The aim of this study is to examine if a shorter palliative radiotherapy fractionation scheme of 20 Gy / 4 F can reduce the early oesophageal toxicity compared to 30 Gy / 10 F in patients with lung cancer in performance status (PS) 0-2. Secondary aims are to examine the effect on lung cancer symptoms, quality of life (QoL) and survival. Furthermore, the investigators aim at standardizing the quality of palliative thoracic radiotherapy in all Danish centres at the highest technical level.

1. Background 1.1 Radiotherapy in lung cancer Non-small-cell lung cancer (NSCLC) is the leading cause of cancer-related death in both men and women [1]. In stage IV disease platinum-based chemotherapy yields a median overall survival (OS) of 8-10 months [2]. When diagnosed, more than 50% of the patients have distant metastases. Forty percent have signs or symptoms originating from the thorax: dyspnoea, cough, haemoptysis, recurrent pneumonia or chest pain [3]. Palliative thoracic radiotherapy can relieve symptoms originating from intra-thoracic malignancy and improves quality of life (QoL) in approximately one third of patients [4]. It is increasingly used in daily clinical practice as an alternative to chemotherapy in patients without local symptoms with the purpose of prolonging life and avoiding local thoracic symptoms. In addition, palliative radiotherapy is used to relieve symptoms before initiating systemic treatment (chemotherapy, targeted agents, immune therapy, etc). In this latter case, it is important not to delay the systemic treatment unnecessary with long schedules of radiotherapy. An optimal radiotherapy regimen will thus palliate symptoms with minimal toxicity and consider the patient's time investment, as well as be compatible with any additional treatments given. In a recent retrospective study of 159 consecutive Danish NSCLC patients who received palliative thoracic radiotherapy (2010-11) median survival was 4.2 months. One third of the patients died within two months and 22% within 30 days [5]. This further underlines the importance of not subjecting patients to prolonged, toxic treatments in their final year of life. A recent Cochrane review [6] found that there was no consistent evidence to support that longer, more fractionated radiotherapy regimens gave better or more durable palliation than shorter regimens. Furthermore, there was no significant survival advantage associated with longer regimens with higher biological radiation doses.

   The most common acute toxicity is oesophagitis. It is often not technically possible to avoid high dose exposure to the oesophagus because of the target localization in proximity to the mediastinum. Radiation-induced oesophagitis is most often a reversible condition. However, it compromises swallowing due to pain, causes weight loss and deterioration of the performance state (PS) and QoL at a vulnerable time for the patient, and it may impact the possibility of receiving further anti-neoplastic treatment. The Cochrane review found that up to 50% of patients receiving palliative thoracic radiotherapy reported severe oesophagitis. The studies differed widely in terms of physician- or patient self-assessment, timing, and the method used, and no overall judgement could be made. The review concluded that toxicity was comparable between different radiotherapy regimens. However, recording and reporting of early toxicity was lacking in the included studies. This absence of early toxicity data means that the current evidence is insufficient for clinical decision making in this patient group.

2. Study design The patients will be invited to participate in the study prior to initiation of palliative radiotherapy. After signing informed consent the patients will be randomized 1:1 between two different radiotherapy dose fractionations: 30 Gy/10 F or 20 Gy/4 F. Physicians and patients will not be blinded to the randomization.

   2.1 Systemic therapy Patients can receive chemotherapy, immunotherapy and targeted therapy during radiotherapy.

   2.2 Study sample size and time frame According to the Danish Lung Cancer Registry, 50% of the 4700 lung cancer patients diagnosed yearly are in advanced stage. Of these, one third is supposed to benefit from palliative thoracic radiotherapy and of these, 400 patients (~50%) are expected to be enrolled on the trial each year. Since 1184 consecutive patients are required (see chapter 7) the study is expected to enrol patients over a 3-year period. Given the declared support of all Danish radiotherapy centres, this is considered a realistic time frame.

   Following study closure, 3 weeks of follow-up after radiotherapy initiation will be allowed for all patients, at which point data will be collected for primary analysis. Further analysis will be performed 1 and 3 years after the last patient is enrolled. Data will be kept in databases 15 years after the last patient is included.

