AI & Radiomics for Stratification of Lung Nodules After Radically Treated Cancer
- Conditions
- Indeterminate Pulmonary NodulesLung MetastasesSecond Primary CancerLung Cancer
- Interventions
- Other: Non-Interventional Study
- Registration Number
- NCT05375591
- Lead Sponsor
- Royal Marsden NHS Foundation Trust
- Brief Summary
This study will assess the utility of radiomics and artificial intelligence approaches to new lung nodules in patients who have undergone radical treatment for a previous cancer.
- Detailed Description
Improvements in cancer detection and diagnosis have led to increasing numbers of patients being diagnosed with early stage cancer and potentially receiving curative therapy with improved survival outcomes. Recent retrospective studies in cancer survivors have demonstrated such patients possess an increased risk of further cancer in their lifetime compared to the general population, in part potentially due to shared lifestyle risk factors (e.g. smoking), genetic cancer pre-disposition or downstream oncogenic side effects of anti-cancer therapies (eg. radiotherapy). Lung cancer remains the leading cause of cancer related deaths worldwide and the lungs also represent a common site for metastatic disease in patients with non-pulmonary malignancy. Furthermore, lung cancer is one of the most common second primary malignancy in patients with a prior history of treated cancer. Therefore, discerning the significance of a pulmonary nodule in the context of a previous cancer remains a clinical challenge given it may possess the potential to represent benign disease, metastatic relapse or new primary malignancy.
This study will assess the utility of radiomics and artificial intelligence approaches to new lung nodules in patients who have undergone radical treatment for a previous cancer. This will entail use of machine learning (ML) approaches and later, exploration of deep-learning/convolutional neural network approaches to nodule interpretation for differentiation of benign, metastatic and new primary lung cancer nodules/lesions. Development of a ML classifier or deep learning based tool may help guide which patients would benefit from earlier investigations including additional imaging, biopsy sampling and lead to earlier cancer diagnosis, leading to better patient outcomes in this unique cohort. This is a retrospective study analysing data already collected routinely as part of patient care. All data will be anonymised prior to any analysis, no patient directed/related interventions will be employed and consent-waiver for study inclusion will be exercised.
Recruitment & Eligibility
- Status
- RECRUITING
- Sex
- All
- Target Recruitment
- 1000
-
Confirmed history of previous radically or curative-intent treated solid organ cancer within 10 years of new index CT thoracic scan demonstrating a new pulmonary nodule and either of the following:
- Biopsy confirming previous malignancy with MDT consensus and successful cancer resolution/remission following anti-cancer treatment on interval imaging or blood assay analysis
- Where biopsy was not possible/confirmed for previous malignancy, MDT consensus outcome confirming cancer (+/- calculated Herder score >80% if applicable) and decision to treat as malignancy with subsequent resolution/remission following anti-cancer treatment on interval imaging or blood assay analysis
-
Radical treatment for previous cancer defined as either of the following:
- Surgical resection
- Radical radiotherapy or stereotactic beam radiotherapy
- Radical chemotherapy
- Radical chemo-radiotherapy
- Multi-modality treatment with any of the above
-
New pulmonary nodule ground truth known
- Scan data showing 2-year stability (based on diameter or volumetry) or resolution in cases of benign disease
- Scan data showing progressive nodule enlargement or increase in nodule number on interval imaging with MDT consensus (+/- PET with Herder score >80% if applicable) determining metastatic disease or new primary malignancy
- Biopsy sampling confirming benign disease or malignancy and in cases of malignancy, metastasis or new primary lung cancer
-
CT scan slice thickness ≤ 2.5mm
-
Nodule size ≥ 5mm
- CT Imaging > 10 years old
- Non-solid haematological malignancies including leukaemia
- Cases of radically treated primary cancer disease with early oligometastatic recurrence treated radically
Study & Design
- Study Type
- OBSERVATIONAL
- Study Design
- Not specified
- Arm && Interventions
Group Intervention Description Metastatic Nodules Non-Interventional Study CT scans of patients with a new lung nodule(s) subsequently confirmed to be metastatic in nature and in the context of a previous history of radically treated cancer, will be identified at participating NHS sites and recruited. Second Primary Lung Cancers Non-Interventional Study CT scans of patients with a new lung nodule(s) subsequently confirmed to be a new second primary lung cancer and in the context of a previous history of radically treated cancer, will be identified at participating NHS sites and recruited. Benign Nodules Non-Interventional Study CT scans of patients with a new lung nodule(s) subsequently confirmed to be benign and in the context of a previous history of radically treated cancer, will be identified at participating NHS sites and recruited.
- Primary Outcome Measures
Name Time Method Development of a CT-thorax based radiomics ML classifier model to predict cancer risk in new lung nodules after previous radically treated cancer. 2 years The study aims to identify distinct clusters of radiomics variables to generate a radiomics predictive vector (RPV), which can be used to stratify benign vs malignant nodules in patients who have previously received radical treatment for a malignancy. The RPV will be used in multivariate analysis and compared to existing risk models used in clinical practice.
Development of the CT-thorax based ML classifier model to predict whether a new malignant nodule represents metastatic lung disease (new cancer vs previous cancer recurrence) or a new primary lung malignancy. 2 years The study aims to identify distinct clusters of radiomic variables to generate a radiomics predictive vector (RPV) which is able to differentiate metastatic lung nodules from new primary lung cancer in patients who have previously received radical treatment for a cancer. No current models exist in clinical practice which address this diagnostic challenge.
- Secondary Outcome Measures
Name Time Method To evaluate performance the developed CT-thorax based ML classifier model in an independent external validation cohort. 2 years The investigators aim to assess performance of the derived radiomics predictive vector (RPV) on an external independent post-cancer lung nodule dataset to evaluate generalisability and potential real-world performance.
Trial Locations
- Locations (2)
The Royal Marsden NHS Foundation Trust (Chelsea Site)
🇬🇧London, United Kingdom
Royal Brompton Hospital
🇬🇧London, United Kingdom