Assessment of ibrutinib plus rituximab in front-line CLL (FLAIR trial): study protocol for a phase III randomised controlled trial

The trial aims to compare the effect on progression-free survival (PFS) of ibrutinib plus rituximab (IR) with that of FCR in young/fit patients with previously untreated CLL.

The primary objective of the trial is to assess whether IR is superior to FCR in terms of PFS, as defined by IWCLL criteria [6]. Secondary objectives include:

The trial is a phase III, multicentre, randomised, controlled, open, parallel-group trial in patients with previously untreated CLL. A total of 754 participants will be randomised on a 1:1 basis to receive therapy with either FCR or IR. Participants randomised to FCR will receive a maximum of six treatment cycles with each cycle being repeated every 28 days. Participants randomised to receive IR will receive six cycles of rituximab with each cycle being received every 28 days. Ibrutinib will be taken daily for 6 years or until MRD negativity has been recorded over a period of 6 months or until disease progression, whichever occurs first. Due to the different treatment modalities it is not possible to blind patients or clinicians to treatment allocation.

The main inclusion criteria are: being between 18 and 75 years old; B-CLL with a characteristic immunophenotype (including small lymphocytic lymphoma); Binet’s stages B, C or progressive stage A; require therapy by IWCLL criteria; able to provide written informed consent; considered fit for treatment with FCR as determined by the treating clinician; and ≤ 20% p53 deletion determined by fluorescent in situ hybridisation (FISH) (full inclusion and exclusion criteria listed in Table 1).

The trial is powered for the primary endpoint of PFS. The German CLL8 [7] trial showed a median PFS rate of 4.5 years in the FCR arm. To assess a superiority hazard ratio of 0.75 (that is, an improvement in median PFS to 6 years) with an overall 5% level of significance and 80% power, assuming a 4-year recruitment and a 4-year follow-up period, 355 participants are required per arm (710 participants overall). Accounting for a 5% dropout rate, which is in line with previous Leeds Institute of Clinical Trials Research (LICTR)-led trials in similar populations, 748 participants are required in order to observe 379 events.

A formal interim analysis on PFS will be carried out and reported to the Data Monitoring and Ethics Committee (DMEC) when half the required number of events have been observed. This is in order to allow any large differences between the randomisation arms to be reported early.

The O’Brien and Fleming alpha-spending function [8] will be used to adjust for multiple testing in order to conserve the overall type I error, which recommends that the interim results are compared to a p value of 0.005, and the final results are then compared to a p value of 0.048. In order to account for the interim analysis, the O’Brien and Fleming method recommends increasing the maximum required sample size and number of events by a factor of 1.008. Therefore, 754 participants will be recruited to the FLAIR trial in order to observe 382 events (191 for the interim analysis).

In order to recruit 754 participants over a 4-year period, the recruitment target is seven patients per month for the first 6 months whilst centres are opening to recruitment and 17 patients per month thereafter.

Participants will be recruited from multiple research centres from around the UK. Research centres will be required to have obtained ethical and management approvals and undertake a site initiation meeting with LICTR prior to the start of recruitment into the trial.

Potential participants will be identified by the clinical team at participating centres based on their diagnosis of CLL, and will be provided with verbal and written details about the trial. Informed consent will be collected by principal investigators at the participating hospitals, after a recommended minimum of 24 h to read and digest the information provided.

Following consent, registration and confirmation of eligibility (Table 1), participants will be randomised into the trial by an authorised member of staff at the trial research site. Randomisation will be performed centrally using LICTR automated 24-h telephone randomisation system.

Participants will be randomised on a 1:1 basis to receive either FCR or IR and will be allocated a trial number. A computer-generated minimisation programme that incorporates a random element will be used to ensure that trial arms are well-balanced for the following participant characteristics: Binet’s stage (B, C or progressive stage A), age (≤65 years, > 65 years), gender (male, female), and randomising centre.

