Lobectomy-first versus lymphadenectomy-first on long-term survival for operable non-small cell lung cancer: protocol of LOFTY trial

Abstract

Background

The standard surgical treatment for early-stage non-small cell lung cancer (NSCLC) involves anatomical lobectomy and mediastinal lymph node dissection (MLND). According to the principle of “from distal to proximal” during tumour resection, MLND should ideally be performed before lobectomy to fit the criteria of surgical oncology. However, no prospective randomised controlled trial (RCT) has compared MLND-first versus lobectomy-first approaches in early-stage NSCLC. This provides the rationale for this multicentre RCT.

Methods

This multicentre, prospective, open-label, RCT investigates the superiority of MLND-first over lobectomy-first in terms of long-term survival in patients with clinical stage T1-2N0-1M0 (stages I–II) NSCLC (LOFTY study). We plan to enrol 620 patients from multiple institutions. Stratified block randomisation will be performed in this trial. The primary endpoint is the 5-year disease-free survival rate, and the secondary endpoints comprise the 5-year overall survival rate and short-term outcomes, such as conversion rate, perioperative complication rate, number of resected mediastinal lymph nodes, operative death rate, postoperative hospital stays, R0 rate, operative time, chest tube duration, number of circulating tumour cells before and after operation, and the total cost of hospitalisation. The primary endpoint will be assessed by a central committee of blinded expert radiologists. Patient recruitment is currently ongoing.

Discussion

This multicentre, prospective, open-label RCT investigates whether MLND-first is superior to lobectomy-first in terms of long-term outcomes in early-stage NSCLC patients who are suitable for lobectomies. The results of this trial will provide high-level evidence to clarify the standard surgical sequence for early-stage operable NSCLC.

Trial registration

Guangdong Association Study of Thoracic Oncology GASTO-10129. Registered on 26 March. 2022. Chinese Clinical Trials Registry ChiCTR2300068586. Registered on 24 February 2023. ClinicalTrial.gov NCT06577792. Registered on 29 August 2024.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13063-025-09285-9.

Background

Anatomical lobectomy combined with mediastinal lymph node dissection (MLND) is still the standard surgical treatment for early-stage solid non-small cell lung cancer (NSCLC) [1–3]. In clinical practice, most thoracic surgeons prefer to perform lobectomy first, because the surgical field becomes more capacious after lobectomy, which might make MLND more convenient [4]. However, some principles of surgical oncology are emphasised in oncology textbooks, including the key requirement to dissect the distant lymph nodes (LNs) first, followed by the proximal LNs [5]. The theoretical assumption is that removing distant LNs first may block the lymphatic pathway of tumour metastasis and therefore reduce the potential spread of tumour cells caused by intraoperative manipulation. Therefore, MLND should be performed before lobectomy.

A previous randomised study comparing vein-first versus artery-first ligation in lobectomy showed differences in circulating tumour cell release and even survival, highlighting that the sequence of intraoperative steps can influence oncological outcomes [6]. Unfortunately, to the best of our knowledge, no prospective randomised controlled trial (RCT) or retrospective study has confirmed the clinical significance of the surgical sequence of MLND and lobectomy in patients with early-stage operable NSCLC. Accordingly, there is no evidence-based recommendation in current clinical guidelines to favour either the lobectomy-first or the lymphadenectomy-first sequence.

Therefore, we have initiated a prospective, multicentre, open-label RCT using 5-year disease-free survival (DFS) as the primary endpoint to investigate the superiority of MLND-first over lobectomy-first in clinical stages I–II NSCLC patients.

Methods/design

Study design

This is a prospective, multicentre, randomised, open-label clinical trial to compare the long-term outcomes of performing MLND first with those of performing lobectomy first for patients with clinical stage T1-2N0-1M0 (stages I–II) NSCLC. The experimental group (MLND-first group) is expected to improve long-term survival compared with the control group (lobectomy-first group). The primary hypothesis is that performing MLND first avoids potential tumour metastasis through the LNs. Figure 1 depicts the flowchart of the LOFTY study.

Fig. 1.

