A systematic literature search was conducted for articles published until December 2020 discussing the efficacy and safety of high irinotecan doses in wild-type patients or heterozygous for the UGT1A1*28 allele. The databases reviewed were Medline and Embase.

The MeSH and Emtree terms used are summarized in table 1. The methods used for the study were based on the recommendations of the PRISMA Statement (Preferred Reporting Items for Systematic Reviews and Meta-Analyses)18, applying both the checklist and the flow diagram.

The inclusion criteria were defined in advance with consideration of the two secondary goals of this review, specifically:

• 1.

  To analyze the magnitude of the differences between the clinical response to high doses of irinotecan and the response to the doses recommended in the product's SmPC for the FOLFIRI regimen in patients with mCRC with the UGT1A1*1/*1 or *1/*28 genotype.

• 2.

  To study the magnitude of the differences between the adverse events resulting from high doses of irinotecan and those arising from the doses recommended in the product's SmPC in patients with mCRC with the UGT1A1*1/*1 or *1/*28 genotype.

Studies inclusion was decided based on whether: (a) they were observational cohort studies or clinical trials; (b) patients with mCRC received higher doses of irinotecan than those currently recommended in the drug's SmPC; (c) patients were genotyped to identify the presence of the UGT1A1*28 allele and patients with the UGT1A1 *1/*1 genotype (wild-type) were included; and (d.1) the association between the UGT1A1 *1/*1 genotype, high doses of irinotecan and the objective response rate (ORR) was mentioned (patients were deemed to be responsive when a full or partial response was obtained), or (d.2) the association between the UGT1A1 *1/*1 genotype, high doses of irinotecan and safety was assessed. Studies not published in English or Spanish were excluded. The inclusion and exclusion criteria are detailed in table 2.

The four authors evaluated independently the article abstracts to determine their eligibility in accordance with the inclusion/exclusion criteria outlined in table 2. All original peer-reviewed articles published in English that evaluated the efficacy and/or toxicity of high doses of irinotecan (> 180 mg/m2) according to the UGT1A1 genotype in adult patients with mCRC were considered for inclusion. Once the eligible abstracts were selected, the full-text versions of all potentially eligible texts were retrieved, and each of the four authors once again independently applied the inclusion/exclusion criteria. Finally, the references included in the study were accurately analyzed to find potentially relevant articles not identified in the initial systematic search. Articles were deemed to be worthy of inclusion when at least three of the four authors agreed with the pre-established criteria. The inter-class correlation (ICC) coefficient was calculated to determine the degree of concordance between the authors regarding article selection.

The reviewers created, tested and improved a data extraction template. The data extracted included: first author and year of publication; type of study and level of evidence; participants and type of cancer; irinotecan regimen and dosage; genotype and prevalence; clinical variables assessed [ORR, overall survival (OS) and progression-free survival (PFS)]; results and discussion (Tables 3 and 4).

The severity of adverse events was established in accordance with the Common Toxicity Criteria of the National Cancer Institute (now called the Common Terminology Criteria for Adverse Events; CTCAE), when they were present (Table 4)25.

The methodological quality of eligible studies was evaluated based on discussions and consensus-building among all the authors of the study, with special focus on the use of objective evaluation criteria and a representative sample. The quality of the observational studies included was evaluated in accordance with the recommendations of the Newcastle-Ottawa (NOS) scale26,27. Eight items were selected for inclusion in the study, including patient selection, comparability between groups, and exposure factors. Studies with NOS scores from 0 to 3, 4 to 6 and 6 to 9 were considered of low, medium and high quality, respectively.

The methodological clinical trials quality was evaluated with Jadad,’s scale, also known as Jadad's score or Oxford quality scoring system, a procedure used to independently evaluate the methodological quality of a clinical trial. The scale assigns a score which rates study as low quality if the score < 3 points, and as rigorous in the case of randomized clinical trials (RCTs), if the score is > 5 points28.

The search strategy yielded a total of 595 references, of which 13 were selected for the analysis of the effectiveness and safety of high doses of irinotecan (>180 mg/m2) when used as part of the FOLFIRI regimen. Specifically, 11 clinical trials and two observational studies were included. ICC was 0.90 (95% CI 0.84–0.94), for the total number of studies included; 1 (95% CI 1–1) for studies dealing with effectiveness; and 0.83 (95% CI 0.71–0.92) for studies dealing with safety.

Figure 1 shows the flow diagram (adapted from PRISMA) showing the process followed to include the 13 studies, 6 of them evaluating toxicity and 7 evaluating both effectiveness and toxicity, in patients treated for mCRC with high doses of irinotecan.

