicates the low pentyloxycarbonyl-5 ′-deoxy-5-fluorocytidine, DPD activity. In 2004, Mattison et al 23 reported a CAP), belonging to a thymidylate synthase (TS) novel, developed, noninvasive, testing method. inhibitor group, is a prodrug chemically synthe - 13 13 After oral administration of 2- C-uracil, CO 2 sized by FU. It can be absorbed rapidly through 12 and CO 2 level in exhaled breath were detected the intestinal mucosa as an intact molecule due to by using IR spectroscopy, and it was concluded containing urethane structure. CAP is a type of 13 that decreasing CO 2 in exhaled breath was asso - oral-administrated fluoropyrimidine nucleoside ciated with partial or complete deficiency of drugs with targeted effect . It can be selectively ac - DPD activity. tivated in tumor tissue and produce high concen - trations of the active cell toxic substances, there - Relationship Between DPD Gene by improving the tolerance of cancer patients and Polymorphism and its Decreased Activity maximizing the anticancer activity. S-1 firstly ap - DPD is encoded by the DPYD, in which nu - peared on the market in Japan and applied in clin - cleotide mutation may cause changes of DPD ical trials in 1994. It has now been approved for structure and activity. To date, genetic polymor - the treatment of advanced gastric cancer, head phism of DPD encoded by DPYD has been grad - and neck cancer, colorectal cancer, and metastatic ually reported. More than 40 mutation sites have breast cancer. S-1 is an oral anticancer drug with been identified and occurrence frequency in some compound ingredients, composed of tegafur, sites has been studied too. The splice-site muta - gimestat, and otastat potassium in a molar ratio of tion IVS14+1G > A was firstly found in DPYD 1:0.4:1, of which gimestat can selectively inhibit gene sequence variant, which is the result of GT the hepatic DPD activity and prevent the catabo - to AT alternation in the nucleotide at the exon 14 lism of 5-FU in vivo , maintaining 5-FU at a high - acceptor splice site leading to lacking of 165 bp er concentration in plasma and tumor tissue and segment in exon 14 and skipping of exon 14 im - prolonging the half-life of 5-FU, while otastat mediately up-stream of the mutated splice donor potassium is also an inhibitor that is intended to site . As a result, the DPD activity is decreased mitigate the 5-FU-related gastrointestinal toxicity markedly. Kuilenburg et al [5] investigated that by preventing the phosphorylation of 5-FU . In re - among the mutations responsible for DPD activity cent years, there has been reported to cause severe decreasing and deficiency, the IVS14+1G > A toxicity during chemotherapy treated by the S-1, mutation was mostly occurred showing a preva - because the gimestat inhibits DPD activity, there - lence in Nordic populations. Jia et al [24] have in - by resulting in severe toxicity for patients with vestigated the polymorphism of three mutation DPD deficiency . sites on DPYD gene (IVS14+1G → A, Exon13 A1627G , Exon11 G1156T) in Chinese Han Relationship Between DPD and Efficacy ethinicity groups. The results showed that the of Fluoropyrimidine Drugs polymorphism of Exon13 A1627G mutation site In 1999, Ishikawa et al 25 and Kirihara et al 26 is present in Chinese Han ethinicity groups, while reported that, in tumor cells in vitro , gene expres - there is no polymorphism for the IVS14+1G → A sion and activity of DPD were related to the ther - and Exon11 G1156T mutation sites. Some of the apy sensitivity of 5-FU. High DPD mRNA ex - mutations, such as A1627G and G2194A sites, do pression and activity level may give rise to resis - not cause a decreased DPD activity . Analysis on tance to 5-FU drugs, otherwise leading to severe polymorphism of these mutation sites and rela - toxicity. Dong et al 27 have studied the relation - tionship between DPD activity and mutation of ship between DPD level and serum concentration these sites needs to be explored in depth. of 5-FU, efficacy and toxicity in colorectal can -

