Pharmacogenomics guided dosing for fluoropyrimidine and irinotecan chemotherapies for patients with cancer (PACIFIC-PGx): study protocol of a multicentre clinical trial

Background

Fluoropyrimidines (5-fluroouracil (5-FU), capecitabine, FP)) are commonly prescribed anti-cancer drugs across many tumour streams such as colorectal, gastric and breast cancers. Although these drugs are generally well tolerated, up to 30% of patients experience severe treatment-related toxicity, and some patients (∼1%) dying consequent to treatment related adverse events [1].

The Clinical Pharmacogenetics Implementation Consortium (CPIC) updated best practice guidelines in 2018 for 5-FU and capecitabine dose optimisation [2]. CPIC recommended that the four common DPYD genetic variants (Table 1) known to affect DPD enzyme activity should be incorporated as part of a DPD screening program [2].

Table 1. DPYD variants tested as part of the trial and activity score.

rsID	Nucleotide change	Allele functional status	Activity score
rs3918290	c.1905 + 1G > A	No function	0
rs55886062	c.1679T > G	No function	0
rs67376798	c.2846A > T	Decreased function	0.5
rs56038477 (rs75017182, rs56276561)	c.1236G > A (c.1129–5923C > G, c.483 + 18G > A)	Decreased function	0.5
rs115232898	c.557A > G^(additional variant tested as part of the trial and commonly observed in the African ancestry as per CPIC)	Decreased function	0.5

Authorities such as Health Canada and The United States Food and Drug Administration (FDA) have added statements to the drug labels for 5-FU and capecitabine that warn against use in patients with known DPD deficiency [3], whilst others advise pre-treatment genomic testing. The European Medicines Agency (EMA) has recommended DPD deficiency genetic testing and United Kingdom government in October 2020 advised all patients should be tested for DPD deficiency before initiation of FP to minimise the risk of these reactions [3]. The Australian Therapeutic Goods Administration (TGA) has not mandated this change [3].

Irinotecan chemotherapy is also associated with severe toxicity especially in patients who are homozygous for the UDP glucuronosyltransferase 1A1 (UGT1A1*28) allele [4,5]. Irinotecan is commonly used in the treatment of gastrointestinal cancers in combination with 5-FU or as a single agent [5]. The most frequently studied UGT1A1 polymorphisms in relation to irinotecan pharmacokinetics/pharmacodynamics are UGT1A1*6 and UGT1A1*28 [5]. The incidence of UGTA1A1*28 genetic variant is relatively high among the Caucasian population and Africans/African Americans. Whereas UGT1A1*6 is more common in the Asian population [5].

In 2005, The FDA recommends consideration of irinotecan dose reduction by one level for patients homozygous for the UGT1A1*28 allele [6]. The Dutch Pharmacogenetics Working Group (DPWG) recommend an initial irinotecan dose reduction of 30% in homozygous patients (*28/*28), with the dose to be increased, guided by the neutrophil count [6]. There is no current guidelines with dose recommendation for patients carrying homozygous UGT1A1*6 allele [5].

International guidelines from the CPIC and DPWG support routine testing and dose adjustment based on DPYD and UGT1A1 genotypes [2,6]. However inconsistent recommendations across regulatory authorities, variable genotype phenotype relationships and limited implementation and feasibility data, impede clinical translation and initiatives to reduce practice variation.

The PACIFIC-PGx study has been designed to assess and evaluate the feasibility in Australia of implementing a pre-treatment PGx Screening Program for patients with cancer commencing on FP or irinotecan-based chemotherapy. This includes a centralised model for consenting, PGx screening for DPYD and UGT1A1*28 genotyping, and follow-up of patients at urban and regional centres. In addition to the feasibility outcomes, the study has a translational research arm facilitating the collection of research blood samples for explorative analyses to identify novel genetic and/or clinical predictors of toxicity for FP and irinotecan. We hypothesise that PGx-guided dose adjustments through the pre-treatment PGx testing coordinated within routine care, will prevent severe FP and irinotecan related toxicity, leading to reduction in hospital admissions, treatment interruptions and discontinuation, improved quality of life (QOL), and reduced health-care costs without impacting on treatment response.

Methods

Study design

A multi-centre, single arm clinical trial to implement and evaluate the feasibility of a PGx Screening Program and the impact of pre-treatment PGx testing and PGx-guided dose adjustments on reducing severe chemotherapy-related adverse events. A single arm design was chosen, as randomisation to no PGx testing was deemed unethical.

