Relevance of CYP3A4*20, UGT1A1*37 and UGT1A1*28 variants in irinotecan‐induced severe toxicity

Abstract

Severe irinotecan‐induced toxicity is associated with UGT1A1 polymorphisms. However, some patients develop side‐effects despite harbouring a normal UGT1A1 genotype. As CYP3A4 is also an irinotecan‐metabolizing enzyme, our study aimed to elucidate the influence of the CYP3A4*20 loss‐of‐function allele in the toxicity profile of these patients. Three‐hundred and eight metastatic colorectal cancer patients treated with an irinotecan‐containing chemotherapy were studied. The presence of CYP3A4*20, UGT1A1*37 and UGT1A1*28 alleles was tested. Associations between these genetic variants and toxicity were evaluated. UGT1A1*28 was significantly associated with severe diarrhoea, neutropenia and asthenia (P = 0.002, P = 0.037 and P = 0.041, respectively). One patient with the UGT1A1*28/*37 genotype presented with grade IV neutropenia and lethal septic shock. One heterozygous UGT1A1 (*1/*28) patient also carried the CYP3A4*20 allele but did not develop toxicity. We confirm that UGT1A1*37 and UGT1A1*28 are associated with severe toxicity and suggest that the CYP3A4*20 allele does not play a role in irinotecan‐induced toxicity.

Introduction

http://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=6823 [also known as camptothecin‐11 (CPT‐11)] is an active drug that is widely used for the treatment of several types of cancer, including colorectal, gastric and pancreatic cancer. Its dose‐limiting and potentially life‐threatening toxicities are delayed diarrhoea and severe neutropenia, which can occur in up to 36% of patients 1.

Several enzymes are involved in the metabolism of irinotecan, highlighting the http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2990, encoded by the UGT1A1 gene. The number of TA repeats in its TATA box is associated with the enzyme activity. Whereas six TA repeats imply normal activity (UGT1A1*1), seven TA repeats (UGT1A1*28) or eight TA repeats (UGT1A1*37) reduce activity markedly. The UGT1A1*28 allele is strongly associated with irinotecan‐induced toxicity 2, 3, 4. The UGT1A1*37 allele is rare in the Caucasian population 5, and its influence on irinotecan toxicity has not yet been reported. However, some patients carrying a normal UGT1A1 genotype (*1/*1) develop severe diarrhoea and/or neutropenia. Identification of additional toxicity biomarkers that could help to individualize irinotecan treatment is therefore of major clinical interest in the management of cancer patients.

The enzyme cytochrome P450 3A4 (http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1337) converts irinotecan into several inactive oxidation products, such as 7‐ethyl‐10‐[4‐N‐(5‐aminopentanoic acid)‐1‐piperidino]‐carbonyloxycamptothecin (APC) and 7‐ethyl‐10‐[4‐amino‐1‐piperidino]‐carbonyloxycamptothecin (NPC). Although the CYP3A4 phenotype is associated with irinotecan pharmacokinetics 6, 7, no significant associations between CYP3A4 genotype and irinotecan toxicity have been found 8. Three CYP3A4 nonfunctional alleles result from premature stop codons in the coding gene *6, *20 and *26 CYP3A4 alleles. In Caucasians, these variants are extremely rare, but a frequency of 1.2% of carriers of CYP3A4*20 has recently been reported in the Spanish population as the result of a founder effect 9.

This state of knowledge prompted us to evaluate whether the presence of the CYP3A4*20 allele could help to identify patients who develop severe irinotecan‐induced toxicity despite having a normal UGT1A1 genotype. We genotyped the UGT1A1 (TA repeats) and the CYP3A4*20 variant in a cohort of 308 metastatic colorectal cancer (mCRC) patients treated with an irinotecan‐containing therapy.

Methods

Patient population

The study included 308 mCRC patients treated with an irinotecan‐containing therapy at Hospital de la Santa Creu i Sant Pau (HSCSP, Barcelona, Spain). Genomic DNA was available in all cases. Left colon cancer was diagnosed in 144 patients, right colon cancer in 76 and rectum cancer in 88. A total of 283 patients received irinotecan combined with fluorouracil and leucovorin (FOLFIRI), 176 of whom were enrolled in clinical trials 10, 11. The remaining patients received either irinotecan monotherapy or another chemotherapeutic regimen containing irinotecan. Patients were administered chemotherapy until unacceptable toxicity or disease progression. Table 1 shows patients' baseline clinical characteristics.

