Introduction

Colorectal cancer (CRC) is caused by uncontrolled cell proliferation in the epithelial cells present in the rectum, colon, or appendix. It is not only the third most commonly diagnosed cancer in males and the second in females, but also the fourth most common cause of cancer death [1]. In addition, CRC is a relative common cancer in China, especially for males in urban areas, since China had an estimated 310,244 new cases and an estimated 149,722 death cases in 2011 [2]. Two-thirds of the patients with CRC were diagnosed at a more advanced stage as the early-stage disease is mostly asymptomatic, and conventional chemotherapeutic regimens have been proven useful against advanced/metastatic CRC [3]. As an established cytotoxic regimen, irinotecan, 5-fluorouracil, and leucovorin (FOLFIRI) is an efficacious treatment for advanced CRC with approximately 2 years of overall survival [4]. However, resistance, toxicity, and side effects of chemotherapy remain great challenges in the long-term treatment of incurable metastatic diseases and may eventually contribute to death as tumors accumulate means of evading treatment [5]. Patients still show no benefit from FOLFIRI, and progress is made in order to find the differences in response of each individual on FOLFIRI, suggesting that genetic factors might have an important role in advanced CRC patients receiving irinotecan-based chemotherapy [6, 7].

Irinotecan (CPT-11), a main drug for the treatment of CRC, has limited clinical applications due to the presence of adverse reactions and obvious individual differences. The genetic polymorphism of drug metabolism is one of the main factors that cause the difference in adverse reactions among individuals [8]. Irinotecan is activated by hydrolysis to 7-ethyl-10-hydroxy-camptothecin (SN-38) by carboxylesterases in human tissues, and inhibition of topoisomerase I by active metabolite SN-38 results in inhibition of DNA replication and transcription, thereby producing cytotoxic effects [9, 10]. Study confirmed that uridine diphosphate glucoronosyltransferase 1A1 (UGT1A1) is a key enzyme in the metabolism and inactivation of SN-38 and UGT1A1 can transform SN-38 into glycosylated SN-38 (SN-38G), suggesting that the function and the gene polymorphism of UGT1A1 enzyme are closely related to irinotecan toxicity [11]. Lethal toxicity may cause failure of irinotecan-based chemotherapy in metastatic CRC patients, and clinical pharmacological evidence has revealed that irinotecan chemotherapy-induced toxicity is related to exposure to SN-38 and genetic polymorphisms in UGT1A1 gene [12]. More than 50 kinds of alterations of UGT1A1 gene loci have been reported in previous studies in which the relationship between UGT1A1*28 and UGT1A1*6 gene polymorphisms and adverse reactions of irinotecan is important, and UGT1A1*6/*28 may be useful as markers to prevent severe adverse reactions of irinotecan administration [11, 13, 14]. Studies have shown that there are significant differences in the distribution of polymorphism of UGT1A1 gene in different ethnic groups, which could lead to differences in the irinotecan-induced adverse reactions in different ethnic groups [15–18]. Specifically, UGT1A1*28 has been suggested as a biomarker for the prediction of irinotecan-induced neutropenia and toxicities in CRC in Caucasians [19]. However, prevalence of UGT1A1*28 mutations in Chinese Han population was significantly lower compared with American and European populations, and whether the UGT1A1*28 mutations can be used as a predictor of irinotecan toxicity remains to be investigated in Chinese Han population [20–22]. In addition, incidence of UGT1A1* 6 mutations (a variant in exon 1 of the UGT1A1 gene) in the Asian population is as high as 20 %, mainly found in East Asians, and UGT1A1* 6 is also an important risk factor associated with severe neutropenia and is important for the prediction of severe toxicity in patients receiving irinotecan treatment in Asian populations, while the potential role of UGT1A1* 6 in Caucasians has not been well clarified concerning irinotecan treatment [23–25]. Xinjiang is a multiethnic region since ancient times with complex national integration and gene exchange. The Han and Uygur were the main body accounting for 85 % of the total population in Xinjiang. Therefore, we carried out the present study to further clarify the relationship between UGT1A1*28/*6 and toxicity and clinical efficacy of irinotecan-based chemotherapy in patients with CRC in Uygur and Han population in Xinjiang, to explore the clinical significance of the detection of UGT1A1 gene polymorphisms on individualized treatment regarding different ethics.