Data will be collected on paper Case Report Forms and entered into a validated trial database by LICTR. A validation check programme will be incorporated into the trial database to verify the data, and discrepancy reports will be generated for resolution by the investigator. Priority validations will be incorporated into the validation programme to ensure that any discrepancies related to participant rights, or safety, are expedited to sites for resolution. Data will be monitored for quality and completeness by LICTR. Missing data will be chased until received, confirmed as not available or the trial is at analysis. LICTR/the sponsor will reserve the right to intermittently conduct source data verification exercises on a sample of participants, which will be carried out by staff from LICTR/the sponsor. Source data verification will involve direct access to participant notes at the participating hospital sites and the central collection of copies of Consent Forms and other relevant investigation reports. Data will be held on a secure server at the University of Leeds and paper Case Report Forms stored in a locked unit, both will be accessible only to authorised trial staff.

QoL and participant-completed health economic data will be collected for patients who have given consent, and will be completed by the participant.

Baseline assessments to be performed within 4 weeks of the start of treatment include: complete medical history, complete physical examination, WHO performance status, local haematology and biochemistry, serology for hepatitis B and C, and presence and level of P53 deletion (within 12 months prior to randomisation).

Participants will also undergo a chest X-ray if required, a computed tomography (CT) scan performed within 8 weeks prior to start of treatment and an electrocardiogram (ECG). Assessment of disease and a pregnancy test (if the participant is a woman of child-bearing potential) must also be performed within 2 weeks prior to start of treatment. Blood and bone marrow samples will be required and sent to the Haematological Malignancy Diagnostic Service (HMDS) for flow cytometry (bone marrow samples are only required if participants have a lymphocyte count of below 10 × 10⁹/L). For those who have given consent; blood, bone marrow and saliva samples will be sent to the UK CLL Trials Biobank (see ‘Substudies’ section).

For those who have given consent to QoL questionnaires and health economic assessments, baseline assessments must be completed prior to the participant becoming aware of their randomisation allocation.

Following randomisation to receive FCR or IR; those participants who receive FCR will receive 24/mg/m²/day of fludarabine and 150 mg/m²/day of cyclophosphamide orally on days 1 to 5, in addition to 375 mg/m² on day 1 of cycle 1 and 500 mg/m² on day 1 of cycles 2 to 6 of intravenous rituximab. Cycles of FCR are repeated every 28 days for a total of six cycles.

Participants randomised to IR will receive 375 mg/m² on day 1 of cycle 1 and 500 mg/m² on day 1 of cycles 2 to 6 of intravenous rituximab, and 420 mg of orally administered ibrutinib starting a day prior to the first dose of rituximab, on day 0 of cycle 1. Cycles of rituximab are repeated every 28 days for a total of six cycles, and ibrutinib will be given daily for up to 6 years, until disease progression, or unacceptable toxicity. The duration of ibrutinib therapy in responding patients will be defined by the eradication of MRD. For participants receiving ibrutinib, MRD will be assessed every 6 months in the peripheral blood, and if MRD negative, repeated after 3 months. If the repeat peripheral blood is MRD negative then after a further 3 months the peripheral blood and bone marrow will be assessed for MRD. If both are MRD negative then the patient will receive the same amount of treatment again as it took for them to become MRD negative, then they will stop taking ibrutinib. Participants would then restart ibrutinib if they become MRD positive, MRD positivity confirmed over 6 months, if MRD positivity is confirmed within 6 years post randomisation.

Participants will be assessed for their suitability for treatment within 1 week prior to day 1 of each cycle of FCR or R (for participants randomised to IR), and 3 monthly for the duration of treatment for participants randomised to IR, according to standard clinical practice. For participants randomised to FCR, participants should be questioned regarding nausea and vomiting or diarrhoea occurring with the prior cycle of therapy and if this is present then the fludarabine and cyclophosphamide should be given via the intravenous route from the next cycle onwards due to concerns over drug absorption. Intravenously administered fludarabine (25 mg/m²/day on days 1 to 3) and cyclophosphamide (250 mg/m²/day on days 1 to 3) regimens are recommended if the orally administered regimen is not tolerated. Patients receiving FCR will receive prophylaxis against Pneumocystis jirovecii pneumonia (co-trimoxazole or equivalent) and against herpes virus reactivation (acyclovir). G-CSF (as per standard dosing) for days 7 to 13 is recommended for all subsequent cycles of rituximab for participants who have had to have a previous dose delay due to neutropenia.