Table 1.

Baseline patient characteristics (n = 308)

Characteristic	n	%
Gender
Male	191	62
Female	117	38
Age (years)
<75	265	86
≥75	43	14
Primary tumour site
Right colon	76	25
Left colon	144	47
Rectum	88	29
Chemotherapy treatment
FOLFIRI	283	92
Irinotecan monotherapy	13	4
Other irinotecan combinations	12	4
Line of chemotherapy
First line	244	79
≥Second line	64	21

FOLFIRI, combination of irinotecan, fluorouracil and leucovorin

Relevant clinical data (gender, age and line of chemotherapy) were collected retrospectively through chart review. Toxicity was graded according to the common toxicity criteria (CTC) v4.0: presence and grade of diarrhoea, neutropenia, febrile neutropenia, asthenia, nausea and mucositis. The most severe grade of toxicity was recorded in all cases. The study was approved by the Institutional Ethics Committee at HSCSP, and all study participants gave informed consent to participate.

Genotyping of UGT1A1 and CYP3A4

UGT1A1 genotyping was performed as previously described 10. CYP3A4*20 (rs67666821) genotyping was carried out by automatic sequencing of a fragment of 244 base pairs in exon 13, which was specifically amplified using the primers FW (5′‐TGAAGGAGTGTCTCACTC‐3′) and RV (5′‐AGGTCTCTGGTGTTCTCAG‐3′).

Statistical analyses

Allele frequencies were calculated by direct counting. The associations between the toxicities assessed and the afore‐mentioned genetic variants were analysed using chi‐square tests. The results were considered statistically significant if P <0.05.

Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY 12, and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18 13.

Results

A total of 137 of the 308 patients developed some type of grade III/IV toxicity. The most common severe adverse effects observed were neutropenia (68 patients, 22%), asthenia (65 patients, 21%) and diarrhoea (55 patients, 18%). Twenty‐seven patients experienced severe nausea, nine patients presented with severe mucositis, eight patients experienced febrile neutropenia and two developed septic shock.

The frequencies of the genetic variants studied were consistent with Hardy–Weinberg equilibrium. For the UGT1A1 variants, we found 44.2% (*1/*1), 46.4% (*1/*28), 9.1% (*28/*28) and 0.3% (*28/*37). The allelic frequency of UGT1A1*28 was 32%, within the range reported in Spanish populations 14. The genotype frequencies for the CYP3A4*20 variant were 99.7% (*1/*1) and 0.3% (*1/*20).

There was a clear association between the UGT1A1*28/*28 genotype and the occurrence of severe diarrhoea (P = 0.002), neutropenia (P = 0.037) and asthenia (P = 0.041) (Table 2). The presence of nausea or mucositis was not significantly associated with the UGT1A1*28/*28 genotype (P = 0.495 and P = 0.595, respectively).

Table 2.

Associations between UGT1A1 genotypes and grade III–IV diarrhoea, neutropenia and asthenia

Genotype	n (%)	Diarrhoea (%)	Neutropenia (%)	Asthenia (%)
*1/*1	136 (44.2%)	18 (13.2%)	23 (16.9%)	23 (16.9%)
*1/*28	143 (46.4%)	25 (17.5%)	34 (23.8%)	31 (21.7%)
*28/*28	28 (9.1%)	12 (42.9%)	10 (35.7%)	10 (35.7%)
*28/*37	1 (0.32%)	0 (0%)	1 (100%)	1 (100%)
P		0.002a	0.037a	0.041a

UGT1A1, UDP‐glucuronosyltransferase isoform 1A1

^aTo perform the chi‐square test, patients *28/*28 and *28/*37 were considered together

The extremely rare UGT1A1 genotype *28/*37 was identified in one mCRC patient who received a reduced dose of irinotecan monotherapy (150 mg m⁻²) as a second‐line treatment. The patient developed grade IV neutropenia and febrile neutropenia; neutropenic colitis was diagnosed and led to lethal septic shock.