In order to further clarify the distribution of human UGT1A1 gene polymorphism in Xinjiang Uygur and Han ethnicities, understand the association of UGT1A1 gene polymorphism and toxic reactions and therapeutic effects of irinotecan, and explore the clinical significance of the detection of UGT1A1 gene polymorphism on individualized treatment, this study conducted UGT1A1 gene polymorphism (UGT1A1*28 and UGT1A1*6) detection on 183 advanced CRC patients receiving irinotecan-based chemotherapy, and follow-up analyses were carried out.

Materials and methods

Study subjects

A total of 183 advanced CRC patients who underwent chemotherapy with irinotecan were chosen from the Affiliated Tumor Hospital of Xinjiang Medical University during November, 2012 to November, 2015. There were 114 Uygur people including 78 males and 36 females (mean age, 43.61 ± 15.82) and 69 Han people including 46 males and 23 females (mean age, 44.48 ± 14.77). The Uygur and Han subjects were selected if three generations of the subjects lived in the Xinjiang area and their families had no history of intermarriage. Inclusion criteria: CRC patients in a relatively good state, confirmed by histology or cytology; performance score from Eastern Cooperative Oncology Group (ECOG): 0–1 [26]; expected survival time over 3 months; above 18 years old; absolute value of neutrophil granulocyte ≥1.5 × 109/L; platelet count ≥80 × 109/L; normal kidney, liver, and bone marrow functions ; total bilirubin level ≤19.2 μmol/L; serum creatinine ≤105 μmol/L; no uncontrolled or severe internal diseases; no active infection; no other primary malignant tumors. This study was approved by Ethics Committee of the Affiliated Tumor Hospital of Xinjiang Medical University, and all the study subjects and their families had given their consent. All the patients were treated with second- or third-line chemotherapy regimen (irinotecan), and complete blood count was performed after treatment or before the next consultation.

Chemotherapy regimens

The chemotherapy regimens included FOLFIRI and a combination of irinotecan and capecitabine. FOLFIRI: first day, intravenous (iv) injection with irinotecan 150 mg/m2; first and second day, iv injection with tetrahydrofolic acid (THFC) 200 mg/m2 and 5-FU 400 mg/m2, and afterward with a continuous infusion pump with 5-FU 1200 mg/m2 for 44 h; repeated every 2 weeks, four cycles of treatment. Combination of irinotecan and capecitabine: first day, iv drip with irinotecan 150 mg/m2 for 90 min; second to the fifteenth day, iv drip with capecitabine 1000 mg/m2 (twice a day); repeated every 3 weeks.

Evaluation criteria for drug toxicity

FOLFIRI and irinotecan-induced severe drug toxicity generally occurred at the first four cycles of treatment. The adverse reactions of patients in the first four cycles were detected in our experimental process. Adverse reactions after chemotherapy were evaluated by National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 3.0 [27]. Adverse reactions included diarrhea and neutropenia. Adverse reactions were recorded in accordance with 0–V stages and III–IV mean severe diarrhea and neutropenia. According to the severity of adverse reactions, diarrhea, and neutropenia were divided into five stages. Diarrhea: Grade I, the number of cacation increased less than four times and the excretion increased slightly compared to the baseline values; Grade II, compared to the baseline values, the number of cacation increased to 4–6 times and the excretion moderately increased but not affect the normal life, and it is necessary to carry out intravenous infusion for less than 24 h; Grade III, compared to the baseline values, the number of cacation increased more than seven times and the excretion extremely increased and had impact on normal life, and patients need to be hospitalized and received intravenous infusion for more than 24 h; Grade IV, patients has hemodynamic disorders and other life-threatening symptoms; Grade V, death. Neutropenia: Grade I, neutropenia (1.5–2.0) × 109 L; Grade II, neutropenia (1.0–1.5) × 109 L; Grad III, neutropenia (0.5–1.0) × 109 L; Grade IV, neutropenia (0.5 × 109 L); and Grade V, death.