Flowchart of the LOFTY study. This diagram outlines the sequence of steps involved in the LOFTY trial, including enrolment, randomisation, treatment allocation, and follow-up assessments. ECOG, Eastern Cooperative Oncology Group; GGO, ground-glass opacity; DFS, disease-free survival; OS, overall survival

Study setting

This trial is scheduled to encompass 26 distinct hospitals across China, all of which have obtained ethical approval. These 26 hospitals are well known for their high patient volumes and are located across various regions of China, including South, East, Northwest, North, and Southwest China. The selection of such a diverse range of sites guarantees wide demographic coverage and bolsters the applicability of the findings to a broader population. Participating sites are listed in the ClinicalTrials.gov registration (identifier: NCT06577792).

Endpoints

The primary endpoint is 5-year DFS. DFS is defined as the time from randomisation to relapse or death from any cause.

Secondary endpoints are grouped as follows: (1) Survival outcomes: 5-year overall survival (OS), 3-year DFS, and 3-year OS; (2) intraoperative outcomes: blood loss, conversion rate, and operative time; (3) postoperative outcomes: perioperative complications, operative death rate, chest tube duration, pain score, analgesic (Grade III) dose, and postoperative length of hospital stay; (4) pathological outcomes: R0 rate and the number of dissected mediastinal LNs; and (5) economic outcomes: total cost of hospitalisation. Additional exploratory endpoints include pre- and postoperative circulating tumour cell (CTC) counts.

OS is defined as the time from randomisation to death from any cause. Blood loss is defined as the blood loss during surgery. The conversion rate is defined as the proportion of patients who undergo open surgery among all patients undergoing surgery in each group. Operative time is defined as the total time from skin incision to suturing, and the time of each step will be recorded simultaneously during each surgery. Perioperative complications are defined as complications that occurred during surgery or within 30 days after surgery in all patients receiving operations, which will be recorded and classified using the Clavien-Dindo classification system [7] daily during hospitalisation and surveillance after discharge. The operative death rate is defined as the proportion of patients who die within 30/90 days after surgery among all patients undergoing surgery in each group. The chest tube duration is defined as the time between the date of surgery and chest tube removal. The pain score will be evaluated using a Numerical Rating Scale daily during hospitalisation after surgery and during surveillance after discharge. The acesodyne (Grade III) dose is defined as the total amount of morphine used by the patient during the postoperative hospital stay, including morphine, pethidine, and fentanyl. The postoperative length of hospital stay is defined as the time between the date of surgery and the date of discharge. The R0 rate is defined as the proportion of patients who undergo complete resection among all patients undergoing surgery in each group. The number of dissected mediastinal LNs will be calculated based on an official postoperative pathological report. The total cost of hospitalisation is defined as the total cost incurred by the patient from admission to discharge. The number of preoperative and postoperative CTCs is calculated using 3 ml of peripheral arterial blood collected before the incision and immediately after closure, using EDTA tubes (BD Diagnostics, Sparks, MD, USA), stored at 4 °C, and analysed within 24 h. CTCs were quantified using the CytoploRare folate receptor-positive CTC assay (GenoSaber, Shanghai, China). The participant’s timetable is shown in Table 1, in accordance with the requirements of SPIRIT 2025 [8].

Table 1.

Participant timeline: schedule of enrolment, interventions and assessments

	Trial period
	Enrollment		Operation	Post-operation								Follow-up
Timepoint	Screening (≤ 7 days before surgery)	T0 (1 day before surgery)	Surgery day	Daily (post-operation in hospital)	Hospital discharge	2 weeks after surgery	4 weeks after surgery	8 weeks after surgery	12 weeks after surgery	26 weeks after surgery	52 weeks after surgery	Every 6 months (years 1–3)	Annually (after year 3)
Visit window	N/A	N/A	N/A	N/A	N/A	+/− 1 day	+/− 3 days	+/− 3 days	+/− 7 days	+/− 7 days	+/− 7 days	+/− 7 days	+/− 7 days
Enrollment
Eligibility screen	X
Informed consent	X
Randomization		X
Interventions
MLND-first			X
Lobectomy-first			X
Assessments
Intraoperative outcomes
Operating time			X
Blood loss			X
Conversion to thoracotomy			X
CTCs (pre- & post-surgery)			X
Intraoperative complications			X
Postoperative outcomes
R0 resection					X
Lymph node count					X
Length of hospital stay					X
Chest tube duration					X
Hospitalisation cost					X
Operative death (30 and 90 days)								X		X
Pain scorea				X	X	X	X	X	X	X	X
Acesodyne dose				X	X	X	X	X	X	X	X
Post-operation complications				X	X	X	X	X	X	X	X
Follow-up
Physical examinations												X	X
Enhanced chest and abdominal CT												X	X
Tumour marker assessments												X	X
Additional visits if new symptoms arise												As needed	As needed

