RESEARCH SYNOPSIS — REVISED
For Submission to the Institutional Ethics Committee
Title
Comparative Effectiveness of Self-Expanding Metal Stents (SEMS) Versus Radiotherapy for Palliation of Dysphagia and Pain in Treatment-Naive Patients with Locally Advanced or Metastatic Oesophageal Cancer Unfit for Standard Anticancer Treatment: A Retrospective Cohort Study at a Tertiary Cancer Centre (January 2017 – December 2023)
Principal Investigator
[Name, Designation, Department, Institution — to be completed]
Co-Investigators
[To be completed]
Department
[Department of Oncology / Gastroenterology / Radiation Oncology — to be completed]
Institution
[Tertiary Cancer Centre — to be completed]
1. Background and Rationale
Oesophageal cancer is the seventh most common cancer globally and ranks sixth in cancer-related mortality. The overwhelming majority of patients present with locally advanced or metastatic disease, where curative intent is not achievable. A particularly underserved subgroup — patients unfit for standard anticancer therapy (surgery, systemic chemotherapy, or radical chemoradiotherapy) due to poor performance status, significant comorbidity, nutritional depletion, or advanced age — constitutes a substantial proportion of the real-world oesophageal cancer population and is consistently under-represented in clinical trials.
In this population, the therapeutic mandate shifts entirely to symptom control and quality of life preservation. The two dominant and debilitating symptoms are:
- Dysphagia — present in 80–90% of patients, leading to nutritional failure, aspiration risk, and progressive functional decline.
- Pain — oesophageal, retrosternal, and back pain arising from tumour infiltration, peritumoural inflammation, and neural invasion; present in a significant proportion of patients and inadequately captured in most palliative intervention trials.
Two principal modalities are employed for locoregional symptom control in this setting:
-
Self-Expanding Metal Stents (SEMS): Endoscopically deployed mechanical stents that restore luminal patency within 24–48 hours. The ESGE 2021 guidelines (Spaander et al.) issue a strong recommendation for SEMS as the first-line intervention for malignant dysphagia, particularly in patients with short life expectancy. However, SEMS are associated with significant complication rates (stent obstruction 30–40%, migration, fistula, haemorrhage) and carry no tumour cytoreductive effect — and therefore limited value in pain palliation arising from tumour mass effect.
-
Radiotherapy (EBRT / Brachytherapy): Palliative external beam radiotherapy (typically 20–30 Gy in 5–10 fractions) or intraluminal brachytherapy acts through tumour volume reduction, offering both luminal relief and reduction of pain from local infiltration. The onset of dysphagia relief is delayed (4–6 weeks), but the duration is more sustained. Critically, radiotherapy has a well-recognised role in pain palliation from locally advanced thoracic tumours — a benefit absent from mechanical stenting.
The landmark ROCS phase III RCT (Adamson et al., Lancet Gastroenterol Hepatol, 2021; n=220) found that adjuvant EBRT after SEMS did not improve dysphagia deterioration or overall survival versus SEMS alone, but significantly reduced bleeding events. Importantly, ROCS enrolled patients receiving combined SEMS + RT — not a direct monotherapy head-to-head comparison.
The largest comparative study of RT versus stenting as independent modalities (Martin et al., J Natl Compr Canc Netw, 2020; n=1,957 US veterans with metastatic oesophageal cancer) found that RT was associated with more rapid and durable pain relief (p < 0.001), equivalent dysphagia relief over time, and a significantly lower rate of severe adverse events (12.4% vs. 21.7% at 6 months, p < 0.001). This provides direct evidentiary support for studying pain as a co-primary outcome alongside dysphagia.
A single-centre Indian study of palliative EBRT (30 Gy/10 fractions) in 70 patients with locally advanced oesophageal squamous cell carcinoma (Bandhate & Diwan, Rep Pract Oncol Radiother, 2022) demonstrated significant improvement in both dysphagia and odynophagia scores (p < 0.0001), as well as pain domain scores on the EORTC QLQ-OES18 at 3 months. Despite this evidence, pain palliation is rarely designated as a primary outcome in oesophageal palliative intervention studies, representing a critical evidence gap.
