Aim

The PIONEER trial aims to determine the feasibility and acceptability of conducting a phase III RCT comparing the impact on HRQOL of surgery versus no-surgery as part of multimodality treatment, in patients with stage III-N2 NSCLC and their carers. Two treatment groups will be compared: (1) multimodality treatment including surgery or (2) multimodality treatment without surgery. This trial, therefore, is focused on those patients with stage III-N2 in whom surgery is technically feasible (termed ‘potentially resectable’). Patients with stage III-N2 in whom the cancer is technically ‘unresectable’ are not eligible for the trial and are not the patient cohort for whom the decision between surgical and non-surgical treatments exist. The following research questions will be addressed: (A) what is the total number and proportion of patients with ‘potentially resectable’ N2-NSCLC, and fit for multimodal therapy, in an unselected population of lung cancer patients? (B) Are patients (and carers) willing to consent to the study and once randomised do they undergo allocated treatment, and if not, why not? (C) Are clinicians willing to recruit eligible patients to the trial, what are clinician-related barriers/facilitators to recruitment? (D) What is the optimum strategy to facilitate trial recruitment? (E) What is the optimal primary outcome assessment and design for a full phase III trial?

Inclusion and exclusion criteria

Participants are eligible to be included in the study only if all of the following criteria apply: Patients with potentially resectable T1-4 N2 M0 NSCLC who have received a multidisciplinary team (MDT) recommendation for multimodality treatment; MDT consensus that the patient has adequate physiological reserve for multimodality treatment and either treatment arm is both technically and clinically appropriate; patient over the age of 18 years. Participants are excluded from the study if any of the following criteria apply: Patients unable to provide informed consent; Patients where the recommendation is for treatment other than those multimodality treatments listed; patients in whom a specific multimodality treatment regimen is recommended over another by the MDT for individualised reasons.

Definition of ‘potentially resectable stage III-N2 NSCLC’

A standardised definition of potentially resectable is provided to all sites (box 1). This definition has been proposed and validated with a high level of agreement in a National survey of practice in stage III lung cancer across the UK.17 18 Patients with stage III-N2 NSCLC will be deemed ‘potentially resectable’ and eligible for the trial if they fulfil all of the criteria set out in the standardised definition.

Treatment arms

Both treatment arms are offered as standard practice. There is no experimental and control arm in this study as no one treatment regimen has been found to be clearly superior to another and therefore multiple treatment options exist as standard of care. This is reflected by international lung cancer guidelines which confirm both surgical and non-surgical multimodality treatment regimens as appropriate strategies for stage III-N2 NSCLC.19 20

Surgical multimodality treatment arm

The surgical arm can include any form of multimodality treatment that includes surgical resection. This includes induction chemotherapy followed by surgery, induction chemoradiotherapy followed by surgery and surgery followed by adjuvant chemotherapy. The choice of surgical multimodality regimen is left to the discretion of the local treating MDT. In the UK, surgery followed by adjuvant chemotherapy is the most common form of multimodality surgical treatment performed and is likely to form a significant component of the surgical arm in this trial.2 However, the recent NICE guideline recommendation for induction chemoradiotherapy followed by surgery may lead to a higher contribution of trimodality treatment to the surgical arm.

Non-surgical multimodality treatment arm

The non-surgical arm can include any form of multimodality treatment that does not involve surgery. This includes sequential chemoradiotherapy and concurrent chemoradiotherapy. Within this treatment arm, there is an important discussion point in regards to the use of durvalumab following concurrent chemoradiotherapy. There is considerable debate among the lung cancer community about how the PACIFIC trial results impact treatment decisions for patients with stage III NSCLC considered suitable for surgery. The PACIFIC trial recruited patients with ‘unresectable’ stage III NSCLC but the protocol provided no standardised definition of ‘unresectable’ and this was left at the discretion of the local MDT. Some MDTs consider any N2/mediastinal nodal metastases as unresectable whereas other MDTs will recommend surgery in N2 NSCLC. In the PACIFIC trial, no data were provided to help clinicians understand the study population in terms of the extent of nodal disease and tumour burden. Everyday practice in the UK has reflected this uncertainty with oncologists considering the use of durvalumab in patients with stage III NSCLC that have completed concurrent chemoradiotherapy regardless of whether the disease could have been considered technically suitable for surgery at the outset, that is, the ‘unresected’ rather than the ‘unresectable’. The PIONEER trial provides a standardised definition of ‘potentially resectable’ N2 NSCLC that is broad and will likely include some patients that are similar to those treated in the PACIFIC trial. Therefore, if a patient within the PIONEER trial is randomised to the non-surgical arm and they complete concurrent chemoradiotherapy, the local oncologist will consider their eligibility for durvalumab. The local oncologist may wish to rediscuss that patient within their local MDT (perhaps with the PD-L1 status) to clarify eligibility for durvalumab towards the end of the chemoradiotherapy treatment.

