For reprint orders, please contact: reprints@futuremedicine.com 733 ReseaRch aRticle ISSN 1462-2416 10.2217/PGS.09.20 © 2009 Future Medicine Ltd Pharmacogenomics (2009) 10 (5), 733–739 UGT1A1*28 genotype predicts gastrointestinal toxicity in patients treated with intermediate-dose irinotecan Irinotecan (7-ethyl-10-[4-(1-piperidino)-1- piperidino]carbonyloxycamptothecin) is a chemo- therapeutic agent efficacious in the treatment of colorectal cancer (CRC) and lung cancer. In the UK, it is approved as a first- and second-line therapy in CRC [1] . Irinotecan can cause severe, unpredictable toxicity in the form of myelo- suppression and delayed-type diarrhea in 20–35% of patients [2] and so is less frequently used as a first-line therapy due to the availability of alterna- tive agents. However, overall survival is increased in patients receiving both irinotecan and oxalipla- tin-based chemotherapy during the course of their treatment [3] , and therefore irinotecan remains important in the treatment of CRC. Irinotecan is a prodrug, which is converted to the cytotoxic metabolite 7-ethyl-10-hydroxycamp- tothecin (SN-38), a potent topoisomerase 1 inhib- itor [4] . The main pathway of SN-38 elimination is glucuronidation to 10-O-glucuronyl-SN-38 (SN-38G) by the liver enzyme uridine diphos- phate glucuronosyltransferase 1A1 (UGT1A1) [5] . Variants in the UGT1A1 gene have been demon- strated to decrease its enzymatic activity, reduc- ing removal and increasing exposure to SN-38 [5] . The UGT1A1*28 allele consists of a two base pair (TA) insertion in the promoter region ([TA] 6 TAA>[TA] 7 TAA). This variant causes decreased enzyme expression and homozygosity is associated with Gilbert’s syndrome, a com- mon cause of mild, benign hyperbilirubinemia [6] . Importantly, homozygosity for UGT1A1*28 has been associated with increased risk of irino- tecan-induced hematological [7–9] and gastroin- testinal toxicities [7,10], prompting the US FDA to recommend a drug-labeling change to highlight the UGT1A1*28/*28 genotype as a risk factor for hematologic toxicity and approve a diagnos- tic test for the UGT1A1*28 allele. Subsequent pharmacogenetic studies to replicate the asso- ciation between UGT1A1 variation and irino- tecan toxicity have produced inconsistent results [11–15] and other UGT1A1 variants, including a promoter variant (c.-3156G>A), have been pro- posed as better predictors of irinotecan toxicity than UGT1A1*28 [8,12] . UGT1A1 testing has not been adopted into standard clinical practice in the UK to predict individuals at risk of irinote- can-induced toxicity. Therefore, we undertook a prospective cohort study to determine the role of testing for the UGT1A1*28 and c.-3156G>A variants in predicting toxicity in CRC patients in a major UK oncology centre. Patients & methods  Study participants A prospective cohort of 96 patients with meta- static CRC were recruited from Christie Hospital Aims: Variants in UGT1A1 have previously been associated with toxicity from irinotecan chemotherapy. We conducted a pragmatic prospective cohort study to establish the relevance of UGT1A1 variants in the prediction of severe diarrhea and neutropenia in patients with colorectal cancer receiving irinotecan in a routine clinical setting. Materials & methods: Genotyping of UGT1A1*28 and c.-3156G>A was undertaken in an unselected, prospective cohort of 96 individuals treated with irinotecan at a single major UK oncology centre. Data on cytotoxic drugs received, and toxicity for all irinotecan treatment cycles were collected from case notes. Over 95% (92/96) of patients received an intermediate dose of irinotecan (180 mg/m 2 , twice weekly). Irinotecan was given in combination with other cytotoxic drugs in 93/96 subjects and Grade 3 or 4 toxicity occurred in 23% of subjects. Results: No association was found between UGT1A1*28 or c.- 3156G>A and neutropenia. However, individuals carrying two copies of UGT1A1*28 (p = 0.04; OR: 14; 95% CI: 1.1–185) or c.-3156G>A (p = 0.03) had a significantly increased risk of diarrhea over all cycles. Conclusion: Our findings indicate that UGT1A1 genotyping is not a good predictor of hematological toxicity in patients treated with intermediate irinotecan doses. However, it may be useful in the identification of patients at risk of severe diarrhea. KEYWORDS: diarrhea  irinotecan  neutropenia  pharmacogenetics  toxicity  UGT1A1 Roberta Ferraldeschi 1,2,3 *, Laura J Minchell 1,3 *, Stephen A Roberts 3 , Simon Tobi 1,3 , Kristen D Hadfeld 1,3 , Fiona H Blackhall 2 , Saifee Mullamitha 2 , Gregory Wilson 2 , Juan Valle 2 , Mark Saunders 2 & William G Newman 1,3† † Author for correspondence: 1 Department of Medical Genetcs, St Mary’s Hospital, Manchester, M13 0JH, UK Tel.: +44 161 276 6264; Fax: +44 161 276 6145; william.newman@ manchester.ac.uk 2 Christe Hospital NHS Trust, Manchester, UK 3 Manchester Academic Health Sciences Centre (MAHSC), University of Manchester, Manchester, UK *Authors contributed equally to this work