Specifically for the treatment-naive, unfit patient subgroup — defined here as patients with locally advanced or metastatic disease who have received no prior anticancer therapy and are not candidates for chemotherapy, surgery, or radical chemoradiotherapy — there are no dedicated comparative effectiveness studies. This population is managed almost entirely on the basis of extrapolated evidence from broader palliative cohorts that often include patients with prior therapy, confounding the assessment of both treatment response and adverse event risk.
This retrospective cohort study will address this gap by systematically comparing SEMS and radiotherapy as first-line palliative interventions for both dysphagia and pain in this precisely defined, homogeneous patient group.
2. Aims and Objectives
Primary Aims
This study has two co-primary aims:
- To compare the effectiveness of SEMS versus radiotherapy in relieving dysphagia at 4 weeks post-intervention.
- To compare the effectiveness of SEMS versus radiotherapy in reducing pain at 4 weeks post-intervention.
Primary Objectives
- Dysphagia: To compare the proportion of patients achieving clinically meaningful dysphagia improvement (≥1-point reduction on the Ogilvie Dysphagia Score, or equivalent) at 4 weeks between the SEMS and radiotherapy groups.
- Pain: To compare the proportion of patients achieving clinically meaningful pain reduction (≥2-point reduction on the Numerical Rating Scale [NRS] 0–10, or equivalent documented pain score) at 4 weeks between the SEMS and radiotherapy groups.
Secondary Objectives
- To compare overall survival (OS) from the date of palliative intervention to death or last follow-up between the two groups.
- To compare time to re-intervention (rescue stent, repeat radiotherapy, or other dysphagia intervention) as a measure of durability of dysphagia relief.
- To compare complication and adverse event rates — including stent migration, tumour ingrowth/overgrowth, perforation, fistula formation, radiation oesophagitis, and haemorrhage — between the two groups.
- To assess health-related quality of life (HRQoL) using available patient-reported outcome data (EORTC QLQ-C30 / QLQ-OES18 dysphagia, pain, and eating restriction domains, where documented).
- To evaluate change in nutritional status (body weight, serum albumin) from baseline to 4 and 12 weeks.
- To identify clinical and tumour factors associated with differential response to each modality in both dysphagia and pain palliation.
3. Study Design
Design: Single-centre, retrospective comparative cohort study.
Setting: Tertiary Cancer Centre.
Study Period: January 2017 – December 2023 (7 years).
Comparison Arms:
- Group A (SEMS): Treatment-naive patients who received SEMS as the primary palliative intervention for dysphagia.
- Group B (Radiotherapy): Treatment-naive patients who received palliative EBRT or intraluminal brachytherapy as the primary palliative intervention.
4. Study Population
Inclusion Criteria
- Histologically or cytologically confirmed oesophageal carcinoma (squamous cell carcinoma or adenocarcinoma) or oesophagogastric junction (OGJ) carcinoma.
- Locally advanced (T3/T4 or N+, unresectable) or metastatic (Stage IV) disease at the time of intervention, as documented by clinical staging workup.
- Treatment-naive: No prior receipt of any anticancer therapy — including systemic chemotherapy, targeted therapy, immunotherapy, surgical resection, radical-intent radiotherapy, or chemoradiotherapy — at any time prior to the study intervention.
- Unfit for standard anticancer treatment as determined by the multidisciplinary team (MDT), based on one or more of the following: ECOG performance status ≥ 3, significant comorbidity precluding systemic therapy or surgery, patient refusal of anticancer treatment, or clinician judgement of treatment futility.
- Receipt of SEMS placement or palliative radiotherapy (EBRT or BT) as the primary intervention for dysphagia or pain between January 2017 and December 2023.
- A documented baseline dysphagia score and/or pain score (NRS or equivalent) within 2 weeks prior to intervention.
- Age ≥ 18 years at time of intervention.
- Adequate follow-up documentation at a minimum of 4 weeks post-intervention.
Exclusion Criteria
- Any prior anticancer treatment (chemotherapy, immunotherapy, targeted therapy, surgery, or radical radiotherapy) before the study intervention.
- Patients undergoing SEMS as a bridge to surgery or before preoperative chemoradiotherapy.
- Patients receiving SEMS and concurrent EBRT simultaneously (not a pure monotherapy arm).
- Benign oesophageal strictures or non-malignant indications for intervention.
- Patients with incomplete records precluding extraction of baseline and at least one post-intervention assessment of dysphagia or pain.
