Article Text
PDF
Abstract
Background Sentinel node biopsy is a surgical technique to explore lymph nodes for surgical staging of endometrial cancer, which has replaced full retroperitoneal lymph node dissection. However, the effectiveness of sentinel node biopsy, its value to patients, and potential harms compared with no-node dissection have never been shown in a randomized trial.
Primary Objectives Stage 1 will test recovery from surgery. Stage 2 will compare disease-free survival at 4.5 years between patients randomized to sentinel node biopsy versus no retroperitoneal node dissection.
Study Hypothesis The primary hypothesis for stage 1 is that treatment with sentinel node biopsy will not cause detriment to patient outcomes (lymphedema, morbidity, loss of quality of life) and will not increase treatment-related morbidity or health services costs compared with patients treated without a retroperitoneal node dissection at 12 months after surgery. The primary hypothesis for stage 2 is that disease-free survival at 4.5 years after surgery in patients without retroperitoneal node dissection is not inferior to those receiving sentinel node biopsy.
Trial Design This phase III, open-label, two-arm, multistage, randomized non-inferiority trial (ENDO-3) will determine the value of sentinel node biopsy for surgical management of endometrial cancer. Patients with endometrial cancer are randomized to receive: (1) laparoscopic/robotic hysterectomy, bilateral salpingo-oophorectomy with sentinel node biopsy or (2) laparoscopic/robotic hysterectomy, bilateral salpingo-oophorectomy without retroperitoneal node dissection. In stage 1, 444 patients will be enrolled to demonstrate feasibility and quality of life. If this is demonstrated, we will enroll another 316 patients in stage 2.
Major Inclusion and Exclusion Criteria Inclusion criteria include women aged 18 years or older with histologically confirmed endometrial cancer; clinical stage 1, who meet the criteria for laparoscopic or robotic total hysterectomy and bilateral salpingo-oophorectomy. Patients with uterine mesenchymal tumors are excluded.
Primary Endpoints The endpoint for stage 1 is surgical recovery, with the proportion of patients returning to usual daily activities at 3 months post-surgery as measured with the EQ-5D. Stage 2 is disease-free survival at 4.5 years.
Sample Size 760 participants (both stages).
Estimated Dates for Completing Accrual and Presenting Results Stage 1 commenced in January 2021 and is planned to be completed in December 2024 when 444 participants have completed 12 months' follow-up. Stage 2 will enroll a further 316 participants for a total of 760 patients.
Trial Registration NCT04073706.
- gynecologic surgical procedures
- lymph nodes
- surgical procedures
- operative
- uterine cancer
Data availability statement
There are no data in this work.
Statistics from Altmetric.com
Introduction
Endometrial cancer is the most common gynecological cancer in Australia (approximately 3000 new cases per annum), and its incidence is expected to rise further.1 The current standard treatment for endometrial cancer involves a total hysterectomy and bilateral salpingo-oophorectomy to remove the primary tumor with or without a retroperitoneal lymph node dissection for surgical staging and to inform decisions about adjuvant treatment. The extent and value of retroperitoneal node dissection for staging has been a topic of controversy for more than 30 years2 and remains unsolved as its advantage to patients has never been conclusively demonstrated.3–5 In a significant proportion of patients, it may cause morbidity, including lymphedema, and prolong the operating time and costs of treatment.3 6
Sentinel node biopsy is used increasingly to remove selected lymph nodes.7 A sentinel node is defined as the lymph node into which the endometrial cancer drains first.8 If the sentinel node is cancer-free, it is assumed that all other nodes are cancer-free also. By contrast, involvement of the sentinel node assumes that the patient has advanced disease and requires adjuvant treatment. In the setting of endometrial cancer, sentinel node biopsy involves the injection of a dye into the uterine cervix, and specialized technical equipment is required to identify the sentinel node in the pelvis. Compared with a full node dissection, where commonly between 10 and 25 nodes are removed, this technique limits the extent of node dissection to only one or two nodes on each side and is thought to reduce surgical morbidity.9 10 Although sentinel node biopsy has been shown to replace a systematic lymph node dissection reliably in prospective observational studies,11–13 no prospective randomized trial data are available on its effectiveness in endometrial cancer for patient-reported quality of life or survival outcomes. Data on survival outcomes following sentinel node biopsy in endometrial cancer are limited to a small number of retrospective series, and studies that compare against historical controls, showing good agreement between sentinel node biopsy with the information from a full node dissection,8 14 while data on patient-reported outcomes are sparse.15 Overall, the value of sentinel node biopsy to patients, its cost-effectiveness, potential harm (procedure-related morbidity), and its effectiveness to guide adjuvant treatment compared with no-node dissection has never been established in a randomized controlled trial.15 Given that the use of sentinel node biopsy is increasing, high-level evidence is urgently needed to answer those questions.