The trial will be conducted using the Teletrial Model (TTM) [7], to improve access for regional patients and enable implementation and an evaluation of a centralised hub-spoke model of care. The latter is foreseen as the likely model to allow successful integration into routine practice beyond trial completion. Under the TTM, patients are consented, have DPYD and UGT1A1*28 gene testing, counselling and follow-up for acute toxicities performed by a trained pharmacist at the lead metropolitan site. The trial is for PGx screening and dose recommendations only. It is the responsibility and choice of the treating clinician to implement or not implement dosing recommendations and to manage all aspects of cancer treatment. The study was approved by the Peter MacCallum Cancer Centre Human Research Ethics Committee (HREC/66681/PMCC-2020) and registered in the Australian New Zealand Clinical Trials Registry (12621000251820).

Participants

Eligible patients are aged over 18 years of age and planned to receive FP or irinotecan based chemotherapy for any cancer diagnosis. Patients have the option to provide research samples for future expanded genomic profiling. Patients will be recruited from four Australian sites including a large metropolitan tertiary specialist cancer centre (primary site leading centralised model of care) and three satellite (regional) centres. Patients that have had prior exposure to FP or irinotecan or known to have DPD deficiency or Gilberts’ syndrome are ineligible. Recruitment will be over 12-months with additional 12-months of patient follow-up and final data collection.

Study endpoints

The primary endpoint is operational feasibility as the proportion of patients who are planned for FP/irinotecan-based chemotherapy at enrolment sites, who were enrolled and underwent PGx testing as part of the PACIFIC-PGx trial. Secondary endpoints include; the proportion of patients undergoing treatment with FP or irinotecan therapy at enrolment sites who undergo PGx testing with results available prior to treatment commencement, treatment toxicity, health service utilisation, user acceptance surveys (patient/clinician) and QOL surveys. Exploratory endpoints include; overall response rate (ORR), progression free survival (PFS) and 12-month overall survival (OS).

Toxicity and economic analysis will be reported within this single arm study of PGx screening Program and compared to a historical cohort (without PGx screening). This approach is consistent with that employed in a similar studies conducted in the Netherlands [8–10]. Patients with DPYD*2A variant allele carriers who received a full dose of FP-based chemotherapy will be compared to DPYD*2A variant allele carriers received PGx adjusted dosing as part of this study. In addition, toxicity of DPYD*2A variant allele carriers who received PGx adjusted dosing will be compared to DPYD*2A wild type (WT) patients receiving the standard dose of FP-based chemotherapy. In terms of irinotecan we will compare toxicity data of grade 3 or higher according to Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 and irinotecan dose intensity for patients enrolled in this study who are heterozygous for the UGT1A1*28 allele versus WT.

An economic analysis will be designed as a cost-effectiveness/cost utility analysis (CEA/CUA) of PGx testing to no PGx, assuming superior safety for PGx screening [11]. A historical model with prospective collected data will be developed and scenario’s will be simulated based on Australian cost inputs from the screening cohort and probabilities from meta-analyses of past studies [12]. Refer to Supplementary material 1 for full details on economic analysis.

Study procedures

All patients will undergo PGx screening prior to commencement of FP or irinotecan chemotherapy. Patient enrolment will be triggered by referral from the primary treating oncologist to the lead site pharmacist who will educate and consent patients, coordinate sample collection, and communicate test results and clinical recommendations to the treating clinician and patient. Patients will be followed to cycle 3 day 1 for acute toxicities and related health resource utilisation/economic outcomes, with secondary follow-up to 12 months for disease-response and survival outcomes. Refer to Supplementary material 2 and Supplementary Table 2 for the schedule of study assessments and full details.

Patients will be genotyped for five actionable variants within the DPYD (Table 1) and UGT1A1*28 allele. Dosing recommendations will be based on screening data and CPIC/DPWG guidelines (Table 2). Acute toxicities will assessed to cycle 3 day 1 according to CTCAE version 5.0. Refer to Supplementary material 2 for full details on data collection.

Table 2. DPYD and UGT1A1 genotyping dose recommendations.

DPYD normal metaboliser	No dose adjustments are required.
DPYD intermediate metaboliser	Reduce starting dose by 50% in the first 2 cycles followed by titration of dose based on toxicity.
DPYD poor metaboliser	Avoid use of 5-FU or capecitabine based regimens in this patient.
UGT1A1*1/*1	No dose adjustments are required.
Normal metaboliser
UGT1A1 *1/*28	No dose adjustments are required.
Intermediate metaboliser
UGT1A1 *28/*28	Reduce starting dose by 30% at cycle 1, followed by titration of dose based on neutrophil count.
Poor metaboliser

Statistical analysis plan

Estimated number of patients to be included in the study is 630 participants during 12 months of recruitment. As there is no formal sample size calculation for this feasibility study, it will conclude at 12 months’ time-point and patients recruited during the specified study period will be included in the study. Descriptive statistics will be used to summarise patient demographics, disease characteristics, gene test results, feasibility, health resource utilisation, and program interventions (dose modifications/supportive medications post chemotherapy cycle). Continuous variables will be described as mean, standard deviation, interquartile range, medians, and ranges; qualitative variables will be described as counts and percentages. Maximum toxicity grade per patient of each adverse event will be derived and presented in table format. All adverse events will be described overall and separately by cycle.