- Patients with a documented life expectancy of less than 2 weeks at time of intervention, where meaningful outcome assessment is not feasible.
5. Sample Size Considerations
This is a retrospective cohort study; the effective sample size is determined by the eligible institutional caseload over the 7-year study period. All eligible patients meeting inclusion criteria will be enrolled. A formal post-hoc power calculation will be performed after data extraction.
Assumptions for the co-primary pain outcome, based on Martin et al. (2020): proportion achieving meaningful pain relief at 4 weeks estimated at 60% for radiotherapy and 35% for SEMS. Using a two-sided α = 0.05, 80% power, and 1:1 allocation, a minimum of 53 patients per arm (n=106 total) is required. This is expected to be achievable within the study period at a tertiary centre.
For the dysphagia co-primary outcome, based on published data: proportion achieving ≥1-point Ogilvie improvement estimated at 78% (SEMS) vs. 58% (radiotherapy) at 4 weeks. With equivalent parameters, a minimum of 61 patients per arm (n=122 total) is required.
The larger of the two estimates (n=122 total) will govern the minimum required sample size, ensuring adequate power for both co-primary outcomes. A Bonferroni correction (α = 0.025 per primary outcome) will be applied to control the family-wise error rate for dual primary testing.
6. Data Collection
Source of Data
All data will be extracted retrospectively from:
- Electronic patient records (EPR) and hospital information systems
- Endoscopy procedure logs and radiology/fluoroscopy reports
- Oncology and gastroenterology clinic letters and MDT records
- Radiotherapy planning, dosimetry, and treatment completion records
- Pain assessment documentation (nursing notes, oncology clinic assessments, analgesic prescription records)
- Pathology and histology databases
- Nutritional assessment records (dietitian notes, weight charts, albumin results)
- Mortality records (hospital, national death registers where accessible)
Variables to be Collected
Baseline / Demographic:
- Age, sex, ethnicity
- ECOG performance status at time of intervention
- BMI, body weight, serum albumin
- Comorbidity burden (Charlson Comorbidity Index)
- Reason for unfitness for standard therapy (performance status, comorbidity, patient choice, MDT decision)
Tumour Characteristics:
- Histological subtype (SCC vs. adenocarcinoma)
- Tumour location (upper / middle / lower third; OGJ)
- Tumour length and degree of luminal obstruction on endoscopy or imaging
- Clinical stage (TNM 8th edition)
- Presence of tracheo-oesophageal or broncho-oesophageal fistula
- Pattern of metastatic spread (nodal only, distant organ, peritoneal)
Intervention Details:
- SEMS group: Stent brand/type, covering (full/partial/uncovered), diameter, technique (fluoroscopic ± endoscopic), procedural complications, stent patency duration
- Radiotherapy group: Modality (EBRT vs. BT), dose and fractionation (total dose, fraction size, number of fractions), technique (2D vs. 3D-conformal), treatment completion (complete vs. incomplete)
Primary Outcome Variables:
- Dysphagia score (Ogilvie scale or equivalent): baseline, 2 weeks, 4 weeks, 3 months
- Pain score (NRS 0–10 or VAS): baseline, 2 weeks, 4 weeks, 3 months
- Analgesic use and WHO ladder step at baseline and follow-up (as proxy for pain burden)
Secondary Outcome Variables:
- Date and cause of death / last follow-up date
- Date and type of re-intervention for dysphagia or pain
- Adverse events (documented complications, graded by CTCAE v5.0 where possible)
- Body weight and serum albumin at 4 and 12 weeks
- QoL scores (EORTC QLQ-C30 / QLQ-OES18 or equivalent, where recorded)
7. Statistical Analysis Plan
Descriptive Analysis
- Continuous variables: mean ± SD or median (IQR) based on distribution (Shapiro-Wilk)
- Categorical variables: frequencies and percentages
- Baseline group comparisons: independent t-test or Mann-Whitney U (continuous); chi-square or Fisher's exact test (categorical)
Co-Primary Outcome Analysis
Both primary outcomes will be tested independently with a Bonferroni-corrected significance threshold of α = 0.025 for each.
- Dysphagia: Proportion achieving ≥1-point Ogilvie improvement at 4 weeks compared using chi-square test; OR with 95% CI. Change in ordinal score compared using Mann-Whitney U test.