Participants can be withdrawn from treatment at the discretion of the treating clinician, and participants are free to withdraw consent from treatment or further follow-up at any time during the trial, without giving a reason, and without future treatment and care being affected in any way.

Participants will be followed up as described in the trial flow diagram (Fig. 1). Participants will be followed up at 6 monthly intervals from 12 months post randomisation until 7 years post randomisation, withdrawal, disease progression requiring treatment, or death, whichever is earliest. Following 6-monthly follow-up, participants will be followed up annually until death or withdrawal, whichever is earliest.

Participants will be evaluated for response to treatment and assessment of MRD via bone marrow aspirate, according to the IWCLL response criteria, 3 months after the end of FCR or R (for participants randomised to IR) and again at the end of ibrutinib for participants randomised to IR. MRD will be assessed throughout the trial according to Fig. 2. A populated Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) figure is also provided in Fig. 3.

QoL and health economics questionnaires will be completed at the end of FCR or R, and then 6 monthly from 12 months post randomisation for 7 years or until disease progression, whichever is earliest.

Safety and toxicity data will be assessed continuously throughout the duration of the trial, as graded by the Common Terminology Criteria for Adverse Events (CTCAE) (v4.03) [9]. Events will be categorised as: suspected unexpected serious adverse reactions (SUSARs); serious adverse reactions SARs); serious adverse events (SAEs); adverse events of special interest (AEoSIs); and adverse events (AEs). AEoSIs are defined as: a major haemorrhage (resulting in: intraocular bleeding causing loss of vision; the need for a transfusion of two or more units of red cells or an equivalent amount of whole blood; hospitalisation; or the prolongation of hospitalisation); or an intracranial haemorrhage. Participants experiencing SAEs and AEs will be assessed from randomisation until 30 days after their last dose of trial treatment. Participants experiencing SARs, SUSARs and treatment-related deaths will be assessed from randomisation throughout the duration of the trial.

Statistical analysis is the responsibility of LICTR statisticians, and a final statistical analysis plan will be written and signed off before any analysis takes place. Health economic analyses will be carried out by a designated health economist at the University of Leeds, and a separate analysis plan will be written and signed off before any analysis takes place.

All analyses will be conducted on the intention-to-treat (ITT) population, where participants will be included according to the treatment that they were randomised to. A per-protocol (PP) analysis, where participants will be included according to the treatment they received, will be considered for the primary endpoint if there are a considerable number of protocol violators. The safety population will consist of all participants who receive at least one dose of treatment, where participants will be included according to the treatment that they received.

The primary endpoint of PFS will be assessed once 382 events have been observed overall (i.e. 382 participants have progressed/died). A formal interim analysis will be carried out when half the number of events (191 events) have been observed overall. Participants not having progressed or died at the time of the analyses will be censored at the last date they were known to be alive and progression-free. The analysis of primacy will use a Cox proportional hazards model [10] to compare the trial arms, adjusting for the minimisation factors excluding centre, to calculate the hazard ratio and associated 95% confidence intervals (CIs). The p value will be compared to 0.048 to account for the formal interim analysis. In addition, median survival estimates and Kaplan-Meier curves [11] will be produced by trial arm.

Overall survival will be assessed at the same time as the primary endpoint using a Cox proportional hazards model to compare the trial arms, adjusting for the minimisation factors excluding centre. Kaplan-Meier curves, 95% CIs and median survival estimates will be produced by trial arm and overall.

The proportion of participants with undetectable MRD (i.e. MRD negative) will be summarised by trial arm and overall at each time point that MRD is assessed. A binary logistic regression model will be used to compare trial arms in terms of the proportion of participants who become MRD negative at any stage during the trial, adjusting for the minimisation factors excluding centre. Time to participants achieving MRD negativity will also be assessed in the IR arm alone.

Binary logistic regression models will be used to compare trial arms in terms of the proportion of participants who achieve a Complete Response (CR or CRi) and who achieve an Overall Response (at least a PR) at any stage during the trial, adjusting for the minimisation factors excluding centre.

Occurrence and frequency of SUSARs, SARs, SAEs, AEoSIs, AEs and treatment-related deaths will summarised by trial arm and overall. In addition, all events will be summarised by maximum CTCAE grade reported and seriousness criteria; and all serious events will be summarised by MedDRA System Organ Class, by trial arm and overall.