Thirty‐eight out of 136 patients (28%) with a UGT1A1*1/*1 genotype presented with severe diarrhoea and/or neutropenia, despite the normal enzymatic activity associated with this genotype. None of these patients harboured the nonfunctional CYP3A4*20 allele. One patient out of 308 carried the CYP3A4*20 variant. This patient, with the UGT1A1*1/*28 genotype, received the FOLFIRI regimen as first‐line treatment and cetuximab–irinotecan as second‐line treatment without any of the toxicities assessed.

Discussion

We investigated the role of the CYP3A4*20 allele in mCRC patients with a normal UGT1A1 genotype treated with an irinotecan‐containing regimen. We found that UGT1A1*37 and UGT1A1*28 alleles were highly associated with severe toxicity to irinotecan treatment, whereas the presence of the CYP3A4*20 allele was not associated with toxicity in our sample.

Patients harbouring the UGT1A1*28/*28 genotype (Gilbert's syndrome) are at risk of severe irinotecan‐induced toxicity 2, 4. Our findings confirmed this association, as the presence of this genotype was statistically significant for diarrhoea, neutropenia and asthenia (P = 0.002, P = 0.037 and P = 0.041, respectively). One patient who carried the infrequent UGT1A1*28/*37 genotype developed fatal toxicity, despite the reduced doses of irinotecan administered. One explanation for this could be that the additive effect of the UGT1A1*37 allele reduces enzyme activity even more than the UGT1A1*28 allele 15. This case study reinforces the clinical relevance of identifying such a low‐frequency nonfunctional variant before prescribing irinotecan. However, further studies are needed in order to confirm the role of this genetic variant in severe irinotecan‐induced toxicity.

Several phenotypic studies have focused on the role of the enzymatic activity of CYP3A4 in irinotecan‐induced toxicity. In 2008, Rouits et al. 16 estimated CYP3A4 activity from cortisol biotransformation into 6β‐hydroxycortisol. They found that the level of cortisol 6β‐hydroxylation predicted the occurrence of diarrhoea. Later, in 2010, van der Bol et al. 17 conducted a randomized trial in which an irinotecan dose of 350 mg m⁻² was administered in the control group, whereas the dose was calculated using an algorithm that considered midazolam clearance in the experimental group. They found a significant reduction in the incidence of grades III–IV neutropenia in the experimental group but no differences in grades III–IV diarrhoea 17. Recently, Makihara et al. 18 conducted a study to evaluate the influence of concomitant use of clarithromycin, a potent CYP3A4 inhibitor, on irinotecan toxicity. The study did not identify an increase in irinotecan‐induced toxicity by coadministration of clarithromycin 18. To the best of our knowledge, to date, only one genotypic study, performed in 177 Japanese cancer patients, has evaluated the influence of variants in the CYP3A4 gene on irinotecan‐induced toxicity; no significant association was found 8.

Recently, a frequency of 1.2% of the loss‐of‐function CYP3A4*20 allele was described in a Spanish population 9. In the light of these data, and taking into account the role of CYP3A4 in the metabolism of irinotecan, we hypothesized that the presence of the CYP3A4*20 allele could help to identify patients who will develop severe irinotecan‐induced toxicity despite having a normal UGT1A1 genotype. We found one patient carrying the CYP3A4*20 allele (0.3%), a frequency similar to that found in a previous study recently performed by our group 19 and lower than that reported previously by Apellániz‐Ruiz et al. 9. The only patient in our cohort who carried the CYP3A4*20 variant was also heterozygous (*1/*28) for UGT1A1. She did not experience any of the toxicities evaluated in our study. Our findings add uncertainty about the effect of a loss‐of‐function CYP3A4 allele on severe irinotecan‐induced toxicity.

Competing Interests

There are no competing interests to declare.

This work was supported by a grant from the Departament d'Economia i Coneixement de la Generalitat de Catalunya (2016FI_B 00368 to Pau Riera).

Riera, P. , Salazar, J. , Virgili, A. C. , Tobeña, M. , Sebio, A. , Gallano, P. , Barnadas, A. , and Páez, D. (2018) Relevance of CYP3A4*20, UGT1A1*37 and UGT1A1*28 variants in irinotecan‐induced severe toxicity. Br J Clin Pharmacol, 84: 1389–1392. doi: 10.1111/bcp.13574.