Recruitment

The eight recruiting sites will be in Manchester (Wythenshawe and Salford), Leeds, Sheffield, University College London Hospitals (UCLH), Birmingham, Glasgow and Nottingham. There are approximately 2500 new cases of N2 NSCLC every year in the UK. Our conservative estimate for recruitment is one patient every second month from each site. Participants will be eligible for the study if they have ‘potentially resectable’ Stage III-N2 lung cancer and have received an MDT recommendation for multimodality treatment. We aim to recruit 66 patients over a 20-month staggered recruitment period. This sample size was chosen to ensure that we can estimate key parameters with sufficient precision based on recommendations by Lancaster et al.21 Recruitment will be led by the respiratory physicians at eight recruiting centres (all high volume lung cancer MDTs). The consenting process will be conducted by respiratory physicians and supported by lung cancer nurse specialists. Our ambition is that the majority of recruitment is undertaken by respiratory physicians as impartial advocates for the patient and to minimise treatment related bias. However, recruitment within surgical and oncology clinics will be allowed to provide flexibility and maximise recruitment opportunities. All potential study participants will be provided with a participant information sheet (PIS) and invited to voluntarily consent to participate in the study. They will also be asked if they have a caregiver who may also want to take part in the study. Due to the anticipated challenges with recruitment we will proactively monitor, evaluate and react to quickly rectify emerging issues. We will apply the principles of the Quintet Recruitment Intervention (QRI) to maximise recruitment as much as possible and inform the recruitment strategy for the follow-on trial.22 Key steps within the QRI process include: (1) understanding the recruitment process; (2) documenting barriers to recruitment and trying to understand them; (3) assessing how the RCT is being delivered by recruiters; (4) presenting evidence of recruitment difficulties to the trial management group and developing strategies to overcome these problems and (5) implementing the plan of improvement. This will include audio recording of recruitment appointments, with real-time data analysis and feedback to the clinical teams to proactively inform recruitment practices including ongoing training. We will conduct trial recruitment workshop sessions with recruiting clinicians across all sites and provide refresher sessions as required based on the results of the QRI process.22 Figure 1 presents the flow of patients through the study.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

CONSORT flow diagram. CONSORT, Consolidated Standards of Reporting Trials; MDT, multidisciplinary team. Hospital Anxiety and Depression Scale (HADS). EuroQOL 5 Dimensions (EQ5D)

Randomisation

Eligible patients and caregivers will provide written informed consent and complete baseline assessments. The recruiting clinician will contact the research team at The Christie National Health Service Foundation Trust who will manage the randomisation process using an online randomisation system. The person responsible for randomisation will have no involvement in the recruiting process. Participants will be randomised on a 1:1 basis to receive either surgery or no surgery as part of multimodality treatment. Once a patient has been randomised, both the clinical team and the patient will be made aware of the treatment allocation; given the nature of the intervention there will be no masking of patients, carers or clinical staff

Assessments and data collection

Various outcomes and success metrics (table 1) will be collected via questionnaires returned by post or completed over the telephone by a member of the clinical team. The outcomes and success metrics collected will determine if it is feasible to do a follow-on randomised controlled trial in the future. A summary of required data, assessment tools and collection time points is provided in table 2.

Feasibility outcomes

• Proportion of patients with N2 disease that are potentially resectable as defined by the lung cancer MDT.

• Proportion of patients with potentially resectable N2 NSCLC that have adequate physiological reserve for multimodality treatment.

• Number and proportion of eligible patients recruited—assessment of screening logs to determine numbers eligible versus not randomised and reasons (such as clinician reluctance, patient/family decline, other).

• Proportion of people lost to follow-up due to withdrawal, death or failure to return questionnaires (ie, assessment attrition) by 3 and 6 months by arm.

• Percentage of participants with complete data at end of trial (6 months), by arm.

• Proportion of randomised patients who receive allocated treatment, by arm.

• Adherence to the allocated treatment arm will be monitored closely (any patients that crossover to receive treatment in the other arm will remain in the trial for data collection but will be recorded as failing to complete the allocated treatment).

• Qualitative data recording recruitment consultations and interviews with participants and non-participants and their carers.