Evaluation criterion for curative effects

The curative effects were evaluated based on Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 [28] every 6 weeks through CT. The curative effects were divided into complete remission (CR): The target lesions disappeared completely, and the lymph nodes were all shorter than 10 mm in minor diameter, and no new lesions were found; partial remission (PR): the sum of major diameter of the patient’s baseline lesions were reduced by more than 30 %, and no new lesions were found; stable disease (SD): the sum of major diameter of the patient’s baseline lesions were reduced but did not reach PR or increased but did not up to PD, and no new lesions were found; progression of disease (PD): The sum of major diameter of the patient’s baseline lesions were increased by 20 % or more, the absolute value increased by 5 mm or more, or new lesions were found. Objective response rate (ORR) was calculated based on CR and PR; disease control rate (DCR) was calculated based on CR and PR and SD; overall survival (OS) was followed up.

Gene detection of UGT1A1

A total of 3 ml peripheral venous blood samples were collected before treatment and were placed in the ethylenediaminetetraacetic acid (EDTA) anticoagulant tube for anticoagulation. Genomic DNA was extracted according to QIAamp Blood kit (Qiagen, Hilden, Germany). After the purity of DNA was confirmed, the primer was designed with the following sequencing: forward 5-CCTGCTACCTTTGTGGACTGAC-3 and reverse 5-TGCCCGAGACTAACAAAAGACT-3. DNA fragment included UGT1A1*6 (211 G>A) and UGT1A1*28 (−53 *1 allele >*28) which were amplificated by polymerase chain reaction (PCR). PCR system (25 μL): 5-ng template DNA, 10 × KOD plus buffer, 1 mmol/L dNTPs, 0.4 mmol/L MgSO4, 0.5 U KOD plus enzyme (TOYOBO CO., LTD., Osaka, Japan), and 0.25 gmol/L forward and reverse primers. Reaction conditions: denaturation at 95 °C for 5 min, then another 15 s at 95 °C, 25 s at 60 °C and 30 s at 72 °C which was cycled for 40 times, and then being extended for 10 min at 72 °C. The amplification products were kept at −20 °C and analyzed using ABI-3730 through two-way analysis. Polyphred 5.04 was used to detect and analyze SNP and perform manual reading of genotype of UGTlA1*28 and UGTlA1*6.

Statistical methods

SPSS 21.0 (SPSS Inc., Chicago, IL, USA) was used to analyze the data. Measurement data were expressed as mean ± SD (\( \bar{x} \) ± s). Comparison between two groups was performed using an independent sample t test. The representativity of the samples was analyzed by Hardy–Weinberg disequilibrium test. Clinical features, differences between adverse reactions of irinotecan and curative effects were analyzed through Chi-square test. Fitting curves were analyzed using Kaplan–Meier survival analysis, and influence factors were analyzed by Cox regression analysis. A P value of less than 0.05 was regarded as significant.

Results

Baseline characteristics

Baseline characteristics of the patients can be seen in Table 1. For the 114 Uygur patients, we can see that they were aged from 18 to 72 (average: 43.61 ± 15.82 years), and there were 78 males and 36 females, while the 69 Han patients were aged from 18 to 71 (average: 44.48 ± 14.77 years), with 46 males and 23 females. There was no difference in age, male/female (M/F) ratio, ECOG score, location of primary tumor, degree of differentiation, carcinoembryonic antigen (CEA) levels, chemotherapy line, and regimen between the two nationalities (all P > 0.05).