Hypothesis
The primary hypothesis for stage 1 is that treatment with sentinel node biopsy will not cause detriment to patient outcomes (lymphedema, morbidity, loss of quality of life) and will not increase treatment-related morbidity or health services costs compared with patients treated without a retroperitoneal node dissection at 12 months after surgery. The primary hypothesis for stage 2 is that disease-free survival at 4.5 years after surgery in patients without retroperitoneal node dissection is not inferior to those receiving sentinel node biopsy.
Methods and Analysis
Trial Design
Study esign
This phase III, open-label, two-group, multistage randomized non-inferiority trial (ENDO-3) is an Australian and New Zealand Gynaecological Oncology Group (ANZGOG) trial (ANZGOG trial 1911/2020) with national and international participation. Ethics approval was obtained from the Royal Brisbane and Women’s Hospital Human Research Ethics Committee (HREC/2019/QRBW/49921). The number of national and international sites will evolve progressively. Participants with endometrial cancer are randomized to receive either (1) laparoscopic/robotic hysterectomy, and bilateral salpingo-oophorectomy with sentinel node biopsy or (2) laparoscopic/robotic hysterectomy, and bilateral salpingo-oophorectomy without retroperitoneal node dissection. In stage 1 of the ENDO-3 trial, 444 patients will be enrolled to demonstrate feasibility of enrollment and to determine short-term outcomes. If this is demonstrated, the ENDO-3 trial will enroll another 316 patients in stage 2 to reach a total of 760 patients.
The ENDO-3 trial is currently funded by the Royal Brisbane and Women’s Hospital Foundation, Wesley Medical Research Foundation, the Cherish Women’s Cancer Foundation, Gynecological Cancer Research Education and Development Society, and an National Health and Medical Research Council Investigator Grant. Figure 1 shows the trial schema.
Trial schema. MMMT, mixed malignant Mullerian tumor.
Surgical Treatment
The surgical standard treatment of endometrial cancer is currently poorly defined. A large number of gynecological oncologists worldwide practice surgical staging with sentinel node biopsy, for which limited evidence is available and no randomized controlled clinical trial has investigated. Omitting surgical staging through a retroperitoneal node dissection is practiced sparsely, but previous evidence comparing outcomes of patients randomized to no-node dissection versus systematic node dissection demonstrated superior short-term outcomes in favor of the no-node dissection group and comparable survival outcomes. The ENDO-3 trial will compare short-term and survival outcomes of patients randomized to sentinel node biopsy versus no-node dissection. Patients in both groups will receive a total laparoscopic/robotic hysterectomy and bilateral salpingo-oophorectomy. Details of the surgical protocol are given in online supplemental appendix 1.
Supplemental material
Sentinel Node Biopsy
The mandatory, optional, and unwarranted steps of a sentinel node biopsy procedure have been described previously.16 In brief, indocyanine green at a concentration of between 0.5 mg/L and 1.5 mg/L is injected into the cervix in two or four positions, superficially or superficially and deeply. If a uterine manipulator is used it is inserted after the tracer injection. At laparoscopy, a white-light inspection of the pelvic and abdominal cavity is performed to exclude peritoneal spread. The round ligament and the infundibulopelvic ligament can be secured and divided or preserved prior to sentinel node biopsy. The external and internal iliac vessels, ureter, obliterated umbilical (hypogastric) ligament, uterine artery, and paravesical space are identified. Sentinel node dissection of mapped nodes commences at the level of the uterine artery or alternatively, may commence at the level of the most highlighted node and seeks to identify the most proximal mapped node. Surgical dissection aims to avoid disruption of lymphatic channels. A sentinel node is defined either as the single highlighted node closest to the uterus and may include highlighted nodes at the next station (echelon nodes). The lowest number of mapped nodes should be removed. Sentinel node biopsy is completed in one hemipelvis before proceeding to the contralateral side. In patients who do not map a sentinel node, a full ipsilateral lymph node dissection will be performed. Nodes are removed using a containment device and ex vivo green fluorescence is used to prove the sentinel node. Nodes are labeled according to laterality and nodal station (obturator/external iliac/internal iliac/presacral/common iliac/aortic/caval).
