Research Dose-Dependent Association between UGT1A1*28 ... · Begg and Mazumdar (28) and Egger (29). The studies included were stratified into different dose groups. Irinotecan dose

3832

Published OnlineFirst June 18, 2010; DOI: 10.1158/1078-0432.CCR-10-1122

Predictive Biomarkers and Personalized Medicine Clinical
Cancer
Research

Dose-Dependent Association between UGT1A1*28 Genotypeand Irinotecan-Induced Neutropenia: Low Doses AlsoIncrease Risk

Zhe-Yi Hu1, Qi Yu2, Qi Pei3, and Cheng Guo2

Abstract

Authors' ASchool, Chand ClinicTong UniveInstitute, InChangsha,

Note: SuppResearch O

Z.-Y. Hu an

Corresponand GraduRoad, Shan86-21-5080

doi: 10.115

©2010 Am

Clin Canc

Down

Purpose: A previous meta-analysis showed that the association between the UGT1A1*28 genotype andirinotecan-induced neutropenia was influenced by irinotecan dose and that the risk of neutropenia wassimilar at low doses for patients with all genotypes. However, the sample sizes for the low- and high-dosegroups were small. Because additional studies have now been reported, an updated and refined meta-analysis is needed.Experimental Design: Meta-analyses were done to assess the relationship between UGT1A1*28 and

neutropenia. The association between UGT1A1*28 and the relative extent of glucuronidation (REG) ofSN-38 was also examined. The studies included were stratified into different dose groups.Results: A total of 1,998 patients were included for the analysis of neutropenia and 581 patients were

included for REG. The risk of neutropenia at low doses was significantly higher among patients with aUGT1A1*28/*28 genotype than among those carrying the UGT1A1*1 allele(s) [relative risk (RR), 2.43;95% confidence interval, 1.34-4.39; P = 0.003]. In addition, the RR of neutropenia at low doses was com-parable with that at medium doses (2.43 versus 2.00). The RR of neutropenia at high doses was signif-icantly higher than that at low and medium doses (7.22 versus 2.04). We found the weighted meandifference of REG (UGT1A1*1/*1 or UGT1A1*1/*28 versus UGT1A1*28/*28) increased with increasingdose of irinotecan.Conclusions: In conclusion, the UGT1A1*28/*28 genotype was associated with an increased risk of

neutropenia not only at medium or high doses of irinotecan but also at low doses. The dose-dependentmanner of SN-38 glucuronidation explained why the association between UGT1A1*28 and neutropeniawas dose dependent. Clin Cancer Res; 16(15); 3832–42. ©2010 AACR.

Irinotecan is approved for use in combination with5-fluorouracil and leucovorin for first-line treatment of me-tastatic colorectal cancer (1). Irinotecan efficacy is depen-dent on activation by carboxyesterases to form the activemetabolite SN-38. The major route of SN-38 eliminationis via glucuronidation by hepatic UGT1A enzymes (2).UGT1A1*28 is a common allele with seven TA repeats

in the TATA box of the UGT1A1 promoter compared with

ffiliations: 1Shanghai Institute of Materia Medica and Graduateinese Academy of Sciences, and 2Department of Pharmacyal Pharmacology, Sixth People's Hospital, Shanghai Jiaorsity, Shanghai, PR China; and 3Pharmacogenetics Researchstitute of Clinical Pharmacology, Central South University,PR China

lementary data for this article are available at Clinical Cancernline (http://clincancerres.aacrjournals.org/).

d Q. Yu share co-first authorship.

ding Author: Zhe-Yi Hu, Shanghai Institute of Materia Medicaate School, Chinese Academy of Sciences, 555 Zuchongzhighai 201203, PR China. Phone: 86-21-50802121-2102; Fax:2121-2102; E-mail: [email protected].

8/1078-0432.CCR-10-1122

erican Association for Cancer Research.

er Res; 16(15) August 1, 2010

Researcon Januarclincancerres.aacrjournals.org loaded from

the wild-type allele, which has six TA repeats. UGT1A1*28has been shown to be associated with decreased SN-38glucuronidation in humans (3–11). The association be-tween the UGT1A1*28 genotype and irinotecan-inducedneutropenia has been extensively studied (3–5, 12–23).Seven of these studies found that UGT1A1*28/*28 patientshad an elevated risk of neutropenia compared with thosecarrying UGT1A1*1 allele(s) (refs. 4, 12, 13, 17, 19–21).In 2005, the U.S. Food and Drug Administration (FDA)

recommended that Pfizer amend the package insert of ir-inotecan to warn of the elevated risk of neutropenia forUGT1A1*28/*28 patients. A subsequent meta-analysisshowed that the association between UGT1A1*28 andneutropenia was influenced by the dose of irinotecanand that the risk of neutropenia was similar at lower dosesfor patients with all genotypes (1). It was noteworthy thatthe sample sizes for the low- and high-dose groupswere small (229 patients and 81 patients). In addition,the studies included in this meta-analysis seemed to beweighted by sample size. This weighted method is notthe recommended method in the Cochrane Handbookfor Systematic Reviews of Interventions version 5.0.2(http://www.cochrane-handbook.org/). Because additional

h. y 27, 2021. © 2010 American Association for Cancer

http://clincancerres.aacrjournals.org/

Translational Relevance

In 2005, the Food and Drug Administration rec-ommended that patients with a UGT1A1*28/*28genotype should receive a lower starting dose of iri-notecan. Then came conflicting results from ameta-analysis by Hoskins et al., suggesting thatUGT1A1*28 had no effect on irinotecan-induced neu-tropenia among patients treated with a low dose ofirinotecan. In this study, we did an updated and re-fined meta-analysis. We found the UGT1A1*28/*28genotype to be associated with a 2-fold increased riskof neutropenia not only at medium doses but at lowdoses as well. We suggest that genotype-based dosingmay be necessary for patients treated with low dosesof irinotecan.