Figure 1.

Flow diagram of the articles included and excluded at each stage.

(0,2MB).

Table 3 shows the descriptive characteristics of the reviewed studies, including the type of study, participants, regimen, dose and rate of irinotecan administration, genotypes and prevalence, clinical variables evaluated, results and discussion.

A total of 992 patients treated with irinotecan with a FOLFIRI regimen for mCRC were included. In all cases, irinotecan was administered as first-line treatment (12, 92.3%) except for one study where it was administered as second-line treatment (1, 77%).

The assessment of the quality of the observational studies was carried out using the NOS scale, with all observational studies exhibiting a highquality level (Table 5). The evaluation of the quality of clinical trials, conducted using Jadad's scale, revealed that the phase I-II and IV clinical trials included in the present review, which analyzed the effectiveness and/or safety of high doses of irinotecan, were of low quality as all 11 (100%) clinical trials obtained a score ≤ 3 points, which means that none was assigned a score ≥ 5 points (Table 6).

As for the studies evaluating the effectiveness and safety outcomes resulting from the use of high doses of irinotecan in patients not carrying the UGT1A1*28 allele, six out of the seven studies were in favor of a dose increase (Table 3). Páez D et al.16, who were among those in favor, randomized patients into a group receiving a high dose FOLFIRI regimen (HD-FOLFIRI) (experimental arm) and a group receiving FOLFIRI at the standard dose (control group). The patients in the experimental arm received 300 mg/m2 if they were wild-type and 260 mg/m2 if they harbored the UGTIA1 *1/*28 genotype, as compared with 180 mg/m2 in the control group. The patients’ response was measured in terms of ORR according to RECIST 1.1, PFS and OS criteria. Higher ORRs were obtained in patients on higher doses of irinotecan (67.5% vs 43.6%; p = 0.001), without a significant increase in associated toxicity (severe toxicity: 22.5% vs 20.5%), and without higher levels of PFS or OS.

A study by Lu et al.19, was mainly aimed at evaluating clinical response rate (CRR), PFS and OS, with statistically significant differences being found across all variables (Table 3). Patients were classified into two groups: one with and the other without dose escalation. In the first group, patients with UGT1A1 *1/*1 or *1/*28 genotypes received an initial dose of 180 mg/m2 of irinotecan and patients with an homozygous genotype for the allele *28 received an initial dose of 120 mg/m2. In the dose escalation group, the irinotecan dose was increased by 30 mg/m2 every three cycles. The authors found that the clinical response was higher is patients with dose escalation (69.4 vs 46.4%; p = 0.028), without any statistically significant difference with respect to grade 3/4 adverse events between the two groups (p = 0.189).

These results are in line with those reported by Tsai et al.17, who studied toxicity as well as effectiveness (in terms of ORR) of different irinotecan doses in patients treated with a FOLFIRI-bevacizumab regimen as first-line therapy. An experimental and a control group were established. The experimental group was treated with increasing doses of irinotecan depending on the nature of their genotype (wild types received 260 mg/m2 while heterozygotes received 240 mg/m2). The primary endpoint was evaluation of the PFS, while the secondary endpoints were measurement of the ORR, the disease control ratio (DCR) and OS. The results obtained were as follows: PFS: 14 vs 10 months (HR 0.539 [0.39–0.730], p < 0.001); ORR p < 0.001, DCR: p = 0.007, and OS: 30 vs 22 months (HR 0.693 [0.503–0.955], p < 0.025). The authors concluded that, in patients with mCRC, regardless of their KRAS status, UGT1A1 genotyping may increase the tolerability of higher doses of irinotecan, allowing favorable clinical results without significant increases in toxicity.

Another study to conduct a joint evaluation of increasing the irinotecan dose according to the patients’ UGT1A1 genotype and the safety profile was conducted by Phelip et al.20. This study included patients without previous treatment who presented with potentially resectable liver metastases, treated with six cycles of HD-FOLFIRI+cetuximab. Heterozygous patients (*1/*28) and those with a wild-type genotype (*1/*1) received a dose of irinotecan of 260 mg/m2, while patients with a mutated homozygous genotype (*28/*28) received 220 mg/m2. The primary endpoints were the resectability rate and the safety of the regimen administered. The results showed that high rates of tumor response (82.6% in patients receiving the six cycles) and of liver metastasis resection (80.7%) could be achieved, without intolerable levels of toxicity.