Sentinel nodes will undergo pathological ultrastaging following procedures as detailed by Kim et al.17
Adjuvant Treatment
The risk of relapse is determined by a nomogram (Coral Gynae) from uterine tumor factors (FIGO grade, depth of invasion, lymphovascular space invasion).18 Patients at high risk of relapse and patients with positive nodes will be offered adjuvant treatment.18 Each site will specify its policy for radiation treatment and chemotherapy prior to enrolling patients. Compliance with this policy will be monitored by the trial office.
Post-operative and Follow-up Study Procedures
All study participants, irrespective of treatment group, will be followed up for 4.5 years according to the schedule of clinical assessments detailed in online supplemental appendix 2. Participants will be assessed for local pelvic disease, extrapelvic recurrence, quality of life status, and treatment morbidity. Data collected will include documentation of the status of disease, adverse events lymphedema, pain level, health status, and quality of life. The presence of lymphedema will be assessed via bioimpedance spectroscopy and/or self-report questionnaire.19 Patients will be seen every 3 to 6 months for the first 2 years and then 6 monthly until 4.5 years as per current National Comprehensive Cancer Network guidelines.20
Management of Participants with Intra-operative Findings of Advanced Stages of Endometrial Cancer
In participants of both treatment groups who have signs of suspected intra-abdominal disease, a frozen section should be performed to confirm the disease involvement. If disseminated intraperitoneal disease is confirmed histologically, the aim of treatment will shift from staging to cytoreduction, if feasible. In such a case, it will be at the surgeon’s discretion whether to continue surgery, continue any staging efforts, or continue efforts towards sentinel node biopsy. If the surgeon decides to continue with a retroperitoneal node dissection, the extent of the node dissection is at the surgeon’s discretion. In some participants this would entail conversion from laparoscopic/robotic surgery to open surgery. We anticipate the number of participants with extra-uterine peritoneal disease will be small as patients with evidence of extra-uterine disease on presurgical medical imaging are ineligible for this trial.
Participants
Site Selection
Potentially eligible hospitals will be contacted by the project manager to assess interest in participating in the trial. Sites must have access to technical equipment to perform sentinel node biopsy with indocyanine green on a near-infrared fluorescence platform, and be willing to follow the procedures as outlined in the study protocol.
Participant Screening
All sites will maintain a screening log. It will record each patient screened for the trial, eligible or otherwise. If the patient is not randomized, then the reasons why will be indicated. Patient eligibility and exclusion criteria are listed in Table 1.
Participant inclusion and exclusion criteria
Participating Surgeons
To be considered as a surgeon for this clinical trial, participating surgeons are required to be qualified gynecological oncologists with a proven track record in clinical research and hospital privileges to perform laparoscopic/robotic hysterectomy and sentinel node biopsy for endometrial cancer. Prospective surgeons must have completed a minimum number of 30 cases performed as the main surgeon and sentinel lymph node mapper. Surgeons should perform sentinel node biopsy at least twice a month (averaged over 6 months) on uterine cancer with sentinel node biopsy (bilateral) being detected in at least 7 of the 10 most recent cases. Surgeons will submit two videos showing a bilateral sentinel node biopsy procedure, which will be assessed by the trial Surgical Quality Assurance Committee using a validated competency assessment tool in sentinel node biopsy.16 The Surgical Quality Assurance Committee will consider if the prospective surgeon meets the standard required for the ENDO-3 trial, needs further training to meet the standards, or is not meeting the standards set out for this trial.