Table 1 Baseline characteristics in Ugyur and Han patients
Full size table

Distribution of genotype and allele frequencies

The distribution of genotype and allele frequencies of UGT1A1*28/*6 were consistent with the Hardy–Weinberg disequilibrium (all P > 0.05). In the 114 Uygur patients, the variants of UGT1A1*28 and UGT1A1*6 were divided into three types: double wild type (DW) (*1/*1 and G/G, n = 55, 48.24 %), single variant (SV) (G/A and *1/*1 or G/G and *1 allele *1/*28, n = 32, 20.07 %), and double variant (DV) (G/G and *28/*28 or A/A and *1/*1 or G/A and *1/*28, n = 27, 23.69 %). In the 69 Han patients: DW (*1/*1 and G/G, n = 30, 43.48 %), SV (G/A and *1/*1 or G/G and *1/*28, n = 34, 49.28 %), and DV (G/A and *1/*28 or A/A and *1/*1, n = 5, 7.25 %). Obvious differences in the distribution frequencies of *1/*1, *1/*28, and *28/*28 of UGT1A1*28 and DW, SV, and DV of UGT1A1*28/*6 were found between Uygur and Han patients (P = 0.02 and P = 0.002). There was no difference in distribution frequencies of UGT1A1*28 allele and UGT1A1*6 genotype and allele between the two ethnicities (P > 0.05) (Table 2).

Table 2 Distribution of genotype and allele frequencies of UGT1A1*28/*6 in Ugyur and Han patients
Full size table

Gene polymorphism and clinical features

There was no difference in age, sex, ECOG score, location of primary tumor, degree of differentiation, CEA level, chemotherapy line and regimen among Ugyur patients with DW, SV, and DV genotypes (all P > 0.05); there was also no difference found in age, sex, ECOG score, location of primary tumor, degree of differentiation, CEA level, chemotherapy line, and regimen among Han patients with DW, SV, and DV genotypes (all P > 0.05) (Table 3).

Table 3 Association of gene polymorphisms and clinical features in Uygur and Han patients
Full size table

UGT1A1*28/*6 polymorphisms and adverse reactions of irinotecan

Among the 114 Uygur patients, 27 (23.69 %) suffered from severe delayed diarrhea and 14 (12.28 %) had severe neutropenia; among the 69 Han patients, 16 (23.19 %) had severe delayed diarrhea and 11 (15.94 %) had severe neutropenia (Table 4). Uygur patients carrying UGT1A1*28 mutant *1/*28 and *28/*28 genotypes were more likely to suffer from diarrhea (I/II and III/IV) compared with those with wild-type *1/*1; compared with mutant G/A and A/A in UGT1A1*6, wild-type G/G genotype significantly decreased the risk of diarrhea (I/II and III/IV) (all P < 0.05). Han patients carrying mutant UGT1A1*28 *1/*28 and *28/*28 genotypes were more likely to suffer from diarrhea (I/II and III/IV) compared with those with wild-type *1/*1; in UGT1A1*28, incidence of diarrhea (I/II and III/IV) was lower in patients with wild-type G/G than those with mutant G/A and A/A (all P < 0.05). In addition, there was no difference in incidence of neutropenia (I/II and III/IV) among Uygur and Han patients with different UGT1A1*28 and UGT1A1*6 genotypes (all P > 0.05). Compared with DW genes, there were significant differences in incidence of neutropenia (I/II and III/IV) in Uygur and Han patients with heterozygous and homozygous type (P < 0.05) (Table 5).