^aPain score is a patient-reported outcome, while all other parameters are clinician-reported outcomes assessed by blinded research nurses

Eligibility criteria

Inclusion criteria

1. Age from 18 to 80 years

2. The first clinical diagnosis before surgery is lung tumour considering NSCLC.

3. Clinical stage T1-2N0-1M0 (cI–II stage): Maximum diameter of tumour ≤ 5 cm and short diameter of mediastinal LN ≤ 1 cm in thin-layer computed tomography (CT)

4. The patient’s physical condition is able to tolerate lobectomy: (1) Goldman index between grade 1 and 2, (2) predicted postoperative forced expiratory volume in 1 s ≥ 40% and diffusing capacity of the lung for carbon monoxide ≥ 40%, (3) total bilirubin ≤ 1.5 upper limit of normal, (4) alanine aminotransferase/aspartate aminotransferase ≤ 2.5 upper limit of normal, and (5) creatinine ≤ 1.25 upper limit of normal and creatinine clearance rate ≥ 60 ml/min

5. Performance status of Eastern Cooperative Oncology Group = 0–1

6. All relevant examinations completed within 28 days before the operation

7. Patients who understand this study and have signed an informed consent form

Exclusion criteria

1. Patients undergoing radiotherapy, chemotherapy, targeted therapy, or immunotherapy before surgery

2. Patients with a previous medical history of other malignancies

3. Patients with secondary primary cancer when enrolled

4. Patients who meet all of the following criteria which make them eligible for sublobar resection (segmentectomy/wedge resection): (1) Ground-glass opacity (GGO) with a solid component ≤ 50%, (2) the largest diameter of nodule is ≤ 2 cm, and (3) the nodule is located in the outer third of the lung field.

5. Patients diagnosed as showing pure GGO before surgery

6. Patients with prior unilateral open thoracic surgical procedures

7. Female patients who are pregnant or breastfeeding

8. Patients with active bacterial or fungal infection that is difficult to control

9. Patients with serious psychosis

10. History of severe heart disease, heart failure, myocardial infarction, or angina pectoris in the past 6 months

Regarding minimally invasive diagnostic procedures, patients who underwent video-assisted needle biopsy were allowed, whereas those who underwent video-assisted wedge resection were excluded.

Sample size calculation

The primary endpoint of this trial is 5-year DFS. Based on previous studies, the 5-year DFS rate for the control group is assumed to be 65% [9], with a superiority margin of 5%, and the experimental group is expected to achieve a 10% improvement. This corresponds to an assumed hazard ratio of 0.67 for the experimental group compared with the control. The type I error rate is set to 0.025 (one-sided), and the statistical power is 0.80 [10–12]. Factoring in a 10% dropout rate, approximately 310 patients per group are required for a total sample size of 620.

Randomisation

Stratified block randomisation will be performed in this trial. An independent randomisation committee will generate a random number table using SPSS software (version 25.0; IBM Corp, Armonk, NY, USA) before the study begins, and the table will remain confidential to researchers and related personnel. Third-party staff members will upload the table into the INCREASE clinical research platform (https://www.c-increase.com/). Principal investigators from each centre will receive a unique account on the platform. Once a patient is eligible for enrolment, the investigator can log into the platform to enrol the patient and proceed with the randomisation. This study will stratify patients by clinical stage (stages I and II), with block sizes randomly set at 4, 6, or 8, and assignments of the MLND-first and lobectomy-first groups will be made in a 1:1 ratio. Neither the researchers nor patients are blinded to the treatment methods used in this trial.

Treatments

Requirements for the surgeon

All participating surgeons in this study are attending thoracic surgeons from high-volume thoracic surgery centres. Each performs more than 100 lobectomies with MLND annually. They are experienced in both thoracotomy and minimally invasive approaches for lobectomy combined with MLND and are proficient in performing these procedures regardless of whether lobectomy or MLND is performed first.