UGT1A1*28 Genotype and Irinotecan-Induced Neutropenia


studies have now been reported, an updated and refinedmeta-analysis is needed. Moreover, the pharmacokineticmechanism for the dose-dependent association betweenUGT1A1*28 and neutropenia is currently unclear.In this study, we carried out an updated and refined

meta-analysis to study the association between theUGT1A1*28 genotype and irinotecan-induced neutrope-nia. To explore the pharmacokinetic mechanism for thedose-dependent association between UGT1A1*28 andneutropenia, the association between UGT1A1*28 andthe relative extent of glucuronidation (REG) of SN-38was also examined.

Materials and Methods

Search strategy and selection criteriaTwo investigators (ZYH, QY) independently searched

Medline, PubMed, and Embase (from 1980 until April15, 2010) databases using the terms “irinotecan,”“UGT1A1,” and “neutropenia (or pharmacokinetics)”.Furthermore, we reviewed citations in the retrieved articlesto search for additional relevant studies. Data derivedfrom abstracts were also used.A priori we defined strict criteria for inclusion of studies.

Studies were included if (a) they could be defined as clin-ical trials, (b) the exposure of interest was the UGT1A1*28genotype, (c) the outcome of interest was irinotecan-induced neutropenia (grade III-IV) or REG of SN-38, and(d) the numbers of patients with and without neutropeniawere provided (or could be calculated by relevant data).We excluded studies that were not published in English,studies that included <20 patients, studies that included<4 UGT1A1*28/*28 patients (a criterion not consid-ered when we study the association between UGT1A1*28and REG), studies that reported hematologic toxicityinstead of more specific neutropenia, and studies thatincluded children patients.

www.aacrjournals.org

Researcon Januarclincancerres.aacrjournals.org Downloaded from

Data extractionThe following information was abstracted from in-

cluded publications: study design, year of publication,race, irinotecan dose, number of patients with and withoutneutropenia (grade III-IV) in each genotype group(UGT1A1*1/*1, UGT1A1*1/*28, and UGT1A1*28/*28),REG values in each genotype group (converted to the unitsof ng × h/mL versus ng × h/mL), mutation detection meth-od, plasma sampling scheme, analytical method, andpotential confounders.Two irinotecan-containing regimens were given to pa-

tients in the N9741 study (21), and in our analyses, weanalyzed the patients treated with each regimen as twoseparate samples. Two irinotecan regimens (250 mg/m2

and 350 mg/m2) were administered to patients in anoth-er study (13). In this study, only three UGT1A1*28/*28patients were included in the 350 mg/m2-dose group.Hence, this group was excluded from our analysis ac-cording to the inclusion criterion. The patients treatedwith different regimens were analyzed as one sample ifseparate data were not available. If the patients in a trialreceived different irinotecan doses and only combinedtoxicity-related data (or REG data) were available, wecalculated the average dose. In one study (16), thenumbers of chemotherapy cycles with neutropeniain each genotype group were provided instead of thenumber of patients. We calculated the number ofpatients as C × tN/tC, where C is the number of chemo-therapy cycles with neutropenia in each genotype group,tC is the total number of chemotherapy cycles withneutropenia, and tN is the total number of patientswith neutropenia. REG values were reported as mediansand ranges instead of means and SDs in three studies(5, 6, 9); we imputed the means and SDs as describedby Hozo et al. (24). The REG values were obtained byvisual measurement of the figures in two studies (10,11). The REG values obtained from UGT1A1*1/*28patients and UGT1A1*1/*1 patients were combined bythe formula suggested in the Cochrane Handbook forSystematic Reviews of Interventions version 5.0.2(http://www.cochrane-handbook.org/).

Assessment of study qualityThe use of a quality scoring system in meta-analyses of

observational studies is controversial. Methodologic com-ponents of study designs, rather than aggregate scoresthemselves, may be important (25). Here, we did not as-sign a single grade or scores to represent the quality of astudy. Instead, we focused on certain items that are reflec-tive of methodologic and reporting quality of the studiesas delineated in the published guidelines for reportingof pharmacogenetic studies (26). In addition, we paidattention to other quality criteria that were specific toour study. These issues of concern were source of popula-tion, mutation detection method, analytical method,plasma sampling scheme, type of cancer, chemotherapyregimens, and grade criteria for neutropenia (see Supple-mentary Tables S1 and S2).

Clin Cancer Res; 16(15) August 1, 2010 3833



Hu et al.

3834


Statistical analysisWith regard to the association between UGT1A1*28 and

neutropenia, the effect measures of interest were relativerisks (RR). The effect measures of interest were meandifferences (MD) for the association between UGT1A1*28and REG of SN-38. Statistical heterogeneity among studieswas evaluated using the χ2 test, P values, and I2 statistics(27). We considered both the presence of significantheterogeneity at the 10% level of significance and valuesof I2 >56% as an indicator of significant heterogeneity.A fixed effects model was used to obtain summary RR(Mantel-Haenszel method) or weighted mean difference(WMD; inverse variance method). We evaluated potentialpublication bias statistically with the methods described byBegg and Mazumdar (28) and Egger (29).The studies included were stratified into different dose

groups. Irinotecan dose levels were pooled into the fol-lowing three subgroups: low (<150 mg/m2), medium(150-250 mg/m2), and high (≥250 mg/m2) doses on the

Clin Cancer Res; 16(15) August 1, 2010


basis of the three most commonly used dosage regimens(1). We used meta-regression analyses to investigate theeffect of irinotecan dose on the association betweenUGT1A1*28 and neutropenia (or REG). Hardy-Weinbergequilibrium (HWE) was tested using χ2 test. We conducteda sensitivity analysis in which one or two studies were re-moved, and the rest were analyzed to evaluate whetherthere was a statistically significant effect on the results.All statistical tests were two-sided. Meta-analysis was donewith Stata version 10.1 (Stata Corp.).