Clinical data

Clinical data will be collected directly from the patients’ hospital records. Data collected will include: age; gender; performance status at recruitment; clinical frailty score; co-morbidities (aligned to the Charlson Comorbidity Index); diagnosis of interstitial lung disease; body mass index (BMI), smoking status; forced expiratory volume (% predicted); diffusing capacity of the lungs for carbon monoxide % predicted; blood results at recruitment; NSCLC histology; clinical (and pathological where available) TNM8 staging; location, size and standard uptake value (SUV) of primary tumour; N2 nodes involved; N2 sub classification (single station vs multistation); Highest stage N2 nodal size (short axis); Highest stage N2 nodal SUVmax and mediastinal blood pool SUVmax. When performed the following additional data will be collected, accepting this is not mandated: 6 min walk test (metres) or incremental shuttle walk (metres); Vo2max (mls/kg/min) V02max%predicted; left and right ventricular function on transthoracic echocardiogram (% ejection fraction and/or categorised as normal, mild impairment, moderate impairment and severe impairment).

Participant completed data

Patient and caregiver reported outcome measures are collected at baseline 6, 9 and 12 weeks and 6 months. Unless the participant has a coinciding clinic appointment, questionnaires will be sent by post. Reminder phone calls will be made if questionnaires are not returned. Real-time completion of questionnaires via a telephone call with research staff may also be considered if this is the preference of the patient. Prior to contacting the participant by telephone or post, the participants’ survival status will be checked with the clinical team. Patients and caregivers will also complete diaries describing their experiences in more detail.

Qualitative interviews

Qualitative interviews will be conducted at various stages of the follow-up process with 12–16 patient and caregiver dyads. Participants will be asked at the time of consenting to the study if they are happy to be approached to participate in an interview. The interview will explore how the treatment has impacted on the daily lives of patient and caregivers from a practical and psychological perspective, their thoughts on the randomisation process and the treatment options will also be explored. Interviews will last approximately 60 min; they will be audiorecorded and transcribed verbatim. Interviews will be conducted by a member of The Christie Research team. We will also interview 8–10 patients who declined to take part in the study or opted not to receive their allocated treatment and explore their decision-making processes. Interviews will be conducted face to face or over the telephone. A topic guide will be developed and interviews audiorecorded with consent and transcribed verbatim.

Data analysis

Given that this is a feasibility study, we will not conduct hypothesis testing to determine if either treatment option has a greater impact on HRQOL. Instead, analyses will be descriptive, focusing on calculating confidence intervals for key trial parameters.23 For example: (1) estimating eligible participant numbers, recruitment, fidelity and attrition rates, rate of missing data; (2) estimating the SD for HRQOL and likely trajectory over time, in order to inform/refine a sample size calculation for a future phase III trial and (3) determining whether the study design was acceptable to participants and their caregivers. A feature of the study data will be that some participants will have their HRQOL outcomes censored due to death. The study will allow us to estimate the frequency of this occurring, and to develop an analytical strategy for a phase III trial. We plan to use joint modelling techniques (joineR package in R24), which allows for the joint analysis of survival data and endogenous repeated measures. Our analysis will focus on estimating variability, correlation between repeated measurements from the same individual and modelling the trajectory of HRQOL over time. We will adjust for baseline functional status. We will use the data from this study to design a detailed analysis plan for a confirmatory study. Qualitative data generated from the audiorecorded interviews with patient–carer dyads and clinicians will be analysed using framework analysis.25 Audiorecordings of recruitment consultations will be analysed using content analysis26 to highlight specific issues and enable immediate feedback to the trial management group.

Patient and public involvement

Patients have played a central role in the design and development of this study. Our early work with the Wythenshawe/Christie lung cancer Patient and Public Invovlement and Engagement (PPIE) group and North Trent Consumer Research Panel clearly supported the need for quality of life and burden to be included in treatment decision making. The issue of recruitment with previous trials comparing surgery with radiotherapy has been fully explored with our patient groups. The randomisation process was a concern for some patients so it will be clearly specified in the patient information sheet and at the time of consent that the trial has been suggested to them because there was no clear indication which treatment would have the best outcome for them. Given the patients concern, it was also decided that a clinician would lead the recruitment consultation.

Anticipated risks and mitigation

As described earlier, optimising recruitment will be addressed using the QRI methodology. Other risks include that the treatment paradigm in stage III-N2 NSCLC may change during the course of this feasibility trial (eg, immunotherapy in surgical treatment pathways). However, the research question will remain relevant regarding the impact on HRQOL from surgical versus non-surgical based multimodality treatment regimens. We have designed this feasibility study in a pragmatic way which allows different types of regimens within each arm that reflect changing standards of care. If we can prove this trial is feasible to recruit to then for a phase III trial we will need to clearly define the standard of care in both the surgical and non-surgical arms that reflects the evidence base and guidelines at that moment. Finally, data completeness will be important particularly in the first 9 weeks where multiple different HRQOL questionnaires are needed. The ability to complete this information via telephone and the use of reminder telephone calls should help to mitigate this risk.