Quality Assurance
All sentinel node biopsy cases that are performed as part of this trial will be video recorded and submitted to the Trial Management Committee. Thirty-five percent of all submitted videos will be randomly selected for review by the Surgical Quality Assurance Committee to ensure that the mandatory, optional, and unwarranted steps of sentinel node biopsy are continuously adhered to throughout the trial. The independent Data Safety Monitoring Committee will also monitor the bilaterality rate of each surgeon and the number of all removed nodes (sentinel, non-sentinel, enlarged/suspicious) to ensure that surgical quality assurance is maintained, along with the number of times suspicious nodes are removed and their size to ensure the clinical endpoints of this trial will not be affected. This information will be recorded, monitored, and reported to the Trial Management Committee.
Should a surgeon’s bilaterality rate drop below 70%, the surgeon will be informed. Should these rates not improve in the following 3 months, appropriate measures will be discussed with the surgeon—for example, further training required, or inability to continue to enroll participants to the ENDO-3 trial.
Primary Outcomes
The primary stage 1 endpoint is to compare the recovery of participants, defined as return to daily activities at 3 months, incidence of adverse events, lower limb lymphedema, health-related quality of life, and costs to the healthcare system of sentinel node biopsy for the surgical treatment of endometrial cancer at 12 months post-surgery.
The primary stage 2 endpoint is to compare disease-free survival for participants randomized to receive hysterectomy, and bilateral salpingo-oophorectomy with sentinel node biopsy compared with participants randomized to hysterectomy, and bilateral salpingo-oophorectomy without retroperitoneal node dissection at 4.5 years post-surgery.
The secondary stage 2 endpoints include: compare patterns of recurrence and overall survival between the groups; determine the cost-effectiveness of sentinel node biopsy; compare peri-operative outcomes (duration of surgery, length of hospital stay, intra-operative blood loss, blood transfusion requirements) and the incidence of intra-operative and post-operative adverse events within 12 months from surgery between the groups; compare lower limb lymphedema at 12 months after surgery; compare the need for post-operative (adjuvant) treatments between groups; determine the impact of body composition and frailty on survival, quality of life, lymphedema, peri-operative, intra-operative, and post-operative outcomes; compare follow-up strategies (clinical vs symptom checklist); and a biobanking strategy to enable translational research (Trans-ENDO 3).
Sample Size
The MRC ASTEC trial21 showed a hazard ratio (HR) for disease-free survival (adjusted disease-specific survival) of 1.12 (95% CI 0.75 to 1.69) and 1.35 (95% CI 1.06 to 1.73) for recurrence-free survival; for the Italian trial,22 the HR was 1.1 (95% CI 0.7 to 1.71). Using the inverse-variance method, this yields a pooled HR of 1.112 (95% CI 0.82 to 1.51) or based on recurrence-free survival in the MRC ASTEC trial, 1.24 (95% CI 0.92 to 1.68). It is anticipated that sentinel node biopsy management would improve disease-free survival particularly in patients with microscopic or positive nodes as they would receive a course of appropriate chemotherapy. The minimum clinical benefit of sentinel node biopsy management over the current standard of care (no treatment) is expected to be at least 5% in the 4.5 year disease-free survival rate. The estimate of the 4.5-year rate in the control group is 90% in the Italian study and ~93% in the MRC study. Thus, a control rate in this cohort of 4.5-year disease-free survival of 92% would seem reasonable.
A non-inferiority margin of, at most, 5% at 4.5 years would be clinically worthwhile in trading off the extra morbidity associated with lymph node dissection.
Based on a 5% non-inferiority margin, and a disease-free survival control rate with sentinel node biopsy management of 92%, a sample size of 720 participants would have 80% power with 95% (one-sided) confidence to detect this difference at 4.5 years assuming a 4-year accrual and a minimum 4-year follow-up time. Allowing for a modest 5% loss to follow-up, the total sample size for both stages of the study would be 760 participants.
For stage 1 of the trial, sample size calculations were based on our prospective LEGS study6 that the incidence of lower limb lymphedema will be approximately 30% at 12 months from surgery. The hypothesis for this stage of the trial is that the incidence at 12 months is comparable between the two treatment groups. A difference of less than 10% in lower limb lymphedema would be clinically insignificant. A minimum of 444 patients would be required to detect this difference. For the targeted final sample size of 760 patients, this would have >90% power to detect a 10% difference from 30% to 20% and 80% power to detect an 8% difference from 30% to 22%.