Results

Study characteristics and methodologic quality(association between UGT1A1*28 and neutropenia)A total of 68 potentially relevant studies were evaluated

(Fig. 1 shows the numbers of studies evaluated at eachstage). Fifteen clinical trials including 1,998 patients wereidentified. Table 1 details the studies' characteristics. Three

Clinical Cancer Research


Fig. 1. Studies evaluated at eachstage of the meta-analysis(*, supplementary search includesreference lists and contactingauthors).




included studies were published as abstracts (21–23).Patients were predominantly Caucasians in 11 trials. Fourtrials did not clearly report the race of the participants (16,20, 22, 23). The patients in these four trials, however,might have been Caucasians because these trials were con-ducted in Europe or America and the reported frequency ofthe UGT1A1*28 allele was similar to that of Caucasians.No deviation from HWE was detected in any of the identi-fied studies (P > 0.05, χ2 test).Quality-assessment tables are shown in full in Sup-

plementary Table S1. Study sample sizes were small(median, 89; range, 20-628). Of the 15 trials, 4 (27%)described a sample size calculation and a priori de-fined the power to detect effect sizes. Eleven trials(73%) were investigated prospectively. Genotyping pro-cedures were described by 10 trials (67%). None ofthe trials described the mode of inheritance or checkedfor the presence of population stratification. Six trials(10%) checked for HWE and none of the trials founddeviation from HWE. Eight trials (53%) were multi-center trials. DNA sequencing method was used infive trials (33%). In 13 trials (87%), the patients weresuffering from the same type of cancer. Single chemo-therapy regimen was employed in 11 trials (73%).Toxicities were evaluated on the basis of NationalCancer Institute Common Toxicity Criteria in most ofthe included trials (80%).The results of meta-analysis are summarized in Table 3.

Study characteristics and methodologic quality(association between UGT1A1*28 and REG)A total of 72 potentially relevant studies were evaluated

(Fig. 1 shows the numbers of studies evaluated at each



stage). Nine clinical trials including 581 patients wereidentified. Table 2 details the studies' characteristics.Quality-assessment tables are shown in full in Supple-

mentary Table S2. Study sample sizes were small (medi-an, 61; range, 20-134). Two trials (22%) described asample size calculation and a priori defined the powerto detect effect sizes. Four trials (44%) were investigatedprospectively and the other four trials (44%) did not re-port the study design. Genotyping procedures were de-scribed by five trials (56%). Three trials (33%) checkedfor HWE and none of the trials found deviation fromHWE. Four trials (44%) were multicenter trials. DNA se-quencing method was used in five trials (56%). In onetrial (11%), the patients were suffering from the sametype of cancer. Single chemotherapy regimen was em-ployed in five trials (56%). Five trials (56%) determinedthe concentrations of SN-38 glucuronide (SN-38G) direct-ly (without hydrolysis of SN-38G) and the other twotrials (22%) did not report it clearly. In six trials (67%),the plasma samples from each patient seemed enoughfor the accurate determination of the pharmacokineticparameters.The results of themeta-analysis are summarized inTable 3.

Association between UGT1A1*28 and neutropenia(UGT1A1*28/*28 versus UGT1A1*1/*1 orUGT1A1*1/*28)Relevant data for the comparison of the risk of neutro-

penia between UGT1A1*28/*28 patients and those with aUGT1A1*1/*1 or UGT1A1*1/*28 genotype were availablein 15 included trials (3–5, 12–23). Considering the indi-vidual subgroup, there was no evidence of publication biasaccording to either Begg's test or Egger's test (Table 3).

Table 1. Characteristics of trials included in the meta-analysis of the association between UGT1A1*28and neutropenia

Study
Study design No. ofpatients
Age (medianor mean)

h. y 27, 202

Source ofpopulation

Clin

1. © 2010 Ame

Mutationdetection methods

Cancer Res; 16(15) Au

rican Association for C

Race

Iyer 2002 (3)
Prospective 20 Unknown Single center SPR C Innocenti 2004 (4) Prospective 59 60 Single center SPR Mainly C Toffoli 2006 (5) Prospective 250 61 Multicenter PYRS C Cote 2007 (12) Prospective 89 Unknown Multicenter SPR C Kweekel 2008 (13) Retrospective 138 62 Multicenter PYRS C Ferraldeschi 2009 (14) Prospective 92 63 Single center SPR C Marcuello 2004 (15) Prospective 95 68 Unknown SPR C Massacesi 2006 (16) Prospective 56 64 Multicenter Sequencing Unknown Rouits 2004 (17) Retrospective 73 62 Single center PYRS C Carlini 2005 (18) Prospective 62 61 Multicenter SPR Mainly C Glimelius 2010 (19) Retrospective 136 62 Multicenter SPR C Pillot 2006 (20) Retrospective 34 60 Single center PYRS Unknown McLeod 2006* (21) Prospective 103 or 109 Unknown Multicenter Unknown Mainly C Roth 2008† (22) Prospective 628 Unknown Multicenter SPR Unknown Tan 2008† (23) Prospective 54 Unknown Unknown Unknown Unknown
(Continued on the following page)

gust 1, 2010 3835

ancer


Table 1. Characteristics of trials included in the meta-analysis of the association between UGT1A1*28and neutropenia (cont'd)

Type of tumors Chemotherapy regimens Irinotecan dose(mg/m2)/schedule

Toxicity gradecriteria

Solid tumors IRI alone 300 every 3 wk NCISolid tumors IRI alone 350 every 3 wk NCImCRC Modified FOLFIRI or FOLFIRI 180 biweekly NCIStage III mCRC FOLFIRI 180 biweekly NCImCRC CAPIRI 250 every 3 wk NCImCRC FOLFIRI, CAPIRI, or IRI plus VEGF inhibitor 180 biweekly NCImCRC IRI alone, IRI plus Tomudex, or IRI plus FU or LV 80/wk, 180 biweekly or 350 every 3 wk WHOAdvanced CRC IRI plus RAL 80/wk NCImCRC Modified FOLFIRI or FOLFIRI 85/wkor 180 biweekly NCImCRC CAPIRI 100 or 125/wk NCImCRC Modified FOLFIRI or FOLFIRI 180 biweekly NCINSCLC IRI plus CAR 200 every 3 wk NCIAdvanced CRC IROX or FOLFIRI 200 every 3 wk or 100/wk NCIStage III CRC FOLFIRI 180 biweekly NCICCRC Unknown Weekly, biweekly or every 3 wk Unknown

