Introduction

Since combination chemotherapy showed survival benefits compared to best supportive care (BSC) in early 1990s, many phase III trials of first-line chemotherapy including molecular target agents have been conducted. Nowadays, combination chemotherapy with fluoropyrimidine and platinum, with trastuzumab in case of positive HER2, is globally recognized as the standard first-line chemotherapy for advanced gastric cancer [1,2,3]. In 2010’s, survival benefits of monotherapy with irinotecan and docetaxel in the second-line chemotherapy were confirmed in phase III trials [4,5,6], and irinotecan and weekly paclitaxel showed equivalent efficacy in the WJOG4007 trial [7]. Since superiority of ramucirumab plus weekly paclitaxel over weekly paclitaxel alone was confirmed in the RAINBOW trial, it has been recognized as the standard second-line chemotherapy [8], and irinotecan monotherapy has been widely used as the third-line chemotherapy especially in Asian countries. Recently, nivolumab has shown survival benefits compared to best supportive care as the salvage line treatment after at least two prior chemotherapy regimens [9]. Therefore, both nivolumab and irinotecan monotherapy are recommended as the third or later line treatment in the Japanese Gastric Cancer Treatment Guideline.

Irinotecan is hydrolyzed to 7-ethyl-10-hydroxy-camptothecin (SN-38) by carboxylesterases, and this active metabolite SN-38 inhibits topoisomerase I leading to inhibition of DNA replication and transcription [10, 11]. In clinical practice, however, irinotecan sometimes causes serious adverse events such as neutropenia and delayed-onset diarrhea. Uridine diphosphate glucuronosyltransferase 1A1 (UGT1A1) is a key enzyme in the metabolism and excretion of SN-38 as it transforms SN-38 into glycosylated SN-38 (SN-38G) which is excreted into the bile juice [12,13,14]. UGT1A1 has genomic polymorphisms such as the UGT1A1*28 and *6, and activity of UGT1A1 is reduced in individuals with these polymorphisms. Actually, previous studies reported that UGT1A1 polymorphisms, especially UGT1A1*28 and UGT1A1*6, is associated with irinotecan-induced severe toxicities in metastatic colorectal cancer patients [15]. Approximately 10% of the North American population is homozygous for the UGT1A1*28 allele [16], and patients with homozygous genotype for the UGT1A1*28 allele showed an increased risk of neutropenia caused by irinoetecan. Similarly, the UGT1A1*6 variant is frequent in the Asian population (approximately 20%, mainly in East Asia), and metastatic colorectal cancer patients with UGT1A1*6 variants also showed a higher incidence of severe neutropenia [17, 18]. Therefore, it is recommended that dose reduction of irinotecan should be considered for patients with UGT1A1*28 and/or UGT1A1*6.

However, there are few reports on the associations between UGT1A1*6/*28 polymorphisms and irinotecan-induced toxicities and efficacy in patients with advanced gastric cancer. The purpose of this study is to investigate the relationship between UGT1A1 genotype and safety and efficacy of irinotecan alone as third-line chemotherapy for advanced gastric cancer.

Patients and methods

Patients

As the source of the subjects in this retrospective analysis, medical records of the 208 patients, who were tested for UGT1A1*6 (G/G, G/A, and A/A) and *28 (6/6, 6/7, and 7/7) genotypes and treated with irinotecan-based chemotherapy for advanced gastric cancer from 2009 to 2014 at National Cancer Center Hospital, Tokyo, were reviewed. Main selection criteria for this study were as follows: histologically confirmed gastric adenocarcinoma, unresectable or recurrent disease, refractory both to a fluoropyrimidine-containing first-line chemotherapy, including relapse < 24 weeks after the final dose of adjuvant chemotherapy with S-1, and to a taxane-containing second-line chemotherapy (paclitaxel, nab-paclitaxel or docetaxel), 20 years or older, Eastern Cooperative Oncology Group (ECOG) performance status (PS) from 0 to 2, sufficient bone marrow, hepatic and renal function, no active concomitant malignancy, and no serious complications.

Treatment

As a standard regimen, irinotecan at a dose of 150 mg/m2 was administered intravenously every 2 weeks. The treatment was repeated until disease progression, unacceptable toxicity, or patient refusal. Dose modifications of irinotecan, including the initial dose, and treatment delays were decided by the physician’s discretion according to the patient general condition, organ function, severity of hematological or non-hematological toxicities, and UGT1A1 genotype. In general, the dose was reduced when grade 4 neutropenia, grade 3 or higher febrile neutropenia, and unacceptable grade 2 or higher diarrhea, nausea, and anorexia occurred.

UGT1A1 genotyping

The genotype of UGT1A1*6 and UGT1A1*28 were tested using the Invader® UGT1A1 Molecular Assay kit (Sekisui Medical Co. Ltd, Tokyo, Japan) Genomic DNA for determined to genotype was extracted from mononuclear cells in peripheral blood. genotype was classified to the following three groups: wild-type (WT) with UGT1A1*6 G/G and *28 6/6, single heterozygosity (SH) with UGT1A1*6 G/G and *28 6/7, or *6 G/A and *28 6/6, and homozygosity/double heterozygosity (Homo/DH) with *6 A/A and *28 6/6 (Homo for *6), *6 G/G and *28 7/7 (Homo for *28), or UGT1A1*6 G/A and *28 6/7 (DH).