Surgical standards

All patients will receive surgery under general anaesthesia with double-lumen endotracheal intubation. The surgical approaches are tailored according to the surgeon’s preference at each centre. Both the thoracotomy approach and the minimally invasive approach can be used for this trial. Video-assisted surgery can be performed using a single-port, single-utility port, or a multiport surgical approach. Robot-assisted surgery can also be performed. Both groups of patients should undergo a standard minimally invasive anatomical lobectomy (with bilobectomy permitted for the right lung) combined with systematic MLND, in accordance with the surgical guidelines of the International Association for the Study of Lung Cancer (IASLC) [13]. According to the IASLC guidelines, the criteria for systematic MLND are as follows [13]:

• At least three mediastinal LN stations should be examined, and the subcarinal LNs must be examined in all patients.

• Patients with left-sided NSCLC must undergo an examination of level 5/6/7 LNs, those with right-sided NSCLC must undergo an examination of level 2R/4R/7 LNs, and those with lower-lobe NSCLC must undergo an examination of level 9 LNs.

• The total number of dissected LNs must be greater than 6, including at least three mediastinal LNs, and all LNs and their surrounding tissues within the anatomical landmark must be completely removed.

• In the experimental group (MLND-first group), the patients will receive systematic MLND before lobectomy. In the control group (lobectomy-first group), patients will receive an anatomical lobectomy before starting MLND. These two surgical steps (lobectomy and MLND) should not be combined during surgery.

Postoperative therapy

Postoperative adjuvant treatment strategies will be determined by each centre based on the relevant clinical guidelines and patient wishes. If recurrence or metastasis is identified during follow-up, appropriate treatment protocols will be implemented promptly. A multidisciplinary approach is recommended to develop a scientific treatment protocol.

Follow‑up

Postoperative follow-up assessments are conducted by dedicated research nurses assigned by the principal investigator (PI) and blinded to treatment group allocation. The nurses are responsible for documenting all pre-specified outcome measures, including postoperative pain levels and the incidence of complications. All enrolled patients will be scheduled for regular follow-up assessments every 6 months in the first 3 years after surgical treatment and once a year thereafter until death or the end of the study. If new symptoms are identified during follow-up, the patients will be re-examined promptly. The review will include physical examinations, enhanced chest and abdominal CT scans, and tumour marker assessments. All imaging endpoints will be evaluated by independent radiologists who are blinded to treatment allocation to ensure objectivity in the assessments. In addition, the patients’ symptoms will need to be tested appropriately. If recurrence or metastasis occurs during the follow-up period, the sites of recurrence and metastasis, related symptoms, and subsequent treatments will be recorded.

Statistical analysis

All primary and secondary endpoint analyses will be conducted according to the intention-to-treat principle. Per-protocol and sensitivity analyses will also be performed. Subgroup analyses will be performed to evaluate the impact of other factors on prognosis, such as gender, age, clinical and pathological stage, histological subtype, use of adjuvant therapy, and surgical approach.

The Kaplan–Meier method will be used to calculate the survival rate of OS and DFS, and the log-rank test will be performed to compare OS and DFS between groups. In addition, a multivariate Cox proportional hazards regression model will be used to minimise the impact of confounding factors. Differences in categorical variables, such as R0 resection rate, conversion rate, complication rate, and surgical mortality, between the two groups will be assessed using the chi-square test or Fisher’s exact test. For differences in continuous variables such as operation time, blood loss, postoperative hospital stay, and chest tube duration, the t-test or Wilcoxon rank-sum test will be used for the two groups.

For studies involving multiple centres, we will address centre effects by using a mixed-effects model and performing stratified analyses by centre to ensure reliable results.

Data management and monitoring

Case report forms (CRFs) are completed promptly and accurately by investigators, with triplicate copies archived post-trial at the lead institution, trial organiser, and collaborating hospitals. Completed CRFs underwent clinical monitor review prior to data entry. The statisticians establish the database in a timely manner, and once reviewed, the database is locked by the principal investigator, sponsor, statisticians, and clinical monitors. All patient-related information is handled in strict accordance with applicable confidentiality regulations. Identifiable data are accessible only to authorised study personnel, such as investigators and research nurses. CRF contains only participant codes and initials, and any personal identifiers appearing in supporting documents (e.g. pathology or imaging reports) are removed from copies. Electronic data are stored securely and comply with local data protection laws. Data security is ensured, with authorised personnel allowed to access or modify the data, and all data must be backed up. In accordance with China’s Good Clinical Practice principles, data should be retained for at least 5 years. Regular inspections and on-site monitoring of the progress and quality of the study at each collaborating centre will be conducted by quality control personnel.