NOTE: Sequencing indicates other DNA sequencing methods. Solid tumors indicate multiple solid tumor types.Abbreviations: SPR, sizing of PCR products (analysis of fragment size); PYRS, pyrosequencing; NSCLC, non–small cell lungcancer; mCRC, metastatic colorectal cancer. C, Caucasian; IRI, irinotecan; CAR, carboplatin; FU, 5-fluorouracil; LV, leucovorin;RAL, raltitrexed; CAPIRI, capecitabine plus irinotecan; FOLFIRI, irinotecan plus FU and leucovorin; IROX, irinotecan plus oxaliplatin;VEGF; vascular endothelial growth factor; NCI, National Cancer Institute common toxicity criteria; NCIC, National Cancer Instituteof Canada Common Toxicity criteria.*Published as abstract. A total 103 patients were administrated with 200 mg/m2 of irinotecan every 3 weeks; 109 patients wereadministrated with 100 mg/m2 of irinotecan weekly.†Published as an abstract.

Hu et al.

3836


When all the dose groups are combined, however, publi-cation bias may exist (Egger's test, P = 0.043; Table 3).This may be caused by the three high-dose studies withhigh RR estimates.Overall analyses suggest an increased risk of neutrope-

nia among UGT1A1*28/*28 patients as compared withthose with a UGT1A1*1/*1 or UGT1A1*1/*28 genotype[RR, 2.20; 95% confidence interval (95% CI), 1.82-2.66;P < 0.001; Fig. 2A]. Heterogeneity was not statisticallysignificant across all studies (I2 = 16.3%, P = 0.267). Thevalue of I2 suggested the existence of weak heterogeneityacross all studies, although without statistical significance.Hence, we did meta-regression to determine whether iri-notecan dose was a significant source of heterogeneityacross studies. Not unexpectedly, meta-regression showedthat irinotecan dose was a significant source of heteroge-neity (P = 0.006; Fig. 3A1).Analyses were further stratified by irinotecan doses. Un-

expectedly, the risk of neutropenia at low doses was signifi-cantly higher amongUGT1A1*28/*28 patients than amongthose with at least one UGT1A1*1 allele (RR, 2.43, 95% CI,1.34-4.39; P = 0.003; Fig. 2A). The risk of neutropenia atmedium doses was also higher amongUGT1A1*28/*28 pa-tients than among those with at least one UGT1A1*1 allele



(RR, 2.00; 95% CI, 1.62-2.47; P < 0.001; Fig. 2A). With re-gard to the high-dose subgroup, the risk of neutropenia wasmuch higher among UGT1A1*28/*28 patients than amongthose with at least one UGT1A1*1 allele (RR, 7.22; 95% CI,3.10-16.78; P < 0.001; Fig. 2A). The RR of neutropenia athigh doses was significantly higher than that at low andmedium doses (RR, 7.22; 95% CI, 3.10-16.78 versus RR,2.04; 95% CI, 1.67-2.49; Table 3).Figure 3A1 shows the relationship between the RRs of

neutropenia and irinotecan doses. There was only a slightincrease in RRs of neutropenia when irinotecan dosesincreased from 80 to 250 mg/m2. However, the RRs ofneutropenia increased dramatically when doses increasedfrom 250 to 350 mg/m2.It should be pointed out that the N9741 trial included

only grade IV toxicity (21). The study by Roth et al. (22)was given themost weight (54%) in ourmeta-analysis. How-ever, exclusion of both studies from the meta-analysis didnot change the general result. For example, the risk of neu-tropenia at low doses was still higher among UGT1A1*28/*28 patients than among those with at least one UGT1A1*1allele (RR, 2.61; 95% CI, 1.39-4.91; P = 0.003). In addition,the RR of neutropenia at low doses was comparable withthat at medium doses (2.61 versus 2.14).






Association between UGT1A1*28 and neutropenia(UGT1A1*1/*28 versus UGT1A1*1/*1)Fourteen trials compared the risk of neutropenia be-

tween patients with a UGT1A1*1/*28 genotype and thosewith a wild-type genotype (3–5, 12–19, 21–23). Onestudy was excluded from analysis because none of the pa-tients suffered from neutropenia (3). Egger's test showedevidence of publication bias (Table 3). When we excludedthe study by Roth et al. (22), the signs of publication biasdisappeared (Egger's test, P = 0.183 for low- and medium-dose group; P = 0.980 for medium-dose group). It is note-worthy that the study by Roth et al. (22) was published asan abstract.The pooled RR was 1.43 (95% CI, 1.16-1.77; P = 0.001)

for all studies (Table 3). No statistical heterogeneity wasdetected (I2 = 0, P = 0.529). Meta-regression showed thatirinotecan dose was not a significant source of heterogene-ity (P = 0.626; Fig. 3B1). The RR of neutropenia did notshow significant difference among three subgroups (Table 3).In a sensitivity analysis excluding two studies (N9741

trial; ref. 21) and the study by Roth et al. (22), the resultswere unchanged. For example, the RR of neutropeniastill did not show significant difference among threesubgroups. Irinotecan doses had no influence on RR ofneutropenia (P = 0.454).