Randomization and Blinding
Participants will be randomly assigned in a 1:1 ratio to the intervention and control groups. Blinding of participants and surgeons is not feasible due to the surgical intervention, but statistical analysis will be conducted in a blinded fashion, and an outcomes adjudication committee will also assess outcomes blinded.
All eligible participants will be required to provide informed consent before being randomized. Prior to randomization, each patient will be screened for eligibility according to the inclusion and exclusion criteria. Once the above criteria have been met, participants will be randomized using an interactive voice response system that will be coordinated centrally from the National Health and Medical Research Council Clinical Trials Center. The method of randomization will be minimization and participants will be stratified by histological cell type (endometrioid vs other), treatment center, body mass index (<30 kg/m2, 30–45 kg/m2, >45 kg/m2), and age (<50 years, 50–70 years, >70 years). Participants are required to have surgery within 30 days of randomization.
Statistical Methods
CONSORT and the SPIRIT-PRO Extension guidelines will be followed for the reporting of trial data. Analysis will be performed according to the principle of intention to treat. Standard descriptive statistics will be used to summarize patient characteristics.
Return to Usual Activities, Health-related Quality of Life, Fear of Recurrence
For the stage 1 primary outcome, we will calculate the proportion of women who report problems with daily activities at 3 months post-surgery and use Χ2 statistics to compare treatment groups. We will report the proportion of women whose health-related quality of life and other patient-reported outcomes changed from baseline as a clinically relevant outcome.23 By convention, at least a 5% difference or more in health-related quality of life scores are considered clinically relevant. The maximum improvement will be categorized as: no improvement <5%, 6%–10%, 11%–15%, or 16–>20% improvement. For early recovery, we will use the maximum improvement from weeks 1 and 6, unless the patient did not complete both measurements, in which case the single available measurement will be used. Forest plots will be used to examine early (up to 6 weeks post-surgery), intermediate (3–6 months post-surgery), and late (up to 12 months post-surgery) recovery from baseline and assess the relative recovery, comparing patients treated with sentinel node biopsy with those receiving no-sentinel node biopsy, with positive values favoring the no-sentinel node biopsy group. Contingency tables will be used to compare improvement for sentinel node biopsy and no-sentinel node biopsy, and Χ2 statistics will be used to compare groups. Regression analyses will be used to assess the impact of demographic and clinical factors associated with return to daily activities, fear of recurrence, and health-related quality of life.
Peri-operative Outcomes, Intra-operative and Post-operative Adverse Events
We will compare the incidence of (1) intra-operative, (2) post-operative CTC grade ≥3, and (3) serious adverse events at early time points (within 6 weeks) versus delayed (3 to 12 months post-surgery). Χ2 tests will be used to compare the incidence of adverse events between the two treatment groups; and a t test to compare operating time, blood loss, and length of hospital stay, if appropriate. Regression analyses may be used to assess the impact of demographic and clinical factors associated with adverse events.
Lower Limb Lymphedema
We will compare the point prevalence (presence of lymphedema: yes/no at each assessment), cumulative burden (presence of lymphedema: ever/never at each assessment), and cumulative incidence (presence of lymphedema: ever/never, excluding those with evidence of lymphedema at baseline) of lymphedema using either bioimpedance spectroscopy and/or self-report over time (baseline, 6 weeks, 3 months, 6 months, 9 months and 12 months post surgery) between those with and without sentinel node biopsy, using generalized estimating equations (time x group), with adjustment for relevant baseline patient, treatment, and behavioral characteristics.
Requirements for Post-operative Radiation Treatment and/or Chemotherapy
Whether radiation treatment and/or chemotherapy will be recommended and commenced within 6 months from surgery will be compared between both groups using Χ2 statistics.
Blood Samples
For a future translational studies, baseline blood samples will be collected, processed, and stored following standard operating procedures from patients who consent to this optional component of the study.
Disease-free Survival
The percentage of 4.5-year information available for disease-free survival will be calculated to determine if sufficient follow-up has occurred to warrant a final analysis of the primary endpoint. As a rule, follow-up will continue, if this percentage is <75%.