Data access and sharing

Following publication of the primary manuscript, de-identified individual participant data will be made available to the scientific community upon reasonable request for a scientifically sound proposal. All subsequent publications using shared data must acknowledge the study’s funding source.

Adverse events

Adverse events (AEs) will be documented and graded in accordance with the Common Terminology Criteria for Adverse Events Version 5.0. An AE is defined as any unfavourable clinical sign or complication occurring after the participant has provided informed consent. The most probable AEs include cardiovascular events (such as arrhythmia, myocardial infarction, and thromboembolic events), pulmonary complications (such as prolonged air leak, pneumonia, and atelectasis), and procedure-related events such as chylothorax, haemorrhage, and wound complications. Serious adverse events (SAEs) are those resulting in death, life-threatening conditions, hospitalisation or prolongation of hospitalisation, or significant disability/incapacity.

Investigators are responsible for ensuring participant safety and maintaining thorough documentation. Any SAE occurring during the study must be reported to the adverse event monitoring centre, study sponsor, and ethics committee, using appropriate reporting forms duly signed and dated.

Participants retain the right to withdraw at any time without penalty or compromise to their medical care. Investigators may discontinue a participant’s involvement if alternative treatment is necessary, a study-related injury occurs, or for any other valid reasons.

Protocol amendments

Any modifications to the study protocol will be documented as formal amendments and submitted promptly to the ethics committee for review and approval. No changes will be implemented until the ethics committee’s approval has been obtained. Following approval, any substantial changes will be updated on ClinicalTrials.gov.

The current protocol is version 2, dated 13 May 2025. Compared with the previous version (27 August 2024), only administrative updates were made. Specifically, the primary completion date and study completion date were extended by 1 year due to slower-than-expected patient enrolment. Several centres that withdrew from the trial were removed from the sponsor/collaborators and contacts/location sections. No changes were made to the study design, endpoints, or analysis plan.

Patient and public involvement

Neither patients nor the public participated in designing, conducting, or reporting this study. The findings will be disseminated via appropriate channels and social organisations to enhance their impact and inform future therapeutic approaches. Upcoming studies could involve direct engagement of patients and the public in the planning and decision-making stages of research.

Dissemination of results

The trial results will be disseminated to participating investigators, referring physicians, and the wider medical community through presentations at scientific conferences and publications in international peer-reviewed journals. Both positive and negative results will be reported. Authorship will follow the criteria established by the International Committee of Medical Journal Editors.

Discussion

For early-stage NSCLC patients, particularly those with tumours > 2 cm in diameter or located in the middle to inner third of the lung, clinical guidelines recommend anatomical lobectomy combined with MLND [1]. Standard MLND can not only reduce tumour recurrence to improve long-term survival for patients but also serve as an important condition for accurate staging [3, 14, 15], which is a crucial basis for determining whether patients need adjuvant therapy.

It is noteworthy that oncology textbooks advocate for the “no-touch isolation” technique during tumour resection, which emphasises dissecting distant LNs before proximal ones [5]. Based on this principle, MLND should precede lobectomy in lung cancer surgery. However, in the surgical treatment of NSCLC, the performance of MLND before or after lobectomy varies among surgeons largely depending on their preferences. In clinical practice however, most thoracic surgeons tend to perform lobectomy first and then proceed with MLND [4], because the surgical field in the chest cavity becomes capacious after lobectomy, which might make MLND more convenient. Unfortunately, we have noted that no evidence confirms the principle of dissecting tissues from distant to proximal in NSCLC surgery, and whether the surgical sequence (MLND first or lobectomy first) affects the long-term survival of NSCLC patients remains unclear.