Calculation of relative dose intensity (RDI)

The dose intensity of irinotecan was calculated by dividing its total actual administered dose by the number of weeks of treatment. The relative dose intensity (RDI) was calculated as a percentage of the dose intensity to the planned target dose (75 mg/m2).

Evaluation and statistical analysis

The efficacy and safety of third-line irinotecan monotherapy were compared among the three groups classified by the genotypes for UGT1A1*28 or UGT1A1*6 (wild, SH and Homo/DH). Tumor response was evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST), version 1.1. Adverse events were assessed according to the Common Terminology Criteria for Adverse Events (CTCAE), version 4.0. Overall survival (OS) was calculated from the date of initiating irinotecan to the date of death from any cause or censored on the last follow-up. Time to treatment failure (TTF) was defined as the interval between the first and last administration of irinotecan. Kaplan–Meier estimates were used to calculate survival probabilities, and the log-rank test was used to compare survival curves. Survival differences between among groups were also evaluated by multivariate analyses using the Cox proportional hazard regression model, and presented as the hazard ratio (HR) and 95% CI. Multivariate analysis included well-known prognostic factors for advanced gastric cancer such as age (≤ 65 versus > 65 years), gender (male versus female), ECOG PS (0–1 versus ≥ 2), number of metastatic sites (1 versus ≥ 2), histology (intestinal versus diffuse), as well as genotype (WT, SH versus Homo/DH). Statistical analyses were performed using the statistical software ‘EZR’ (Easy R), which is based on R and R commander.

Results

Patient characteristics

A total of 208 patients were the source of this retrospective study. Excluding 91 patients receiving combination of chemotherapy containing irinotecan and those receiving irinotecan monotherapy as second- (n = 41) or later than 3rd (n = 2) line treatment (Fig. 1). Finally, 74 patients were identified as the subjects of this retrospective study, who were classified into three groups by the genomic types of the UGT1A1*28 and UGT1A1*6: WT (n = 37, 50%), SH (n = 27, 36.5%), and Homo/DH (n = 10, 13.5%). Their baseline characteristics are shown in Table 1.

Fig. 1
figure 1

CONSORT diagram of gastric cancer patients tested for the UGTA1 genotype

Full size image
Table 1 Patients’ characteristics in third-line chemotherapy
Full size table

Initial dose and drug delivery of irinotecan

The initial dose of irinotecan was reduced in 27% (10/37) of the patients in the WT group (120 mg/m2, 6 patients; 100 mg/m2, 4 patients), in 33% (9/27) of those in the SH group (120 mg/m2, 8 patients; 100 mg/m2, 1 patient), and in 70% (7/10) of those in the Homo/DH group (120 mg/m2, 3 patients; 100 mg/m2, 4 patients). In the total 26 patients whose initial dose was reduced, six patients (23%) required further dose reduction due to adverse events. The relative dose intensity of irinotecan was lower in the Homo/DH group (52.5%) than in the WT (77.6%) and SH groups (70.3%). The median number of administration of irinotecan was smaller in the Homo/DH group (2 cycles, range: 1–6 cycles) than in the WT (4 cycles, range: 1–34 cycle) and SH groups (3 cycles, range: 1–22 cycles). With a median follow-up of 7.2 months, the median time to treatment failure (mTTF) was 2.4 months (95% CI 1.6–3.6) in the WT group, 2.3 months (95% CI 1.3–3.7) in the SH group, and 1.3 months (95% CI 0.3–1.9) in the Homo/DH group (Fig. 2). Reasons for treatment discontinuation were disease progression (78.3%), hematological adverse events (2.7%) and non- hematological adverse events (5.4%) and others (13.6%) in the WT group, and those were 59.2%, 7.4%, 7.4% and 26.0% in the SH group, 30.0%, 40.0%, 10.0% and 20.0% in the Homo/DH group, respectively.

Fig. 2
figure 2

Time to treatment failure of third-line irinotecan monotherapy in the WT, SH, and Homo/DH groups

Full size image

Toxicity of irinotecan in third-line chemotherapy

The hematological and non-hematological adverse events observed during the irinotecan monotherapy are listed in Table 3. The incidences of grade 3 or 4 neutropenia (50%) and febrile neutropenia (40%) in the Homo/DH group were significantly higher than in the other groups. The incidences of grade 3 or 4 diarrhea in the WT, SH, and Homo/DH groups were 8.1%, 3.7%, and 10%, respectively. No treatment-related deaths were observed in any group (Table 2).

Table 3 Response in measurable lesions
Full size table
Table 2 Adverse events of special interest
Full size table

Efficacy of irinotecan as third-line treatment

Among the 36 patients with measurable lesion at baseline, 2 achieved a partial response (5.5%), 14 showed stable disease (39%), and 20 showed progression disease (55.5%), resulting in a response rate of 5.5%, and disease control rate of 44.5%. Disease control rate was 47.6% in the WT group, 41.7% in the SH group, and 33.3% in the Homo/DH group, respectively (Table 3). Median overall survival was 4.6 months (95% CI 3.4–6.2) in all patients. Median overall survival was 6.9 months (95% CI 3.9–7.9) in the WT group, 6.3 months (95% CI 3.3–8.3) in the SH group, and 2.8 months (95% CI 1.2–3.4) in the Homo/DH group, respectively (Fig. 3).