2774 Correlation between dihydropyrimidine dehydrogenase and efficacy and toxicity of fluoropyrimidine drugs cer patients. DPD level and serum concentration so the potential patients with severe toxicity can of 5-FU significantly vary among the different be treated with lower initial doses of chemothera - colorectal cancer patients. DPD level is negative - py drugs to avoid the occurrence of severe toxici - ly correlated with serum 5-FU concentration and ty. Along with the emergence of personalized toxicity, while serum concentration of 5-FU is chemotherapy era, it is believed that the patients positively related to adverse response and treat - can be treated according to different tumor iden - ment efficacy, which correspond well to most tity card, which is now within our grasp. conclusions reported previously . Horiguchi et al 28 investigated DPD expression and prognosis in a total of 191 patients with invasive breast cancer ––––––––––––– –– ––– –-–– Conflict of interest treated by 5-FU, showing that patients with DPD expression-positive tumors had a significantly The Authors declare that they have no conflict of interests. poorer prognosis in disease-free and overall sur - vival compared to those with DPD-negative tu - mors (p < 0.05). But, the efficacy of 5-FU was correlated with the other enzyme (such as TS) References expression involved in metabolism. In addition, 1) TAKAI S, F ERNANDEZ -S ALGUERO P, K IMURA S, K IMURA S, DPD activity may be regulated by post-transcrip - GONZALEZ FJ, Y AMADA K. Assignment of the human tional level, which needs to be further studied. dihydropyrimidine dehydrogenase gene (DPYD) to chromosome region 1p22 by fluorescence in Relationship Between DPD and Toxicity situ hybridization. Genomics 1994; 24: 613-614. of Fluoropyrimidine Drugs 2) WEI X, E LIZONDO G, S APONE A, M CLEOD HL, R AUNIO H, F ERNANDEZ -S ALGUERO P, G ONZALEZ FJ. Characteri - A reduced DPD activity will result in increas - zation of the human dihydropyrimidine dehyrdo - ing serum concentrations of 5-FU drugs, thus, in - genase gene. Genomics 1998; 51: 391-400. creasing the risk of patients suffering from severe 3) DIASIO RB, L U Z. Dihydropyrimidine dehydrogenase toxicity. Van Kuilenburg et al 29 demonstrated that activity and fluorouracil chemotherapy. J Clin On - in 59% of the cases (22 of 37 patients suffering col 1994; 12: 2239-2242. from severe toxicity after administration of 5-FU 4) DOBRITZSCH D, S CHNEIDER G, S CHNACKERZ KD, drugs) , DPD activity deficiency can be detected LINDQVIST Y. Crystal structure of dihydropyrimidine dehydrogenase, a major determinant of the phar - in PBMCs. Comparing the two groups of patients macokinetics of the anti-cancer drug 5-fluo - with a low DPD activity and a normal DPD ac - rouracil. EMBO J 2001; 20: 650-660. tivity, no differences in hematological, gastroin - 5) VAN KUILENBURG AB, V REKEN P, A BELING NG, B AKKER testinal, or flu-like symptoms were observed be - HD, M EINSMA R, V AN LENTHE H, D E ABREU RA, tween both