Table 4 Distribution of diarrhea and neutropenia stages of Uygur and Han patients with advanced colorectal cancer
Full size table
Table 5 Association of genotypes of UGT1A1*28/*6 and adverse reactions of irinotecan in Uygur and Han patients
Full size table

UGT1A1*28/*6 polymorphisms and chemotherapy efficacy

Chemotherapy effects of the 114 Ugyur patients treated with irinotecan were evaluated: CR (n = 0), PR (n = 20), SD (n = 48), and PD (n = 46). In UGT1A1*28, there were differences in clinical responses (ORR and DCR) to *1/*1, *1/*28, and *28/*28 (P = 0.001 and P = 0.004) polymorphisms but no difference was found to UGT1A1*6 polymorphism (P > 0.05). The therapeutic efficacy of the 69 Han patients was assessed: CR (n = 0), PR (n = 33), SD (n = 14), and PD (n = 22). There was no difference in clinical responses (ORR and DCR) to UGT1A1*28 (P = 0. 618 and P = 1.00) and UGT1A1*6 (P = 0.821 and P = 0.946) polymorphisms (Table 6).

Table 6 UGT1A1*28/*6 polymorphisms and chemotherapy efficacy in Uygur and Han patients
Full size table

UGT1A1*28/*6 polymorphisms and OS

The median survival time was 25 months in 114 Ugyur patients, there were differences in OS in patients with UGT1A1*28 *1/*28, *28/*28, and *1/*1 (P < 0.05); differences were also found in patients with UGT1A1*6 AA and GG (P < 0.05); compared with double wild-type genes, single-point mutant and double-point mutant genes had different influences on OS (P < 0.05) (Fig. 1). The median survival time was 24 months in 69 Han patients, there were differences in OS in patients with UGT1A1*28 *1/*28, and *1/*1 (P < 0.05); differences were also revealed in UGT1A1*6 AA and GG (P < 0.05); compared with double wild-type genes, single-point mutant and double-point mutant genes had significantly different influences on OS (both P < 0.05) (Fig. 2). Multivariate survival analysis was performed among Ugyur and Han patients using Cox regression model to analyze influencing factors on OS included age, M/F ratio, ECOG score, location of primary tumor, degree of differentiation, CEA level, number of mutant integrated allele, chemotherapy regimen, and line. The results showed that for Ugyur patients, UGT1A1*28, UGT1A1*6, the number of mutant integrated alleles and chemotherapy line had significantly different influences on OS (all P < 0.05), but for Han patients, only the number of mutant integrated alleles had influenced OS (P < 0.05) (Table 7).

Fig. 1
figure 1

Comparison of survival curves of UGT1A1*28, UGT1A1*6, and UGT1A1*28/*6 among Ugyur patients. a Survival curve of UGT1A1*28 in Ugyur patients; b survival curve of UGT1A1*6 in Ugyur patients; c survival curve of UGT1A1*28/*6 in Ugyur patients

Full size image
Fig. 2
figure 2

Comparison of survival curves of UGT1A1*28, UGT1A1*6, and UGT1A1*28/*6 among Han patients. a Survival curve of UGT1A1*28 in Han patients; b survival curve of UGT1A1*6 in Han patients; c survival curve of UGT1A1*28/*6 in Han patients

Full size image
Table 7 Cox regression model to analyze the factors affecting overall survival (OS) in Uygur and Han patients
Full size table

Discussion

Currently, irinotecan is widely used as a clinical topoisomerase I inhibitor, especially in the treatment of gastric cancer, CRC, small cell lung cancer, and other solid tumors, and clinical trials have demonstrated that irinotecan can significantly prolong the patient’s life [13, 29, 30]. However, clinical application of irinotecan is limited due to its important dose dependent toxicity, and many scholars are working on evaluating predictors of irinotecan-induced toxicity in order to improve safety and efficacy of irinotecan application, especially in advanced and/or metastatic CRC [31, 32]. Several important clinical trials suggest that genetic factors may play important roles in drug metabolism, distribution, and toxicity of irinotecan. UGTIA1 is currently known as the main isozyme involved in irinotecan metabolism [33, 34]: UGTIA1 gene polymorphism influences the UGTIA1 enzyme activity, and pharmacogenetic studies on irinotecan have revealed the close relationship between UGT1A1 polymorphisms and adverse reactions of irinotecan and therapeutic effects in patients with advanced CRC receiving irinotecan-based chemotherapy [25, 35].