3. Radiotherapy treatment planning 3.1 Target and organs at risk The gross tumour volume (GTV) is defined by the referring oncologist based on a planning CT scan, and if available, a diagnostic PET/CT scan. The oncologist will note in the CRF if the GTV fully or only partially encompass all active malignant disease in the thorax. The clinical target volume (CTV) is identical to the GTV, and thus not delineated.

   Spinal cord or spinal canal, total lung, heart and the oesophagus are delineated for all patients based on the guidelines in [5,6]. The heart is cranially limited defined by the division of truncus pulmonalis. The oesophagus is delineated from cartilago criocoideum to the gastro-oesophageal junction.

   3.2 Dose planning Patients are treated with conformal treatment plans, conventional or intensity-modulated.

4. Treatment verification and quality assurance 4.1 Verification of treatment position Daily imaging must be performed and used for daily correction of treatment position.

5. Participants Patients with lung cancer stage III-IV not suitable for curative treatment, being referred for palliative radiotherapy either alone or in addition to systemic treatment.

   5.1 Evaluations during treatment and follow up 5.1.1 Expected side effects Palliative RT comes with a risk of side-effects including fatigue, pain and soreness in the throat and chest, reddening of the skin in the irradiated area and dry cough and dyspnoea.

   5.1.2 Scoring of symptoms and QoL Symptom scores on oesophagitis, cough, pain, dyspnoea, bronchopulmonary haemorrhage, performance status and QoL will be registered at baseline, end of treatment, 2 weeks, 3 weeks, 8-10 weeks and 6 months after radiotherapy completion. If the patient has symptoms which need clinical assessment, the patient will get an appointment with a physician. All complication (side effects) will be evaluated according to the NCI-CTCAE v 5.0 during all follow-up visits or calls [7]. See appendix C for an overview of the CTCAE v 5.0. Table 6.1 shows the evaluation plan for the trial. Symptom scores on oesophagitis, cough, pain, dyspnoea, bronchopulmonary haemorrhage, and assessment of performance status will be registered during the patient on-site visit at baseline, end of treatment, 8-10 weeks after treatment and 6 months after treatment completion. At 2 weeks and 3 weeks, symptom scores will be registered by a phone call. Questionnaires for registration of QoL will be filled in on-site by the patient at baseline, end of treatment, 8-10 weeks after completion of treatment and 6 months after completion of treatment. In addition, two questionnaires for registration of QoL to be filled in 2 weeks and 3 weeks after completion of treatment will be handed out at end of treatment. During the phone calls at 2 weeks and 3 weeks, the patient will be reminded to fill in the questionnaire. The questionnaires will be collected at the on-site visit at 8-10 weeks after treatment. One or both phonecalls at 2 weeks and 3 weeks may be substituted by on site visits.

   After the end of the study-specified visit, patients will be seen every 3rd month according to the guidelines by the Danish Health Authorities. Evaluation of toxicity will be performed at every occasion. End of study is 3 years after commencement of radiotherapy.

   5.2 Quality of Life To examine the participants' quality of life (QoL) during follow up, they will be asked to fill in a QoL questionnaire. The EuroQoL 5D (EQ-5D-5L) questionnaire will be used.

6. Statistical considerations and planned data analyses 6.1 Sample size calculation for primary endpoint - acute oesophagitis The primary study endpoint is reduction in early oesophageal toxicity, as assessed two weeks after completion of radiotherapy. The investigators assume that 50% of patients who receive the standard treatment (30 Gy in 10 fractions) will experience grade 2 or worse oesophagitis, compared to their baseline prior to treatment. A reduction in acute toxicity of 10%-point with the experimental treatment (odds ratio between arms of 0.67) will be clinically meaningful and will justify a change of practice for this patient group. This corresponds to an expected 40% rate of acute oesophagitis in the experimental arm. With 90% power and a 2-sided type 1 error rate of 5%, 1184 patients will be required (assuming 1:1 allocations between treatment arms, and incorporating a 10% loss to follow-up). There are some clinical indications that oesophageal-related toxicity will mainly be seen in patients with centrally located tumours (where palliative radiotherapy results in irradiation of mediastinal structures). These patients make up approximately two thirds of the palliative radiotherapy patient population. An incidence of oesophagitis of 70% in these patients (and ~10% in patients with peripheral tumours) is consistent with the observed overall toxicity rate. Enrolling 790 patients with central tumours in the study will result in >95% power to detect a 15%-point reduction of toxicity in this subgroup, from 70% to 55% (consistent with an overall reduction of 10%, assuming very limited effect in patients with peripheral tumours). Alternatively, there will be nearly 80% power to detect a 10%-point toxicity reduction in the central tumour subgroup. Any such subgroup analyses will be completely exploratory. All sample size and power calculations for the primary endpoint are based on two-group chi-squared tests (continuity corrected) of equal proportions.