Association between UGT1A1*28 and REG(UGT1A1*1/*1 or UGT1A1*1/*28 versusUGT1A1*28/*28)Nine included trials compared REG of SN-38 between

patients carrying UGT1A1*1 allele(s) and those with aUGT1A1*28/*28 genotype (3–11). There was no evidenceof publication bias according to either Begg's test orEgger's test (Table 3).Overall analyses suggest an increased REG of SN-38

among patients carrying UGT1A1*1 allele(s) as comparedwith those with a UGT1A1*28/*28 genotype (WMD, 2.44;95% CI, 1.73-3.14; P < 0.001; Fig. 2B). No statistical hetero-geneitywas detected (I2 = 0, P = 0.642). Analyseswere furtherstratified by dose. WMD of REG at low and medium doses(1.62; 95% CI, 0.57-2.68; P = 0.002) was lower than thatat high doses (3.08; 95% CI, 2.14-4.02; P < 0.001; Fig. 2B).Heterogeneity between subgroups was significant (P =0.043). Figure 3A2 shows a nonsignificant positive correla-tion betweenWMDof REG and irinotecan doses (P = 0.124).

Association between UGT1A1*28 and REG(UGT1A1*1/*1 versus UGT1A1*1/*28)Nine included trials compared REG of SN-38 between

patients with a UGT1A1*1/*1 genotype and thosewith a UGT1A1*28/*28 genotype (3–11). There was no

Table 2. Characteristics of trials included in the meta-analysis of the association between UGT1A1*28and relative extent of glucuronidation of SN-38

Study
Study design No. ofpatients
Age (medianor mean)

Race
Type oftumors
h. y 27,

Sampling scheme

Clin Ca

2021. © 2010 America

Irinotecan dose(mg/m2)/schedule

ncer Res; 16(15) Au

n Association for

Analyticalmethod

Iyer 2002 (3)
Prospective 20 Unknown C Solidtumors
15 time points,up to 24 h

300 every 3 wk
HPLC-Flu
Innocenti2004 (4)

Prospective
61 60 Mainly C Solidtumors
14 time points,up to 25.5 h

350 every 3 wk
HPLC-Flu
Toffoli2006 (5)

Prospective
71 61 C mCRC 16 time points,up to 50 h
180 biweekly
HPLC-Flu
de Jong2007 (6)

Retrospective
134 55 C Solidtumors
Multiple time points,up to 20 time points

350 every 3 wk
HPLC-Flu
Mathijssen2003 (7)

Unknown
53 53 Mainly C Solidtumors

200 to 350 every3 wk

HPLC-Flu

Mathijssen2004 (8)

Prospective
30 55 C Solidtumors

350 every 3 wk*
HPLC-Flu
Paoluzzi2004 (9)

Unknown
86 54 C Solidtumors
Unknown
350 every 3 wk* HPLC-Flu
Minami2007 (10)

Unknown
85 61 A Solidtumors

60-150/wk orbiweekly

HPLC-Flu

Sai 2004 (11)
Unknown 41 62 A Solidtumors

60-150/wk orbiweekly

HPLC-Flu

NOTE: Solid tumors indicate multiple solid tumor types.Abbreviations: A, Asian; HPLC-Flu, high-performance liquid chromatography with fluorescence detection.*Patients were treated with irinotecan once every 3 weeks at a fixed dose of 600 mg. A fixed dose of 600 mg was similar to a doseof 350 mg/m2.

gust 1, 2010 3837

Cancer


Hu et al.

3838


evidence of publication bias according to either Begg's testor Egger's test (Table 3).Analyses were stratified by dose. WMDof REG showedno

significant difference between the low/medium-dose group(1.85; 95% CI, 1.00-2.70; P < 0.001) and the high-dosegroup (1.03; 95% CI, −0.09 to 2.16; P = 0.072; Table 3).Heterogeneity between subgroups was not significant (P =



0.225). There was no correlation at all between WMD ofREG and irinotecan doses (Fig. 3B2).

Discussion

The primary finding of this study is that the UGT1A1*28/*28 genotype was associated with an increased risk of

Table 3. Summary of meta-analysis

Comparisonor outcome

No. oftrials

No. ofparticipants

Irinotecandoses

Statisticalmethod

Effect size(95%confidenceintervals)

h. y 27, 2021

P

. © 2010

Test forheterogeneity

Cl

American Ass

inical

ociat

Begg'stest

Cancer

ion for C

Egger'stest

P
I2
(%)
P P
Neutropenia(*28/*28 vs.*1/*1 or*1/*28)

4
300 Low RR (fixed) 2.43(1.34-4.39)
0.003*
0.964 0 Ne Ne
9
1,481 Medium RR (fixed) 2.00(1.62-2.47)
<0.001*
0.325 13.1 0.348 0.234
13
1,781 Low andmedium
RR (fixed)
2.04(1.67-2.49)
<0.001*
0.612 0 0.300 0.279
3
217 High RR (fixed) 7.22(3.10-16.78)
<0.001*
0.713 0 Ne Ne
15
1998 Total RR (fixed) 2.20(1.82-2.66)
<0.001*
0.267 16.3 0.893 0.043†
REG (*1/*1 or*1/*28 vs.*28/*28)

3
197 Low andmedium
WMD (fixed)
1.62(0.57-2.68)
0.002‡
0.854 0 Ne Ne
6
384 High WMD (fixed) 3.08(2.14-4.02)
<0.001‡
0.897 0 Ne Ne
9
581 Total WMD (fixed) 2.44(1.73-3.14)
<0.001‡
0.642 0 0.175 0.060
Neutropenia(*1/*28 vs.*1/*1)

4
270 Low RR (fixed) 2.94(1.36-6.35)
0.006§
0.747 0 Ne Ne
8
1288 Medium RR (fixed) 1.29(1.04-1.62)
0.023§
0.581 0 0.902 0.001†
12
1558 Low andmedium
RR (fixed)
1.40(1.14-1.74)
0.002§
0.462 0 0.451 0.004†
2
180 High RR (fixed) 2.65(0.70-9.94)
0.149
0.547 0 Ne Ne
14
1738 Total RR (fixed) 1.43(1.16-1.77)
0.001§
0.529 0 0.274 0.001†
REG (*1/*1 vs.*1/*28)