Mean quality of life (QoL) scores and 95% CIs, adjusted for the baseline score, will be calculated for all domains of the European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30 and CLL-specific module EORTC QLQ-CLL16 for each trial arm and at each assessment time point and overall. Wilcoxon’s rank sum tests will be used to compare the change in score for the trial arms from baseline to each time point.

A cost-effectiveness analysis, incorporating the development of a decision analytic cost effectiveness Markov/semi-Markov state model, using quality-adjusted life years (QALYs) estimated from utility weights and health state scores, will be carried out using the euroQol five dimensions (EQ-5D) and 12-item Short Form Health Survey (SF12) collected at baseline, end of FCR/R and 6 monthly from 12 months post randomisation.

Pre-specified exploratory analyses will be conducted for subgroups of interest as appropriate, but will be interpreted with caution and treated as hypothesis-generating.

An independent DMEC will review the safety and ethics of the trial whilst participants are receiving trial treatment; reviewing unblinded safety updates at 3-monthly intervals, and more detailed unblinded reports annually.

The DMEC, in light of the interim data, and of any advice or evidence they wish to request, will advise the TSC if there are any concerns or reasons that the trial should not continue, or if they believe that the trial results should be shared further.

The trial was developed by the FLAIR Trial Management Group (TMG), with the support of the UKCLL/NCRI CLL Clinical Trials Subgroup. The trial is sponsored by the University of Leeds ((R&D Office, 34 Hyde Terrace, Leeds, LS2 9LN); sponsor reference: HM13/10747)), run and co-ordinated by LICTR, and is registered (ISRCTN01844152, EudraCT Number 2013-001944-76). The trial will be conducted in accordance with the principles of Good Clinical Practice (GCP) in clinical trials, as applicable under UK regulations, the NHS Research Governance Framework and through adherence to LICTR standard operating procedures (SOPs). LICTR and the sponsor have systems in place to ensure that serious breaches of GCP of the trial protocol are identified and reported. A sample of data will be assessed by on-site source data verification by LICTR and this is not independent from the sponsor. The University of Leeds will be liable for negligent harm caused by the design of the trial. No additional compensation for clinical negligence will be provided for trial participants over that which is available to NHS patients. Ibrutinib was only provided by the sponsor for the duration of protocol treatment and subsequent treatment will be as per standard care.

A core Internal Project Team (IPT), a TMG, a TSC, and a DMEC will be established. The independent DMEC will review the safety and ethics of the trial, as overseen by the independent TSC. The TSC will monitor trial progress and the overall direction. The results of the study will be published in peer-reviewed publications and will be presented at relevant national and international conferences. There are no plans to use professional writers for presenting the outcomes of this trial. Any protocol changes are disseminated by LICTR to the relevant parties. A Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) Checklist has been prepared for this manuscript (Additional file 1).

The trial opened to recruitment in August 2014 using protocol version 1.0 (6 March 2014).

The protocol was amended to version 2 on 4 September 2014 to include the additions of long-term orally administered anticoagulant use and cardiac events requiring dual antiplatelet therapy to the exclusion criteria on safety grounds. This change was implemented before any participants were randomised.

The protocol was amended to version 3.0 on 29 September 2016 to modify the MRD stopping criteria based on updated external MRD data. A criteria for restarting treatment based on MRD relapse was also added at this point. Secondary endpoints relating to pattern of MRD relapse, retreatment and MRD response over time in participants randomised to IR were also added. These changes were implemented before any participants had reached either the original or the amended MRD negative stopping criteria.

Both amendments were reviewed and approved by the sponsor, the REC and MHRA.

Trial participants will also be invited to take part in the UK CLL Trials Biobank at the University of Liverpool. Participants are approached at baseline with a separate consent form (REC reference 14/NW/1014) and biological samples are sent to the University of Liverpool.

The FLAIR trial opened to recruitment in August 2014 using protocol version 1.0 (6 March 2014) and is currently recruiting above target. As of 24 February 2017, 537 participants have been randomised. Recruitment is due to close in August 2018.