- Pain: Proportion achieving ≥2-point NRS reduction at 4 weeks compared using chi-square test; OR with 95% CI. Change in continuous NRS score compared using independent t-test or Mann-Whitney U test as appropriate.
Secondary Outcome Analysis
- Overall survival: Kaplan-Meier curves with log-rank test; multivariable Cox proportional hazards regression (adjusted HR, 95% CI), adjusting for age, ECOG PS, stage, histology
- Time to re-intervention: Kaplan-Meier analysis; Fine-Gray competing risks regression (death as competing event)
- Adverse events: Chi-square / Fisher's exact test for between-group comparison
- Nutritional parameters: Paired Wilcoxon signed-rank test within groups; Mann-Whitney U between groups
- QoL: Descriptive; paired within-group comparison using Wilcoxon signed-rank test where longitudinal data exist
Adjustment for Confounding
Given the inherently non-randomised, observational design — and the expected difference in baseline characteristics between patients selected for SEMS (typically more severe dysphagia, worse PS) versus radiotherapy — propensity score analysis will be performed. Propensity scores will be estimated via logistic regression using baseline covariates (age, sex, ECOG PS, tumour stage, histology, tumour location, baseline dysphagia score, baseline pain score). Both 1:1 nearest-neighbour matching and inverse probability of treatment weighting (IPTW) will be used, with results compared for consistency.
Subgroup Analyses
- Histological subtype (SCC vs. adenocarcinoma)
- Performance status (ECOG 2 vs. ≥3)
- Tumour location (proximal/middle vs. distal/OGJ)
- Stage (locally advanced vs. metastatic)
- Radiotherapy modality (EBRT vs. BT) — within the radiotherapy arm
Statistical analyses will be performed using SPSS v28 and/or R v4.3 (packages: survival, cmprsk, MatchIt, WeightIt).
8. Ethical Considerations
Study Type and Consent Waiver
This is a retrospective, observational, non-interventional study involving analysis of existing clinical records. No new interventions, procedures, or patient contact are planned. An application for waiver of individual informed consent is made on the grounds that:
- All patients are being managed under routine clinical care; the study involves no additional burden
- Retrospective record review of this nature is recognised as appropriate without individual consent under national and international ethical guidelines for non-interventional research
- Data will be fully anonymised prior to analysis
Data Privacy and Security
- All patient identifiers will be replaced with unique study codes prior to data entry into the analytical database.
- Data will be stored in a password-protected, encrypted electronic database on an institutional server, accessible only to named investigators.
- Data will be handled in strict compliance with applicable data protection legislation (GDPR / national equivalent) and institutional information governance policy.
- Original identifiable extraction sheets will be stored securely and destroyed at study completion per institutional policy.
Risk-Benefit Assessment
This study carries no risk to participants. The anticipated benefits are substantial: generating real-world evidence on both dysphagia and pain palliation in a poorly studied, vulnerable patient group will directly inform clinical practice, MDT decision-making, and future prospective study design at this institution and beyond.
Declaration of Interests
The investigators declare no financial or non-financial conflicts of interest. No commercial or external funding is involved. This is an investigator-initiated study.
9. Limitations
- Retrospective design with selection bias: Patients were not randomised; the choice of SEMS vs. radiotherapy was determined by clinician judgement and patient factors. Patients with more severe dysphagia, worse PS, or tumour-related fistulae may have been disproportionately directed to SEMS. Propensity score methods will mitigate but cannot fully eliminate this.
- Pain documentation variability: Systematic pain scoring (NRS/VAS) may not have been uniformly recorded across all clinic encounters throughout the study period. Analgesic prescription records will be used as a supplementary proxy where formal scores are absent.
- Treatment heterogeneity within arms: Radiotherapy fractionation schedules and stent designs varied over the 7-year period. Subgroup and sensitivity analyses will address intra-arm heterogeneity.
- Outcome ascertainment completeness: Not all patients will have documented 4-week assessments; those lost to follow-up before the primary endpoint will require appropriate missing data handling (multiple imputation or complete-case sensitivity analysis).
- Unmeasured confounders: Factors such as opioid co-prescription, performance status trajectory, nutritional support intensity, and social support — which influence both pain experience and functional outcomes — may not be systematically captured in the medical record.