3
182 Low andmedium
WMD (fixed)
1.85(1.00-2.70)
<0.001∥
0.148 47.6 Ne Ne
6
356 High WMD (fixed) 1.03(−0.09 to 2.16)
0.072
0.591 0 Ne Ne
9
538 Total WMD (fixed) 1.55(0.87-2.23)
<0.001∥
0.357 9.4 0.917 0.999
NOTE: Ne, Begg's and Egger's tests were not done if <8 studies were included in the analyzed subgroup.*The risk of toxicity was significantly higher among patients with a UGT1A1*28/*28 genotype than among those with a UGT1A1*1/*1or UGT1A1*1/*28 genotype.†Publication bias may exist.‡The REG was significantly higher among patients with a UGT1A1*1/*1 or UGT1A1*1/*28 genotype than among those with aUGT1A1*28/*28 genotype.§The riskof toxicitywassignificantly higher amongpatientswithaUGT1A1*1/*28genotype thanamong thosewith aUGT1A1*1/*1genotype.∥REG was significantly higher among patients with a UGT1A1*1/*1 genotype than among those with a UGT1A1*1/*28 genotype.

Research

ancer




neutropenia not only at medium or high doses of irinote-can but at low doses as well. The RRs of neutropenia amongUGT1A1*28/*28 patients were comparable at low andmedium doses. The secondary finding is that there was no



correlation between the RRs of neutropenia among patientsheterozygous for UGT1A1*28 (as compared with wild-typegenotype) and irinotecan doses. The last finding is thatWMD of REG (UGT1A1*1/*1 or UGT1A1*1/*28 versus

Fig. 2. A, summary RR ofirinotecan-induced neutropeniafor UGT1A1*28/*28 versusUGT1A1*1/*1 or UGT1A1*1/*28;B, WMD of REG for UGT1A1*1/*1or UGT1A1*1/*28 versusUGT1A1*28/*28. A fixed-effectsmodel was used for all analyses.Squares, study-specific estimates(size of the square reflects thestudy-specific statistical weight);horizontal lines, 95% CI;diamonds, summary estimateswith corresponding 95% CI.




Hu et al.

3840


UGT1A1*28/*28) increased with increasing dose of irinote-can. In contrast, when comparing REGbetween patientswitha UGT1A1*1/*1 genotype and those with a UGT1A1*1/*28genotype,WMDof REG did not change with increasing doseof irinotecan. This finding sheds light on the mechanism ofthe dose-dependent association between UGT1A1*28 andneutropenia.On the basis of the findings of four initial studies (4, 15,

17, 30), the FDA made the recommendation that patientswith a UGT1A1*28/*28 genotype should receive a lowerstarting dose of irinotecan. Then came conflicting resultsfrom a meta-analysis (821 patients) by Hoskins et al.(1), suggesting the association between UGT1A1*28 andneutropenia was dose dependent; UGT1A1*28 had no ef-fect in patients treated with a low dose of irinotecan (<150mg/m2). Our current meta-analysis based on a large sam-ple size (1,998 patients) indicates that UGT1A1*28/*28patients are at an increased risk of neutropenia not onlyif they are being treated with medium (RR, 2.00) and highdoses (RR, 7.22) of irinotecan but also if they are beingtreated with low doses (RR, 2.43; 80-145 mg/m2).



Diarrhea is another important side effect related to irino-tecan administration. Recently, we found thatUGT1A1*28/*28 patients were at an increased risk of diarrhea atmedium(RR, 1.79; 95%CI, 1.08-2.97) or high doses (RR, 2.32; 95%CI, 1.25-4.28) of irinotecan, but not at low doses (RR, 0.65;95% CI, 0.27-1.58; ref. 31). As a result, when selecting thedose of irinotecan for UGT1A1*28/*28 patients, this infor-mation should be considered together with the results ofthe current meta-analysis.The implications for clinical practice should be consid-

ered. When regimens with a high dose of irinotecan arebeing considered, dose reduction is advisable forUGT1A1*28/*28 patients (neutropenia RR, 7.22; diarrheaRR, 2.32). When regimens with a medium dose of irinote-can are being considered, dose reduction is also recom-mended for UGT1A1*28/*28 patients (neutropenia RR,2.00; diarrhea RR, 1.79). However, dose reduction is op-tional for UGT1A1*28/*28 patients (neutropenia RR,2.43) treated at low irinotecan doses. In this regard, wesuggest that UGT1A1*28/*28 patients with other predic-tors of irinotecan-induced toxicity be given a reduced dose

Fig. 3. A1, RR of irinotecan-induced neutropenia (UGT1A1*28/*28 versus UGT1A1*1/*1 or UGT1A1*1/*28) against dose of irinotecan; A2, weighted meandifference (WMD) of REG (UGT1A1*1/*1 or UGT1A1*1/*28 versus UGT1A1*28/*28) against dose of irinotecan. B1, RR of irinotecan-induced neutropenia(UGT1A1*1/*28 versus UGT1A1*1/*1) against dose of irinotecan; B2, WMD of REG (UGT1A1*1/*1 versus UGT1A1*1/*28) against dose of irinotecan. Size ofcircle is proportional to the study-specific statistical weight.Marked “Excluded” indicates the excluded study by Iyer et al. (ref. 3; zero incidence of neutropenia).