Some researchers have illustrated the impact of surgical sequence on operative outcomes in other cancer types. Williams et al. found that in robot-assisted radical prostatectomy, performing pelvic LND before prostatectomy could achieve higher efficiency and better outcomes in prostate cancer patients [16]. Liang et al. found that the medial-to-lateral laparoscopic dissection approach for rectosigmoid cancer is quicker, less expensive, less invasive, and equally effective oncologically and postoperatively compared with the lateral-to-medial approach [17]. Wei et al. designed an RCT to explore the impact of vein-first versus artery-first surgical sequences on survival in patients with NSCLC. Their results demonstrated that ligating effluent veins first during surgery might reduce tumour cell dissemination and improve survival outcomes in patients with NSCLC [6]. This trial suggests that the sequence of ligation of different vessel types may affect the long-term prognosis of patients with NSCLC. Lymphatic metastasis is another crucial route of lung cancer metastasis, and the principle of dissecting LNs from distant to proximal sites has a theoretical foundation. Unfortunately, no evidence currently confirms which surgical approach—lymphadenectomy-first or lobectomy-first—offers greater long-term survival in patients with operable NSCLC. Therefore, we believe that large-scale multicentre RCTs and long follow-up periods are necessary to determine whether the MLND-first approach can significantly improve long-term outcomes for patients with operable NSCLC.

Meanwhile, perioperative liquid biomarkers, particularly CTCs, have emerged as valuable tools for real-time monitoring of tumour dynamics and minimal residual disease, enabling more precise evaluation of recurrence risk [18, 19]. Consistent with these advances, our trial incorporates perioperative CTC assessment as an exploratory endpoint to investigate whether different surgical sequences influence intraoperative tumour cell dissemination and long-term prognosis.

Limitations of this trial should be noted. The open-label design may introduce bias, though standardised protocols and blinded outcome assessment were used to reduce this risk. Surgical heterogeneity among centres and potential contamination of treatment sequence are possible, but predefined operative principles and regular monitoring were applied to maintain consistency. Finally, as the study is conducted in China, the generalisability of the findings to other healthcare systems might be limited.

To the best of our knowledge, the LOFTY study is the first prospective, multicentre RCT to evaluate the superiority of the MLND-first over the lobectomy-first surgical approach in terms of long-term and short-term outcomes in patients with clinical stages I–II NSCLC. The trial will include 620 patients with clinical stages I–II NSCLC, with 5-year DFS as the primary endpoint, and 5-year OS and short-term outcomes as the secondary endpoints. The results of this trial aim to define an evidence-based surgical sequence for operable NSCLC, promoting greater consistency and standardisation in clinical practice.

Trial status

The first participant was recruited on 16 August, 2023. The current protocol is version 2 of 2025 May 13. Currently (22nd of June 2025), we included 523 patients. Patient recruitment is estimated to be completed around December 2025.

Abbreviations

NSCLC

Non-small cell lung cancer

MLND

Mediastinal lymph node dissection

RCT

Randomised controlled trial

DFS

Disease-free survival

OS

Overall survival

CTCs

Circulating tumour cells

GGO

Ground-glass opacity

IASLC

International Association for the Study of Lung Cancer

CRFs

Case report forms

AEs

Adverse events

SAEs

Serious adverse events

Authors’ contributions

HXY and MZY designed the study protocol. HXY, MZY, YBL, XDS, GWM, HY, ZRZ, QWL, TYS, WD, BQ, JBL, JY, HZ, NZ, DQS, YJG, HNZ, XBC, WSC, XFL, SWC, HT, XZQ, JY, WQG, ML, ZMP, WW, WJJ, GYL, HYL, and JDM will oversee the implementation of the protocol, data collection, and data analysis. HRF and MZY drafted the initial version of the manuscript. All authors reviewed the manuscript.

Funding

This study was fully funded by the Excellent Surgery Study Project of Bethune Charitable Foundation (CESS2021LB15). This is an investigator-initiated trial. The sponsor provides financial support only and has no role in the study design, data collection, analysis, interpretation or paper writing.

Data availability

The research data will be available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study is conducted according to the Declaration of Helsinki and the relevant Chinese laws and regulations. The Ethics Committee of Sun Yat-sen University Cancer Center approved the study protocol in August 2022, and the date of the first participant enrolment was 16 August, 2023. Approval from the institutional review board was obtained before patient enrolment at each institution. Written informed consent was obtained from all enrolled patients. All 26 institutions have passed the ethical review.

Consent for publication

This report does not include any identifiable personal or clinical detail, and such information will not be included in future publications of the trial results.

Competing interests

The authors declare that they have no competing interests.