The unadjusted 4.5-year disease-free survival rates will be compared using the difference between proportions obtained using the method of Kaplan-Meier. Non-inferiority will be confirmed if the 97.5% one-sided confidence interval for this difference does not cross the non-inferiority margin of 5%.
Appropriate regression methods will be used to explore the impact of other factors on disease-free survival and other secondary outcomes. Multivariable regression methods (linear, logistic, proportional hazards) will be used to explore the relationship between outcomes (disease-free survival, lower limb lymphedema, etc) and patient, treatment, and disease factors. No formal missing data value imputation is planned.
Cost Effectiveness
The total costs will be estimated for intervention and control groups. The change to costs from a decision to adopt the intervention will be compared with the change to functional outcome to reveal the cost per unit of improvement in functional outcome, measured in terms of the primary outcome. A complete cost-effectiveness evaluation will be undertaken to measure the quality-adjusted life years gained with the intervention. Calculations of quality-adjusted life years will be based on preference-based health measures for trial participants from the EQ-5D-5L. Several viewpoints will be considered for the economic analyses, including those of health system purchasers, households, and society in general. The CHEERS checklist24 will be used for reporting cost-effectiveness, and decision uncertainties will be addressed with probabilistic analyses.25
Cost-effectiveness analysis will follow processes previously described by Janda et al26 to obtain the necessary data from various cost sectors, including intervention costs; general practitioner and specialist consultations; radiology and imaging; prescriptions and over-the-counter medications; community health and social services; days off work; and informal care by family and friends. Direct costs will be obtained for samples of participants, stratified by hospital, operation and outcome, using a bottom-up approach by recording the volume of resource use in both groups of the trial, and then applying a unit cost to each component. Data on all participants’ use of healthcare services will be collected using a combination of a retrospective questionnaire and clinical files. Where possible, local cost tariffs will be used, and national sources will be used as comparators.
Discussion
The effectiveness of sentinel node biopsy and surgical staging for patients with early-stage endometrial cancer remains a matter of controversy over several decades. Despite lack of high-level evidence for benefit, large numbers of women who require treatment for endometrial cancer are offered surgical staging, which prolongs operating time and may increase patients’ risk of surgical-related morbidity. According to the innovation, development, exploration, assessment, and long-term study (IDEAL) model, no surgical innovation should be implemented without evaluation.27 The planned outcome of this research is to provide definitive evidence as to whether or not sentinel node biopsy is as effective as no lymph node dissection in the surgical management of early-stage endometrial cancer, or whether clinical management can be performed without the information from sentinel node biopsy. If the latter is shown, patients and healthcare systems can avoid the additional surgical burden.
Conclusion
The outcomes from this trial will provide evidence on the effectiveness of sentinel node biopsy, benefit and harm to patients, and will determine the value of retaining or discontinuing this procedure in the treatment of early-stage endometrial cancer.
Data availability statement
There are no data in this work.
Ethics statements
Patient consent for publication
Acknowledgments
The authors acknowledge the contributions of Trudi Cattley and Vanessa Behan from the University of Queensland and Peter McCulloch from the IDEAL Collaboration for their assistance and inspirations in ENDO-3 protocol development.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Twitter @AndreasObermair, @Jim Nicklin, @_SandiHayes, @cfsalomo, @MonikaJanda
Contributors All authors planned, designed, and started the study; VG developed the statistical methods and is the study statistician. NG is the study health economist who will assess the relative effectiveness of the intervention. AO and MJ drafted the initial manuscript. All authors contributed to critical review and revision of the manuscript and approved the final version. AO reports full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish.
Funding This study was funded by Royal Brisbane and Women’s Hospital (RBWH) Foundation, NHMRC Investigator Grant, Cherish Women’s Cancer Foundation, Wesley Medical Research Foundation.
Competing interests AO reports grants, personal fees, and other funding from SurgicalPerformance Pty Ltd, and grants from Medtronic, not directly related to the subject of this manuscript. AO reports consultancy fees from Baxter Healthcare Australia and New Zealand and Astra Zeneca Australia, not directly related to the subject of this manuscript. In addition, AO has a trademark licensed to SurgicalPerformance Pty Ltd. MJ reports speaker fees from Novartis, not related to this project.
Provenance and peer review Commissioned; internally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.