of irinotecan. Other predictors of irinotecan-inducedtoxicity could be nongenetic factors (neutrophil baselinelevels or sex) or genetic factors (UGT1A1*93, ABCC1IVS11 −48C>T, or SLCO1B1*1b; ref. 32).Recently, Toffoli et al. (33) evidenced that the recom-

mended dose of 180 mg/m2 for irinotecan in FOLFIRI isconsiderably lower than the dose that can be tolerated bythe non-UGT1A1*28/*28 patients. This result was sup-ported by our data that the incidence of neutropenia (%)was not increased with increasing dose of irinotecan innon-UGT1A1*28/*28 patients (Supplementary Fig. S1).The limitations of this meta-analysis need to be consid-

ered. Firstly, the possibility of information and selectionbiases cannot be completely excluded because some ofthe included studies were retrospective. Secondly, althoughno statistical heterogeneity was observed among the ana-lyzed studies, there were many sources of heterogeneityamong the analyzed studies, such as study design, sourceof population, dose administered, chemotherapy regimens,mutation detection methods, toxicity grade criteria, sam-pling scheme, type of tumor, and stage of disease. Thirdly,three included trials were published as abstracts. Fourthly,in our analysis, grade III-IV neutropenia data were availablefor most studies, whereas only grade IV neutropenia infor-mation could be extracted from two studies (20, 21). Fifth-ly, for the study of the association between UGT1A1*28and REG of SN-38, two included trials were conducted onAsian patients. Ethnic difference was not considered here.To assess the potential for publication bias to have in-

fluenced the results of our meta-analysis, we calculated thefailsafe numbers using a weighted method (35). A failsafenumber can be defined as the number of nonsignificant,unpublished studies that would be needed to reduce astatistically significant observed result to nonsignificance(34, 35). The failsafe number was 14 for the low-dose



group, 148 for the medium-dose group, and 21 for thehigh-dose group (at α = 0.05). It seems unlikely that suchnumbers of studies with null findings exist and have notbeen published.In the present meta-analysis, we excluded two trials

conducted on children according to the exclusion criterion(36, 37). The children in these two trials were administrat-ed with very low doses of irinotecan (30 and 50 mg/m2).Pooled RR from these two trials (118 patients) showed thatthe risk of neutropenia was similar between UGT1A1*28/*28 patients and those carrying UGT1A1*1 allele(s) (RR,0.63; 95% CI, 0.14-2.72; P = 0.532). However, it is stilltoo early to draw a reliable conclusion based on two trialsincluding only 118 patients. As a result, further studies arerequired in this area.In conclusion, the UGT1A1*28/*28 genotype was asso-

ciated with an increased risk of neutropenia not only atmedium or high doses of irinotecan but also at low doses.The dose-dependent manner of SN-38 glucuronidationexplained why the association between UGT1A1*28 andneutropenia was dose dependent.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We gratefully acknowledge Dai-Li Xi for valuable comments andsuggestions regarding this article.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely toindicate this fact.

Received 04/29/2010; revised 06/07/2010; accepted 06/08/2010.

References
1. Hoskins JM, Goldberg RM, Qu P, Ibrahim JG, McLeod HL.
UGT1A1*28 genotype and irinotecan-induced neutropenia: dosematters. J Natl Cancer Inst 2007;99:1290–5.

2. Iyer L, King CD, Whitington PF, et al. Genetic predisposition to themetabolism of irinotecan (CPT-11). Role of uridine diphosphateglucuronosyltransferase isoform 1A1 in the glucuronidation of itsactive metabolite (SN-38) in human liver microsomes. J Clin Invest1998;101:847–54.

3. Iyer L, Das S, Janisch L, et al. UGT1A1*28 polymorphism as a deter-minant of irinotecan disposition and toxicity. Pharmacogenomics J2002;2:43–7.

4. Innocenti F, Undevia SD, Iyer L, et al. Genetic variants in theUDP-glucuronosyltransferase 1A1 gene predict the risk of severeneutropenia of irinotecan. J Clin Oncol 2004;22:1382–8.

5. Toffoli G, Cecchin E, Corona G, et al. The role of UGT1A1*28polymorphism in the pharmacodynamics and pharmacokinetics ofirinotecan in patients with metastatic colorectal cancer. J Clin Oncol2006;24:3061–8.

6. de Jong FA, Scott-Horton TJ, Kroetz DL, et al. Irinotecan-induceddiarrhea: functional significance of the polymorphic ABCC2transporter protein. Clin Pharmacol Ther 2007;81:42–9.

7. Mathijssen RH, Marsh S, Karlsson MO, et al. Irinotecan pathway ge-notype analysis to predict pharmacokinetics. Clin Cancer Res 2003;9:3246–53.

8. Mathijssen RH, de Jong FA, van Schaik RH, et al. Prediction ofirinotecan pharmacokinetics by use of cytochrome P450 3A4phenotyping probes. J Natl Cancer Inst 2004;96:1585–92.

9. Paoluzzi L, Singh AS, Price DK, et al. Influence of genetic variants inUGT1A1 and UGT1A9 on the in vivo glucuronidation of SN-38. J ClinPharmacol 2004;44:854–60.

10. Minami H, Sai K, Saeki M, et al. Irinotecan pharmacokinetics/phar-macodynamics and UGT1A genetic polymorphisms in Japanese:roles of UGT1A1*6 and *28. Pharmacogenet Genomics 2007;17:497–504.

11. Sai K, Saeki M, Saito Y, et al. UGT1A1 haplotypes associated withreduced glucuronidation and increased serum bilirubin in irinotecan-administered Japanese patients with cancer. Clin Pharmacol Ther2004;75:501–15.

12. Côté JF, Kirzin S, Kramar A, et al. UGT1A1 polymorphism can predicthematologic toxicity in patients treated with irinotecan. Clin CancerRes 2007;13:3269–75.

13. Kweekel DM, Gelderblom H, Van der Straaten T, Antonini NF,Punt CJ, Guchelaar HJ. UGT1A1*28 genotype and irinotecan dosagein patients with metastatic colorectal cancer: a Dutch ColorectalCancer Group study. Br J Cancer 2008;99:275–82.

14. Ferraldeschi R, Minchell LJ, Roberts SA, et al. UGT1A1*28 genotypepredicts gastrointestinal toxicity in patients treated with intermedi-ate-dose irinotecan. Pharmacogenomics 2009;10:733–9.




Hu et al.

3842


15. Marcuello E, Altés A, Menoyo A, Del Rio E, Gómez-Pardo M, BaigetM. UGT1A1 gene variations and irinotecan treatment in patients withmetastatic colorectal cancer. Br J Cancer 2004;91:678–82.