Boisdron-Celle et al.21 performed an evaluation of the efficacy and safety of the FOLFIRI-cetuximab regimen with high doses of irinotecan (*1/*1: 210 mg/m2, with a 20% increase at each cycle up to a maximum of 293 mg/m2; *1/*28: 180 mg/m2, with a 20%/increase at each cycle up to 274 mg/m2; and *28/*28: 126 mg/m2, with a 10% increase at each cycle up to a maximum of 153 mg/m2), without a control group. The primary endpoint of this study was ORR, with 25.8% of patients exhibiting a response, which allowed 43.5% of them to undergo a secondary resection of the metastasis. This dose escalation pattern did not result in significant ORR or PFS differences in the three groups analyzed, despite using a widely different irinotecan dose. Nor were there any differences observed with respect to safety.

The study carried out by Lu et al.22, which included patients with mCRC on first-line treatment with the FOLFIRI-bevacizumab regimen, stratified patients into two groups (group 1: *1/*1, *1/*28; and group 2: *28/*28). In the first group, patients with a wild-type or heterozygous genotype were started on a 180 mg/m2 dose of irinotecan, with dose escalation implemented every three cycles (with increases of 30 mg/m2 up to a maximum of 260 mg/m2 for wild-type genotypes and 240 mg/m2 for heterozygous genotypes). In group 2, patients were started on 120 mg/m2 of irinotecan (maximum dose following a 210 mg/m2 escalation). The primary endpoint was ORR and the DCR, as well as PFS, with a comparison being made between the wild-type/heterozygous genotype group (group 1) and the mutated homozygous genotype group (group 2). The following results were observed: ORR: 76.9% vs 20.0%, p < 0.001; DCR: 93.8% vs 40.0%, p < 0.001. No increase in adverse events was observed.

Hebbar et al.23 analyzed whether a combining chemotherapy strategy with FOLFIRI at high doses in UGT1A1 *1/*1 or *1/*28 genotype patients, together with radiofrequency ablation and surgery, could increase the clearance of liver metastases and survival rates in patients with mCRC. Based on their observation of complete clearance of metastasis in 33.3% of patients without any cases of grade 4 toxicity or death, as well as of an increase in OS, the authors concluded that FOLFIRI at high doses may represent an alternative for patients not amenable to targeted therapy, although the study had to be discontinued because of the small size of the sample.

Finally, and in contrast with other reports, Manfredi et al.24 used high doses of irinotecan (260 mg/m2) in heterozygous patients with a wild-type genotype (*1/*1 and *1/*28), and observed an ORR of 45.0% in the *1/*1 group and of 56.5% in the *1/*28 group, which was similar to the ORR observed in other studies without dose escalation. With regard to toxicity, the authors established the following criteria to discontinue the study: 1) ≥ 7 patients with an objective response (OR); and 2) ≥ 3 patients with severe toxicity. As during the preliminary studies, 26 patients in group 1 (UGT1A1*1) developed grade 4 neutropenia, 6 patients (4 in the *1/*1 group and 2 in the *1/*28 group) presented with grade 4 diarrhea, and 2 patients in the *1/*1 group exhibited febrile neutropenia, a decision was made to discontinue the study, even that a reduction of the dose resulted in an acceptable safety profile.

Several studies evaluated the maximum tolerated dose (MTD) of irinotecan when administered as part of the FOLFIRI regimen as a function of the patients’ UGT1A1 genotype. One of the first studies was carried out in 2010 by Toffoli et al.11 who reported that *1/*1 patients were able to tolerate a MTD of 370 mg/m2 and *1/*28 patients could tolerate 310 mg/m2. These results were subsequently corroborated by Marcuello et al.10, who found a MTD of 390 mg/m2 and 340 mg/m2, respectively. A more recent study, by Kim et al.12, published in 2015, reported MTDs of 330 mg/m2 and 300 mg/m2, respectively. In a study published in 2017, Toffoli et al.14 established the MTD for irinotecan in patients with mCRC treated with first-line FOLFIRI-bevacizumab. MTD was 310 mg/m2 and 260 mg/m2 for patients with the UGT1A1*1/*1 and *1/*28 genotypes, respectively. Another study, authored by Kim et al.15, evaluated the MTD in patients treated with the XELIRI (capecitabine-irinotecan) regimen and found a MTD of 380 mg/m2 in patients with a wild-type or heterozygous genotype, and of 240 mg/m2 for those with mutated homozygous genotypes. Finally, Kim et al.12 observed that the MTD of irinotecan when used as part of the FOLFIRI regimen in patients with an unmutated genotype was 330 mg/m2, for those with heterozygous genotypes of up to 300 mg/m2, and for those with homozygous genotypes of 150 mg/m2.