   6.2 Non-inferiority of overall survival Current best evidence indicates that overall survival should not be different with shorter compared to longer palliative treatment regimens, although the literature is not conclusive. The proposed study will have the power to test for non-inferiority of the short (experimental) treatment arm compared to the standard treatment: Assuming a median OS of 4.2 months, inclusion of 1184 patients over 3 years (with one year of additional follow-up) will give approximately 90% power to test that median OS in the experimental arm is at most 3 weeks shorter than in the standard arm (one-sided significance level of 95%). Power calculations for overall survival are based on Cox's proportional hazards model.

   6.3 Planned data analyses All analyses will be conducted on an intention to treat (ITT) basis. The primary endpoint analysis will take place once all enrolled patients have reached the primary endpoint assessment (2 weeks after completion of radiotherapy). A multivariate logistic regression model will be used to compare rates of grade 2+ acute oesophagitis in the two treatment arms, adjusted for stratification factors. Robustness of the primary endpoint to the timing of oesophagitis assessment will be examined in secondary analyses, comparing the two arms with respect to oesophagitis rates at 3 weeks after end of therapy as well as oesophagitis rates at 4 weeks after the start of therapy (two weeks after end of therapy in the standard arm and three weeks after end of therapy in the experimental arm). Significance testing will be two-sided, with a 95% significance level.

   Analysis of secondary endpoints will depend on the follow-up required for each endpoint; with most endpoints (QoL, response assessment) available at the time of primary analysis, and initial analysis of overall survival conducted one year after randomisation of the final patient.

   6.4 Randomization

   • Patients are stratified prior to randomization by treatment institution, performance status 0/1 versus 2, and histology (small cell carcinoma vs squamos cell carcinoma vs non-squamos cell carcinoma).
   • Patients will be randomized 1:1 between the two different radiotherapy dose fractionations

7. Complications All complication (side effects) will be evaluated according to the NCI-CTCAE v 5.0 during all followup visits.

8. References [1] Jemal A, CA Cancer J Clin 2011 Mar;61(2):69-90 [2] Wao H, Syst Rev 2013 Feb 4;2:10 [3] Beckles MA1Chest. 2003 Jan;123(1 Suppl):97S-104S. [4] L angendijk JA, Int J Radiat Oncol Biol Phys 2000 Apr 1;47(1):149-55 [5] Støchkel Frank M, BMC Palliat Care. 2018 Jan 5;17:15. [6] S tevens R, Cochrane Database Syst Rev 2015 Jan 14;1:CD002143. [7] Kong FM, Int J Radiat Oncol Biol Phys 2011 Dec 1;81(5):1442-57.

Inclusion criteria:

• Histologically or cytologically confirmed NSCLC or SCLC.
• Stages III-IV not candidates for curative treatment
• Age ≥18 years.
• Performance status: 0-2
• Signed informed consent.
• Fertile women must have a negative pregnancy test. Fertile men and women must use effective contraception. Fertile women included in the study must use the pill, spiral, depot injection of gestagen, subdermal implantation, hormonal vaginal ring or transdermal patch for the duration of study treatment and one month thereafter.

Exclusion criteria:

• Prior radiotherapy to the thorax that prohibits the delivery of 30 Gy/10 F with respect to OAR dose constraints.
• Patients not able to understand the written or spoken information.