Results in our study suggested significant differences in distribution of *1 allele*1/*1, *1 allele *1/*28*1/*28, and *28 allele*28/*28 of UGT1A1*28, and DW, SV, and DV of UGT1A1*28/*6 gene polymorphisms were found in the Uygur ethnicity when compared to the Han ethnicity, while the allele frequency distribution of UGT1A1*28 and genotype and allele frequency distribution of UGT1A1*6 were not statistically significant between the two ethnic groups. Anthropological studies have shown that the Uygur population contains Caucasian origins and has a unique genetic background. East Asians have less frequent UGT1A1*28 and more frequent UGT1A1*6 compared to that in Caucasians [15]. Intra-ethnic differences in genetic variation of UGT1A1 gene have been analyzed previously in three ethnic groups of Chinese populations, and heterogeneity among different ethnic populations could be a result of microevolution [22]. In our study, the frequency of UGT1A1*28 (*1/*1) in Han was slightly higher than in Ugyur (66.7 vs. 65.78 %), as previously reported, and Chinese patients showed a higher frequency of wild-type UGT1A1*28 (*1/*1) compared with Caucasian population (69.9 vs. 45.2 %) in CRC patients receiving irinotecan [36]. Similarly, the associations between UGT1A1 *28 and *6 polymorphisms and irinotecan toxicity were determined in Thai patients with CRC. Atasilp C, et al. found that the frequencies of *1/*28 and *28/*28 in UGT1A1*28 polymorphism were 20.5 and 2.3 % in Thai, 24. 62 and 9.6 % in Ugyur, and 33.33 % and 0 in Han, respectively, and the frequencies of G/A in UGT1A1*6 polymorphism were 15.9 % in Thai, 29.82 % in Ugyur and 24.64 % in Han, respectively, showing a certain difference. Consistent with our results, the study of Atasilp C, et al. suggested that UGT1A1 *28 and *6 in combination may have an increased risk of irinotecan-induced neutropenia in Thai CRC patients [37]. In addition, Sukasem et al. [38] found that the allele frequencies for UGT1A1 genetic polymorphisms were *1/*1 (54.95 %), *1/*6 (13.19 %), *1/*28 (25.27 %), *28/*6 (4.40 %), and *28/*28 (2.20 %), with no homozygous mutation UGT1A1*6, suggesting that the UGT1A1*28 and UGT1A1*6 alleles were similar in the Asian populations. Genotypes of UGT1A1 *28 and *6 were analyzed by pyrosequencing technique in studies conducted by Atasilp et al. [37] and Sukasem et al. [38]. In our study, the methods for detection of UGT1A1 were PCR amplification and the direct sequencing since PCR amplification and the direct sequencing had the characteristics of strong specificity, high sensitivity, simple operation and were time saving and easy to be standardized and automated [39].