groups, with the exception of grade SMEITINK JA, KAYSERILI H, A PAK MY, C HRISTENSEN E, HOLOPAINEN I, P ULKKI K, R IVA D, B OTTEON G, H OLME E, IV neutropenia, indicating that reducing DPD ac - TULINIUS M, K LEIJER WJ, B EEMER FA, D URAN M, N IEZEN - tivity will significantly suppress bone marrow. KONING KE, S MIT GP, J AKOBS C, S MIT GP, J AKOBS C, S MIT Dong et al 27 have studied serum levels of DPD LM, V AN GENNIP AH. Genotype and phenotype in patients with dihydropyrimidine dehydrogenase distributed in 72 patients, showing that UH 2/U ratio is negatively correlated with serum concen - deficiency. Hum Genet 1999; 104: 1-9. tration of 5-FU. UH /U ratio in patients suffered 6) TUCHMAN M, S TOECKELER JS, K IANG DT, O'D EA RF, R AM - 2 NARAINE ML, M IRKIN BL. Familial pyrimidinemia and from grades II-IV oral mucositis and diarrhea is pyrimidinuria associated with severe fluorouracil lower than that in patients with grades 0-I. toxicity. N Engl J Med 1985; 313: 245-249. 7) MILANO G, E TIENNE MC, P IERREFITE V, B ARBERI -H EYOB M, D EPORTE -F ETY R, R ENÉE N. Dihydropyrimidine de - Conclusions hydrogenase deficiency and fluorouracil-related toxicity. Br J Cancer 1999; 79: 627-630. 8) [N O AUTHORS LISTED ] Toxicity of fluorouracil in pa - DPD plays an important role during tients with advanced colorectal cancer: effect of chemotherapy treated by 5-FU drugs. Detecting administration schedule and prognostic factors. and assessing DPD activity may become a strate - Meta-Analysis Group In Cancer. J Clin Oncol gy to improve the efficacy and avoid the severe 1998; 16: 3537-3541. toxicity prior to starting a fluoropyrimidine- 9) ETIENNE MC, L AGRANGE JL, D ASSONVILLE O, F LEMING R, THYSS A, R ENÉE N, S CHNEIDER M, D EMARD F, M ILANO based therapy . At present, it is imperative for G. Population study of dihydropyrimidine dehydro - comprehensive understanding the variability of genase in cancer patients. J Clin Oncol 1994; 12: individual tolerance to fluoropyrimidine drugs, 2248-2253.

2775 X.-q. Liu, M. Zhuang, Z. Wang, R.M. Huber

10) LU Z, Z HANG R, C ARPENTER JT, D IASIO RB. Decreased 21) BOISDRON -C ELLE M, R EMAUD G, T RAORE S, P OIRIER A L, dihydropyrimidine dehydrogenase activity in a GAMELIN L, M OREL A, G AMELIN E. 5-Fluorouracil- population of patients with breast cancer: implica - related severe toxicity: a comparison of different tion for 5-fluorouracilbased chemotherapy. Clin methods for the pretherapeutic detection of dihy - Cancer Res 1998; 4: 325-329. dropyrimidine dehydrogenase deficiency. Cancer 11) WEI X, M CLEOD HL, M CMURROUGH J, G ONZALEZ FJ, Lett 2007; 249: 271-82. FERNANDEZ -S ALGUERO P. Molecular basis of the hu - 22) FURUHATA T, K AWAKAMI M, O KITA K, K IMURA Y, K IHARA man dihydropyrimidine dehydrogenase deficiency C, T SURUMA T, O HMURA T, Y AMAGUCHI K, H ATA F, K AT - and 5-fluorouracil toxicity. J Clin Invest 1996; 98: SURAMAKI T, S ASAKI K, H IRATA K. Plasma level of a 5- 610-615. fluorouracil metabolite, fluoro-beta-alanine corre - 12) SOHN DR, C HO MS, C HUNG PJ. Dihydropyrimidine lates with dihydropyrimidine dehydrogenase activ - dehydrogenase activity in a Korean population. ity of peripheral blood mononuclear cells in 5-fluo - Ther Drug Monit 1999; 21: 152-154. rouracil treated patients. J Exp Clin Cancer Res 2006; 25: 79-82. 13) YAMAGUCHI K, A RAI Y, K ANDA Y, A KAGI K. Germline mutation of dihydropyrimidine dehydrogenase 23) MATTISON LK, E ZZELDIN H, C ARPENTER M, M ODAK A, gene among a Japanese population in relation to JOHNSON MR, D IASIO RB. Rapid identification of di - toxicity to 5-fluorouracil. Jpn J Cancer Res 2001; hydropyrimidine-dehydrogenase deficiency by us - 92: 337-342. ing a novel 2-13C-uracil breath test. Clin Cancer Res 2004; 10: 2652-2658. 