- Single-centre experience: Results may reflect institutional referral patterns and local practice norms; generalisability to other settings requires external validation.
10. Expected Outcomes and Significance
This study is expected to:
- Generate the first dedicated comparative analysis of pain and dysphagia palliation by SEMS versus radiotherapy in a treatment-naive, unfit oesophageal cancer population, addressing a critical evidence gap.
- Provide real-world, institutional evidence to guide individualised MDT decisions about palliative treatment allocation in this underserved group.
- Characterise the safety profile of each modality in a population who have not received prior therapies that could amplify complications (e.g., radiation-induced wall fragility, chemotherapy-related neutropenia).
- Generate pilot data to inform the design and sample size of a future prospective study or pragmatic RCT targeting this specific patient population.
- Contribute to the national and international evidence base on pain palliation in oesophageal cancer, an outcome currently not designated as a primary endpoint in existing phase III trials.
Findings will be submitted for publication in a peer-reviewed journal (target: Supportive Care in Cancer, Journal of Pain and Symptom Management, or Diseases of the Esophagus) and presented at national oncology or gastroenterology meetings.
11. Timeline
| Phase | Activity | Duration |
|---|
| Month 1 | Ethics committee approval; data governance clearance | 4 weeks |
| Months 1–3 | Patient identification; data extraction from EPR | 8 weeks |
| Months 3–4 | Data cleaning, coding, pseudonymisation, database finalisation | 4 weeks |
| Months 4–5 | Statistical analysis (primary, secondary, propensity score) | 4 weeks |
| Months 5–6 | Manuscript preparation and journal submission | 6 weeks |
| Total | | ~6 months |
12. References
- Adamson D et al. Palliative radiotherapy after oesophageal cancer stenting (ROCS): a multicentre, open-label, phase 3 RCT. Lancet Gastroenterol Hepatol 2021. [PMID 33610215]
- Adamson D et al. ROCS RCT — Health Technology Assessment. Health Technol Assess 2021. [PMID 34042566]
- Spaander MCW et al. Esophageal stenting for benign and malignant disease: ESGE Guideline Update 2021. Endoscopy 2021. [PMID 33930932]
- Martin EJ et al. Palliative Radiotherapy Versus Esophageal Stent Placement in Metastatic Esophageal Cancer (n=1,957). J Natl Compr Canc Netw 2020. [PMID 32380466]
- Bandhate K, Diwan A. Palliative radiotherapy and quality of life in locally advanced thoracic oesophageal cancer. Rep Pract Oncol Radiother 2022. [PMID 35402038]
- Xu Z et al. Palliative radiotherapy combined with stent insertion for dysphagia in advanced oesophageal cancer: systematic review and meta-analysis. Front Oncol 2022. [PMID 36172146]
- Lancellotta V et al. Role of palliative interventional radiotherapy (brachytherapy) in oesophageal cancer: AIRO systematic review. Brachytherapy 2020. [PMID 31636025]
- Yang ZM et al. Radioactive Stent for Malignant Oesophageal Obstruction: Meta-Analysis of RCTs. J Laparoendosc Adv Surg Tech 2021. [PMID 32915105]
- Clinical Gastrointestinal Endoscopy Expert Consult, 3rd ed. Chapter 28: Endoscopic Management of Oesophageal Cancer.
Declaration
I, the undersigned Principal Investigator, confirm that the information provided in this synopsis is accurate and complete, and that this study will be conducted in accordance with the Declaration of Helsinki, Good Clinical Practice (GCP), and all applicable institutional and national regulatory requirements.
Principal Investigator Signature: _________________________ Date: _____________
Head of Department Signature: _________________________ Date: _____________
Completion checklist before submission:
- Fill in all investigator names, designations, and institutional details
- Confirm the pain scoring tool routinely used at your centre (NRS, VAS, or EORTC QLQ-OES18 pain domain) — adjust pain primary outcome definition accordingly
- Confirm your institution's dysphagia scoring tool (Ogilvie, Mellow-Pinkas, or EORTC QLQ-OES18 dysphagia subscale)
- Attach any required supplementary documents per your ethics committee's submission checklist (e.g., PI CV, co-investigator declarations, data extraction template, data storage agreement)
- A formal consent waiver application form may need to be completed separately depending on your institution's requirements