In addition, neutropenia and diarrhea are the dose-limiting toxicities. Our findings revealed that advanced CRC patients carrying wild genotypes of UGT1A1*28 and UGT1A1*6 have significantly lower incidence of diarrhea in I/II stage and III/IV stage than those of mutant genotypes in irinotecan-based chemotherapy in the Uygur ethnicity and Han ethnicity. But this is without distinct difference in neutropenia incidence in I/II stage and III/IV stage, suggesting that UGT1A1*6/*28 could be used as markers to prevent induction by irinotecan administration [13]. The polymorphism of UGT1A1 gene is closely related to the function of the enzyme, and UGT1A1 gene polymorphism may cause decreased or absence of UGT1A1 enzyme activity, and influencing the effects of irinotecan metabolism in human body, resulting in the accumulation of bioactive metabolite SN-38 in the body, thereby producing related adverse reactions [9, 40, 41]. Consistently with our findings, the UGT1A1*28 and UGT1A1*6 could be considered as predictors for severe delayed diarrhea associated with irinotecan, but without association with severe neutropenia in irinotecan-based regimens in CRC [42]. The homozygous variant of UGT1A1*6 served as a risk factor for severe diarrhea, and UGT1A1*6 polymorphisms can be used as potential biomarkers for predicting irinotecan-induced severe toxicity in patients from Asia [35]. The change of TATA box of thymine and adenine (TA) repeat sequences may influence the activity of UGT1A1 gene in the drug metabolism and its (TA) 7 repeats (UGT1A1*28 gene polymorphism) can significantly increase the risk of severe granulocyte reduction and diarrhea in CRC patients treated with irinotecan-based chemotherapy in Caucasians [43]. Moreover, a combined test of UGT1A1*6 and UGT1A1*28 might be a potential biomarker of irinotecan-induced neutropenia in Asians, which still needs to be confirmed by additional research [20].

With respect to association between UGT1A1*28/*6 gene polymorphisms and efficacy and survival of irinotecan-based chemotherapy, genotypes of UGT1A1*28 are associated with ORR and DCR in Uygur patients. Compared with *1 allele *1/*1 wild-type genotype,*1 allele *1/*28*1/*28 mutant genotype of UGT1A1*28 is associated with shorter OS in Uygur and Han nationality. Compared with wild-type DW, SV and DV mutant genotypes of UGT1A1*28/*6 are associated with shorter OS in Uygur and Han nationality. Cox regression model analysis further confirmed that UGT1A1*28, UGT1A1*6, combined genotypes and chemotherapy line are factors who have a significant impact on OS. It has been proposed that UGT1A variants together with UGT1A1*28 might help predicting the outcome of CRC patients treated with FOLFIRI [11]. It has been found that homozygous UGT1A1*28 allele carriers have increased survival rates and higher tumor response rates [44]. The response rate was higher in *28 allele/*28 allele patients compared with *1 allele/*1 allele in metastatic CRC patients treated with FOLFIRI, but a nonsignificant survival advantage was observed for *28 allele/*28 allele when compared with *1 allele/*1 allele patients [45]. However, many studies reported contrary points or even slight evidence regarding the association of response rates and OS and genotypes of UGT1A1*28 and UGT1A1*6 in Chinese patients with advanced CRC in irinotecan-based regimens [36, 42, 46]. UGT1A1*28 polymorphism cannot be considered as a reliable predictor of TR and PFS in CRC patients treated with IRI-based chemotherapy in a meta-analysis in Caucasians [47]. Therefore, the association between clinical response nor prognosis and UGT1A1 gene polymorphisms needs to be further confirmed by more well-designed studies with the ultimate goal of achieving personalized irinotecan-based chemotherapy.

In summary, UGT1A1 gene polymorphism can predict irinotecan-induced adverse reactions in advanced CRC patients in individuals from Xinjiang Uygur ethnicity and Han ethnicity. UGT1A1 gene polymorphism is associated with clinical response and prognosis of irinotecan-based chemotherapy in persons from Uygur ethnicity, and only associated with prognosis in persons from Han nationality. However, the recruited subjects received FOLFIRI regimen including irinotecan and 5-FU. Both irinotecan and 5-FU may induce similar side effects (severe neutropenia and diarrhea). Therefore, part of the toxic side effects in our study may be possibly caused by 5-FU, and the specific effect of UGT1A1 polymorphism in each treatment regimen should be further investigated. In addition, more fully and strongly clinical research evidence is needed to reflect the detection of the relevant genetic polymorphisms which can be used to correctly guide the clinical safety and effectiveness of the use of irinotecan and to achieve the purpose of predicting serious adverse reactions.