Fig. 3
figure 3

Kaplan–Meier analysis of overall survival (third-line chemotherapy) in the WT, SH, and Homo/DH group

Full size image

Prognostic factors

Univariate and multivariate analyses for prognostic factors of OS are shown in Table 4. Risk factors for poor OS were the UGT1A1 genotype (p = 0.034) in univariate analysis and ECOG PS 2 (p < 0.0001) in univariate and multivariate analysis.

Table 4 Univariate and multivariate analysis for overall survival
Full size table

Discussion

In summary, this study showed association between the UGT1A1 genotype and clinical outcomes such as disease control, time to treatment failure, overall survival, and severe adverse events in Japanese patients with advanced gastric cancer who received irinotecan monotherapy in the third-line setting.

As for toxicity, irinotecan monotherapy was well tolerated in the WT and SH groups. In contrast to the results that there were no differences in incidences of any toxicities between the WT and SH groups, the incidences of grade 3 or 4 adverse events in the Homo/DH group were higher than those in the WT and SH groups, especially leucopenia, febrile neutropenia, and diarrhea. The homozygous genotype was associated with a twofold higher risk of severe neutropenia compared with the WT and SH groups, despite using a lower starting dose in the Homo/DH group. These results were consistent to previous similar study of patients with advanced colorectal cancer. Ichikawa et al. evaluated the effects of UGT1A1 genotypes and non-genetic factors on the efficacy and safety of irinotecan-based regimens in 1312 patients with advanced colorectal cancer [19]. As for toxicities, the homozygous genotype was associated with a twofold higher risk of severe neutropenia in the first course as compared with wild-type, despite using a lower starting dose in the homozygous group. In this study, although 70% of patients in the Homo/DH group received initially reduced dose of irinotecan, they experienced severe adverse events. Therefore, it seems difficult to use the standard dose of irinotecan for patients with UGT1A1 Homo/DH polymorphism.

Furthermore, this study showed that survival benefits of third-line irinotecan was lower in patients with the Homo/DH genotype UGT1A1 polymorphism compared with patients with the WT and SH polymorphisms (OS: 2.8 versus 6.9 and 6.3 months, respectively). The median OS in the WT and SH groups was similar to those reported in previous retrospective studies of third- or later-line treatment with irinotecan ranging 4.0–6.6 months [20,21,22]. However, the median OS in the Homo/DH group seemed as short as those of best supportive care in pivotal phase III studies (2.4, 3.6, 3.8, 4.1, and 4.3 months in AIO study [4], COUGAR-02 study [5], Korean study [6], ONO-4538-12 study [9], and GRANITE-1 [23] study, respectively). Although there were no remarkable differences in disease control rates among the three groups, time to treatment failure is shorter in the Homo/DH group than others, and there are more patients whose treatment was discontinued due to adverse events. It is considered that serious adverse events which caused short TTP and deteriorated the patient’s condition might cause lack of benefits of irinotecan for the patients in the Homo/DH group.

Recently, nivolumab has been shown to provide a significant survival benefit in patients with refractory to standard treatment (ONO-4538-12 study), and the Ministry of Health Labour and Welfare of Japan approved nivolumab for advanced gastric cancer as salvage line treatment in 2017. In the Gastric Cancer Treatment Guide Lines 2018 published by the Japanese Gastric Cancer Association, both irinotecan and nivolumab are recommended in the third or later line treatment. Among patients assigned to best supportive care were enrolled in the ONO-4538-12 study, 25% of them did not receive irinotecan as prior treatment and only 5.5% of them received irinotecan after placebo. Therefore, precisely, it is still unclear whether irinotecan or nivolumab should be administered first if UGT1A1 polymorphism is not taken into account. However, together with the efficacy and toxicities of patients with the UGT1A1 Homo/DH polymorphism in this study, genotype testing of UGT1A1 is recommended for selecting the third-line therapy either irinotecan or nivolumab. And it is suggested that irinotecan monotherapy as the third-line therapy may not be selected preferentially than nivolumab for advanced gastric cancer patients with the UGT1A1 Homo/DH polymorphism.

The current study has some limitations. This is a retrospective study which could not collect the precise data of adverse events and quality of life, no data of pharmacokinetics, and there were no pre-specified criteria for dose reduction, rest and discontinuation of irinotecan. The small sample size at single-center, especially, including only 10 UGT1A1 Homo/DH patients, could not adjust the difference in patient’s background even by multivariate analysis. These limitations might make some bias in this study.

In conclusion, since Homo/DH patients showed unfavorable clinical outcomes associated with a high risk of grade 3 or higher adverse events, irinotecan may not be administered as third-line or later-line therapy in advanced gastric cancer patients with the UGT1A1 Homo/DH polymorphism.