14) OGURA K, O HNUMA T, M INAMIDE Y, M IZUNO A, N ISHIYA - MA T, N AGASHIMA S, K ANAMARU M, H IRATSUKA A, 24) JIA L, L I JH, C AI JP, G UO J. Survey of the polymor - WATABE T, U EMATSU T. Dihydropyrimidine dehydro - phism of dihydropyrimidine dehydrogenase gene genase activity in 150 healthy Japanese volun - in the Han nationality population. J China-Japan teers and identification of novel mutations. Clin Friendship Hospital 2005; 19: 202-204, 207. Cancer Res 2005; 11: 5104-5111. 25) ISHIKAWA Y, K UBOTA T, O TANI Y, W ATANABE M, T ERAMO - 15) MORSMAN JM, S LUDDEN J, A MEYAW MM, G ITHANG 'A J, TO T, K UMAI K, K ITAJIMA M, T AKECHI T, O KABE H, INDALO A, O FORI -A DJEI D, M CLEOD HL. Evaluation of FUKUSHIMA M. Dihydropyrimidine dehydrogenase dihydropyrimidine dehydrogenase activity in activity and messenger RNA level may be related South-west Asian, Kenyan and Ghanaian popula - to the antitumor effect of 5-fluorouracil on human tions. Br J Clin Pharmacol 2000; 50: 269-272. tumor xenografts in nude mice. Clin Cancer Res 16) SAIF MW, M ATTISON L, C AROLLO T, E ZZELDIN H, D IASIO 1999; 5: 883-889. RB. Dihydropyrimidine dehydrogenase deficiency 26) KIRIHARA Y, Y AMAMOTO W, T OGE T, N ISHIYAMA M. Dihy - in an Indian population. Cancer Chemother Phar - dropyrimidine dehydrogenase, multidrug resis - macol 2006; 58: 396-401. tance-associated protein, and thymidylate syn - 17) SECK K, R IEMER S, K ATES R, U LLRICH T, L UTZ V, H ARBECK thase gene expression levels can predict 5-fluo - N, S CHMITT M, K IECHLE M, D IASIO R, G ROSS E. Analy - rouracil resistance in human gastrointestinal can - sis of the DPYD gene implicated in 5-fluorouracil cer cells. Int J Oncol 1999; 14: 551-556. catabolism in a cohort of Caucasian individuals. 27) DONG QM, H E YJ, L I S, Z HOU ZM, Z HANG L, Z HOU Clin Cancer Res 2005; 11: 5886-5892. ZW, X IA ZJ, L I YY. Relationship of serum level of di - 18) TAKECHI T, O KABE H, F UJIOKA A, M URAKAMI Y, F UKUSHI - hydropyrimidine dehydrogenase and serum con - MA M. Relationship between protein levels and centration of 5-fluorouracil to treatment response gene expression of dihydropyrimidine dehydroge - and adverse events in colorectal cancer patients. nase in human tumor cells during growth in cul - Chinese J Cancer 2005; 24: 483-487. ture and in nude mice. Jpn J Cancer Res 1998; 28) HORIGUCHI J, Y OSHIDA T, K OIBUCHI Y, I IJIMA K, N INOMIYA 89: 1144-1153. J, T AKEI H, Y OKOE T, I INO Y, M ORISHITA Y. DPD activity 19) JOHNSON MR, W ANG K, S MITH JB, H ESLIN MJ, D IASIO and immunohistochemical DPD expression in hu - RB. Quantitation of dihydropyrimidine dehydroge - man breast cancer. Oncol Rep 2004; 11: 65-72. nase expression by real-time reverse transcription 29) VAN KUILENBURG AB, H AASJES J, R ICHEL DJ, Z OETEKOUW polymerase chain reaction. Anal Biochem 2000; L, V AN LENTHE H, D E ABREU RA, M ARING JG, V REKEN 278: 175-184. P, VAN GENNIP AH. Clinical implications of dihy - 20) JOHNSON MR, W ANG K, D IASIO RB. Profound dihy - dropyrimidine dehydrogenase (DPD) deficiency in dropyrimidine dehydrogenase deficiency resulting patients with severe 5-fluorouracil-associated toxi - from a novel compound heterozygote genotype. city: identification of new mutations in the DPD Clin Cancer Res 2002; 8: 768-774. gene. Clin Cancer Res 2000; 6: 4705-4712.

2776