16. Massacesi C, Terrazzino S, Marcucci F, et al. Uridine diphosphateglucuronosyl transferase 1A1 promoter polymorphism predicts therisk of gastrointestinal toxicity and fatigue induced by irinotecan-based chemotherapy. Cancer 2006;106:1007–16.

17. Rouits E, Boisdron-Celle M, Dumont A, Guérin O, Morel A, Gamelin E.Relevance of different UGT1A1 polymorphisms in irinotecan-inducedtoxicity: a molecular and clinical study of 75 patients. Clin Cancer Res2004;10:5151–9.

18. Carlini LE, Meropol NJ, Bever J, et al. UGT1A7 and UGT1A9polymorphisms predict response and toxicity in colorectal cancerpatients treated with capecitabine/irinotecan. Clin Cancer Res2005;11:1226–36.

19. Glimelius B, Garmo H, Berglund A, et al. Prediction of irinotecan and5-fluorouracil toxicity and response in patients with advanced colo-rectal cancer. Pharmacogenomics J 2010. Epub ahead of print.

20. Pillot GA, Read WL, Hennenfent KL, et al. A phase II study of irino-tecan and carboplatin in advanced non-small cell lung cancer withpharmacogenomic analysis: final report. J Thorac Oncol 2006;1:972–8.

21. McLeod HL, Parodi L, Sargent DJ, et al. UGT1A1*28, toxicity andoutcome in advanced colorectal cancer: Results from Trial N9741.Journal of Clinical Oncology ASCO Annual Meeting Proceedings2006;24:3520.

22. Roth AD, Yan P, Dietrich D, et al. Does UGT1A1*28 homozygositypredict for severe toxicity in patients treated with 5-fluorouracil(5-FU)-irinotecan (IRI)? Results of the PETACC 3-EORTC 40993-SAKK 60/00 trial comparing IRI/5-FU/folinic acid (FA) to 5-FU/FA instage II-III colon cancer. Gastrointestinal Cancers Symposium 2008;2008:277.

23. Tan BR, Zehnbauer B, Picus J, et al. UGT1A1 genotype-based dosemodification of irinotecan regimens: Impact on hematologic toxicities.Journal of Clinical Oncology ASCOAnnualMeeting 2008;2008:13500.

24. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variancefrom the median, range, and the size of a sample. BMC Med ResMethodol 2005;5:13.



25. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observation-al studies in epidemiology: a proposal for reporting. Meta-analysis OfObservational Studies in Epidemiology (MOOSE) group. JAMA 2000;283:2008–12.

26. Jorgensen AL, Williamson PR. Methodological quality of pharmaco-genetic studies: issues of concern. Stat Med 2008;27:6547–69.

27. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58.

28. Begg CB, Mazumdar M. Operating characteristics of a rank correla-tion test for publication bias. Biometrics 1994;50:1088–101.

29. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-anal-ysis detected by a simple, graphical test. BMJ 1997;315:629–34.

30. Ando Y, Saka H, Ando M, et al. Polymorphisms of UDP-glucurono-syltransferase gene and irinotecan toxicity: a pharmacogeneticanalysis. Cancer Res 2000;60:6921–6.

31. Hu ZY, Yu Q, Zhao YS. Dose-dependent association betweenUGT1A1 *28 polymorphism and irinotecan-induced diarrhoea: ameta-analysis. Eur J Cancer 2010;46:1856–65.

32. Innocenti F, Kroetz DL, Schuetz E, et al. Comprehensive pharmaco-genetic analysis of irinotecan neutropenia and pharmacokinetics.J Clin Oncol 2009;27:2604–14.

33. Toffoli G, Cecchin E, Gasparini G, et al. Genotype-driven phase Istudy of irinotecan administered in combination with fluorouracil/leu-covorin in patients with metastatic colorectal cancer. J Clin Oncol2010;28:866–71.

34. Rosenthal R. The file drawer problem and tolerance for null results.Psychol Bull 1979;86:638–41.

35. Rosenberg MS. The file-drawer problem revisited: a general weight-ed method for calculating fail-safe numbers in meta-analysis. Evolu-tion 2005;59:464–8.

36. Stewart CF, Panetta JC, O'Shaughnessy MA, et al. UGT1A1 promot-er genotype correlates with SN-38 pharmacokinetics, but not severetoxicity in patients receiving low-dose irinotecan. J Clin Oncol 2007;25:2594–600.

37. Bomgaars L, Kuttesch N, Bernstein M, Blaney S. Correlation ofUGT1A1 promoter genotype with pharmacokinetics and toxicity inpediatric patients receiving irinotecan (CPT-11). Journal of ClinicalOncology ASCO Annual Meeting 2003;2003:551.




2010;16:3832-3842. Published OnlineFirst June 18, 2010.Clin Cancer Res Zhe-Yi Hu, Qi Yu, Qi Pei, et al. Riskand Irinotecan-Induced Neutropenia: Low Doses Also Increase Dose-Dependent Association between UGT1A1*28 Genotype

Updated version

10.1158/1078-0432.CCR-10-1122doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2010/08/10/1078-0432.CCR-10-1122.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/16/15/3832.full#ref-list-1

This article cites 36 articles, 11 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/16/15/3832.full#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/16/15/3832To request permission to re-use all or part of this article, use this link

Research. on January 27, 2021. © 2010 American Association for Cancerclincancerres.aacrjournals.org Downloaded from


http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-10-1122

http://clincancerres.aacrjournals.org/content/suppl/2010/08/10/1078-0432.CCR-10-1122.DC1

http://clincancerres.aacrjournals.org/content/16/15/3832.full#ref-list-1

http://clincancerres.aacrjournals.org/content/16/15/3832.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/16/15/3832


Documents

Research Dose-Dependent Association between UGT1A1*28 ... · Begg and Mazumdar (28) and Egger (29). The studies included were stratified into different dose groups. Irinotecan dose