The proportion of patients with and without PGx guided dose reduction, and 12 months survival rate will be presented with 95% confidence interval. Refer to Supplementary material 3 for full details.

Surveys and health questionnaires

End-user acceptance surveys have been developed to understand patients and clinicians experiences and acceptance of participation in the PGx Screening Program and to identify improvement opportunities (Supplementary material 4).

A clinician survey planned for national distribution across professional organisation networks, has been developed to ascertain clinicians’ awareness on the current levels of evidence available, and their views on of implementing a program for DPYD and/or UGT1A1 testing. The survey is composed of a series of quantitative questions and one final qualitative question that include multi-item or single-item scale(s) (Supplementary material 4).

All patients enrolled in this study will be invited to complete EORTC QLQ-C30 questionnaire [13] to assess the QOL and identify to explore their perspective on toxicity in different tumour streams.

Patients will be invited to complete EQ-5D-5L version 3.0 (EuroQol 2019) questionnaire which will be used to capture utilities across the study and derive QALYs and will be used to as part of health economic assessement [14]. Refer to Supplementary material 4 for full details.

Refer to schedule of assessment for time points of each survey/health questionnaires (Supplementary Table 2).

Discussion

PACIFIC-PGx trial will be the first study to evaluate the feasibility a PGx Screening Program and the effects of pre-treatment PGx testing and PGx-guided dose adjustments on reducing severe FP and irinotecan adverse events within the Australian healthcare system. A single arm design was chosen, as randomisation to no PGx testing was deemed unethical given evidence from randomised trials of mortality risk in the absence of PGx [1].

This study was designed with the primary and secondary outcomes focussed on operational feasibility in an attempt to generate evidence to overcome perceived barriers for routine implementation of a PGx Screening Program in Australia. Large prospective multicentre studies from the Netherlands [8,15], support the feasibility and cost-effectiveness/cost neutral of prospective DPYD genotyping for patients receiving FP across all tumour types, however it has not been widely adopted [8,15]. Hesitancy may be around the turnaround time from clinician referral to genetic screening until reporting of PGx dose adjustments prior to chemotherapy treatment. Other barriers may include lack of clinician awareness and onsite testing capabilities. Clinician surveys developed as part of this study will explore these barriers in details (Supplementary material 4).

Toxicity is a key secondary endpoint in this study. As previously described, the inability (unethical design) to enrol a comparator arm, required utilisation of a historic control for assessment of treatment-related toxicity with and without PGx screening. Whilst this non-randomised approach will have obvious limitations compared to randomised data, a prior selection of the comparator cohort approximately matched for disease and treatment interventions, aimed to reduce potential bias.

An important feature of the PACIFIC-PGx study is centralised testing and reporting model (facilitated under TTM), geared towards future health service implementation. Supporting a system-wide approach for the implementation of evidence-based change that can lead to improvements in the health care outcomes for all patients with cancer. The centralised testing model allows regional patients to access PGx screening despite receiving cancer care a local centre where volume and geography diminish the viability of local PGx testing, but not the right to access best possible care. This study has the potential to lead to rapid integration of PGx Screening Programs across different Australian hospitals with improvements to current design informed by the patient and clinician feedback, captured within this study.

The implementation of an integrated PGx Screening Program with pre-treatment PGx testing and PGx-guided dose adjustments may promote personalised chemotherapy dosing decisions and establish a new model of care to optimise chemotherapy treatment with potential reduction in severe chemotherapy side effects, hospitalisations, health-care costs, treatment interruptions/discontinuation without affecting the overall response to treatment.

Ethical approval

The study was approved by the Peter MacCallum Cancer Centre Human Research Ethics Committee (HREC/66681/PMCC-2020) and registered in the Australian New Zealand Clinical Trials Registry (12621000251820). All participants must sign an Informed Consent Form prior to study enrolment.

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Additional information

Funding

This work was supported by the International Society of Oncology Pharmacy Practitioners (ISOPP) under ISOPP Research Grant Award.