Identification of dose constraints and evaluation of optimal planning technique for radical thoracic re-irradiation for non-small cell lung cancer

Rulach, Robert John (2023) Identification of dose constraints and evaluation of optimal planning technique for radical thoracic re-irradiation for non-small cell lung cancer. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2023RulachPhD.pdf] PDF
Download (7MB)



There are approximately 48,000 new diagnoses of lung cancer in the United Kingdom. It is one of the most lethal cancers, with a 10% chance of survival at 10 years. A third of patients receive radiotherapy a part of the primary treatment for lung cancer. However, there is an approximately 30% local recurrence rate after radical radiotherapy for non-small cell lung cancer, and there is a 14% risk at 10 years of developing a second lung cancer. There are no treatment guidelines for patients who are diagnosed with intra-thoracic relapsed disease.
Radical thoracic re-irradiation for non-small cell lung cancer has been performed in selected patients from the 1970s with promising efficacy. However, re-irradiation is associated with increased risk of toxicity compared to de novo radiotherapy. Re-irradiation is being delivered more frequently due to the advances in radiotherapy technology and better detection of recurrent disease, despite the lack of evidence in how to deliver re-irradiation safely, or any recent prospective studies that demonstrate efficacy.

Aim of thesis

The aim of this thesis is to investigate how to optimise the safety of radical thoracic re-irradiation, in preparation for a future prospective clinical trial.

(i) An international Delphi consensus process with thoracic oncologists was performed to identify current practice in thoracic re-irradiation, patient selection, develop dose constraints and radiotherapy planning strategies.
(ii) A retrospective review of 39 patients who underwent re-irradiation in the Beatson West of Scotland cancer centre was conducted. Clinical outcomes and cumulative dosimetric information were analysed. Image and dose registration strategies were developed to account for the previous dose delivered from initial radiotherapy with the re-irradiation dose.
(iii) A literature review was performed to collect information (including toxicity, cumulative dose, interval between treatments and use of chemotherapy) about thoracic re-irradiation. This was divided into five datasets for the organs at risk in the chest (spinal cord, oesophagus,
lungs, proximal bronchial tree, and aorta) and logistic regression modelling was performed to determine cumulative dose constraints.
(iv) A literature review was performed to collect information (including cumulative dose, local control, and overall survival rates) from thoracic re-irradiation for non-small cell lung cancer. Logistic regression modelling was performed to determine the dose required for 50% rates of 2-year local control and overall survival.
(v) A radiotherapy planning study using the 39 Beatson patients was conducted using volumetric arc therapy (VMAT) and multi-criteria optimisation (MCO). Patients were re planned to the cumulative dose constraints and the models developed in sections (iii) and (iv) were applied to assess if the re-planned re-irradiation was safer.
(vi) Patients who had completed a course of radical lung radiotherapy were recruited into a qualitative interview study to explore patients’ perspectives on re-irradiation. The interviews were transcribed and underwent thematic analysis.

(i) Fifteen lung oncologists participated in the Delphi process. Patients being considered for radical re-irradiation should be PS 0-2, and radical resection should be discussed. Staging with PET-CT and brain imaging was endorsed. Consensus dose constraints based on clinician expertise were agreed upon for the oesophagus, spinal cord, brachial plexus and aorta. There was no consensus for lung and proximal bronchial tree doses.
(ii) Clinical outcomes and cumulative dose of 39 patients from the Beatson were analysed and divided into patients with local recurrence and second primary lung cancers. The 2-year OS rate was 38.5% in the local recurrence group, and 69.2% in the SPLC group. Sixteen patients (41%) experienced grade 3 toxicities and one patient (2.6%) had fatal haemoptysis. A reproducible process to accumulate dose was developed, which identified that using the whole lung for image registration was the optimal strategy.
(iii) The literature search identified 55 studies with the cumulative dose and toxicity required for modelling. Dose/toxicity models were developed using logistic regression for the spinal cord, oesophagus, the mean lung dose, the lung V20Gy, the proximal bronchial tree and aorta. There was insufficient data to model the heart, chest wall and brachial plexus dose. For the spinal cord, oesophagus, proximal bronchial tree and aorta, the maximum likelihood 5% risk of grade 3 toxicity was seen at 77.2Gy, 94.3Gy, 157.5Gy and 142.5Gy respectively (all doses in equivalent dose in 2-Gray fractions, median values used if other variables were included in the model). The mean lung dose and V20Gy associated with 20% grade 3 toxicity were 19.3Gy and 28.4% respectively. These models were validated on the Beatson data, and dose constraints developed.
(iv) The literature search identified 21 studies with 2-year local control or overall survival data and cumulative dose to the tumour. Dose/outcome models were developed using logistic regression modelling. The modelling predicted a 50% 2-year local control rate at 67.8Gy using a median planning target volume of 112cc. The predicted dose to the tumour for a 50% 2-year overall survival rate was 76.5Gy. A sub-study to assess if 13 locally recurrent patients from the Beatson cohort could have dose escalation identified six patients where their re-irradiation dose could potentially be increased.
(v) The planning study identified 15 patients from the 39 patients in the Beatson cohort that breached the re-irradiation dose constraints. These patients mostly had locally recurrent disease. Seven patients were replanned using VMAT and MCO and met the dose constraints. The remaining eight patients required alternate strategies (such as a change in dose fractionation or modification of the planning target volume) to meet the constraints. Six patients were able to be safely replanned with these alternate strategies. The combination of VMAT and MCO was superior to VMAT alone when planning re-irradiation for sparing the serial organs at risk in the chest.
(vi) Eight patients participated in a qualitative interview exploring their perspectives on radical re-irradiation. Thematic analysis identified two main themes from the interviews: fear and control. The key finding was that all patients would consider re-irradiation. A common reason given was they were not afraid of it having experienced radiotherapy before. Each patient had a very different attitude to risk, with some patients stating that they would accept high risk treatment if the outcomes were better, whereas others preferred to prioritise avoiding toxicity.

The research detailed in this thesis contributes to the delivery of safe re-irradiation in several ways. The consensus statements provided guidance for the selection and staging of patients to be considered for radical re-irradiation, ensuring that only suitable patients proceed with high-risk treatment. The dose constraints developed from the dose/toxicity can be used to limit severe re-irradiation toxicity and allows patients to be better counselled prior to treatment. The dose/outcome study identified that recurrent disease required higher doses for disease control, and that dose escalation may be possible in selected patients. The planning study identified that the optimal planning technique is VMAT with MCO. The qualitative study
demonstrated that patients may consider re-irradiation and require individual counselling regarding their acceptance of risk. This research provides insights to the inclusion criteria, dose constraints, radiotherapy planning technique and the patient and public involvement necessary for a prospective clinical study of re-irradiation.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Additional Information: Supported by funding from The Beatson Cancer Charity, The University of Glasgow and Cancer Research UK.
Subjects: R Medicine > RC Internal medicine > RC0254 Neoplasms. Tumors. Oncology (including Cancer)
Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Cancer Sciences
Supervisor's Name: Harrow, Dr. Stephen and Chalmers, Professor Anthony
Date of Award: 2023
Depositing User: Theses Team
Unique ID: glathesis:2023-84009
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 21 Dec 2023 09:25
Last Modified: 21 Dec 2023 09:28
Thesis DOI: 10.5525/gla.thesis.84009

Actions (login required)

View Item View Item


Downloads per month over past year