Radiomics for Prediction of Survival in GBM
Study Details
Study Description
Brief Summary
Radiomics, the extraction of large amounts of quantitative image features to convert medical images into minable data, is an in-development field that intends to provide accurate risk stratification of oncologic patients. Published prognostic scores only take clinical variables into account. The investigators hypothesize that a combination of CT/MRI features, molecular biology and clinical data can provide an accurate prediction of medical outcome. The long term objective is to build a Decision Support System based on the predictive models established in this study.
| Condition or Disease | Intervention/Treatment | Phase |
|---|---|---|
|
|
Detailed Description
Human oncologic tissues exhibit strong phenotypic differences. Due to advances in both acquisition and analysis methods of medical imaging technologies, the extraction of reliable and informative image features to quantify these differences is currently possible. Radiomics, the extraction of large amounts of quantitative image features to convert medical images into minable data, is an in-development field that intends to provide accurate risk stratification of oncologic patients. Previous studies at Maastro demonstrated the importance of a large number (n=440) of these radiomics features to quantify the tumor phenotype by intensity, shape and texture. A landmark study was the extraction of imaging features from computed tomography (CT) data of 943 patients with non-small cell lung cancer (NSCLC) and head and neck squamous cell carcinoma (HNSCC) cancer in six distinct datasets 1. The model was trained on lung cancer patients and showed that a large number of radiomics features also have strong prognostic power in head and neck cancer patients. These data suggest that radiomics signatures decode a general prognostic phenotype existing in both NSCLC and HNSCC patients.
The investigators anticipate that their results will be a starting point of a novel field applying advanced computational methodologies on to medical imaging data, merging fields of medical imaging and bioinformatics. Also, as CT imaging has been applied in routine clinical oncology practice over decades in almost every hospital worldwide, the application of their analysis has potential to improve decision support in a large number of cancer patients.
Glioblastoma (GBM) are the most common type of primary brain tumors with an annual incidence of approximately 500 patients in the Netherlands. Despite extensive treatment including a resection, radiation therapy and chemotherapy, the median overall survival is only 14.6 months 7. On CT and magnetic resonance imaging (MRI) GBM usually appear as a heterogeneous tumor with central areas of necrosis, surrounded by thick irregular walls of solid, living neoplastic tissue. The gross tumor is often surrounded by extensive edema and it usually exerts considerable mass effect. So far several prognostic and predictive factors have been identified including age, performance status, extent of resection and biomarkers such as MGMT, EGFRvIII and IDH1. However, the value of imaging biomarkers such as radiomics has not yet fully been explored.
Radiomics can have an important role in the prediction of prognosis for patients with a GBM. As some patients only survive a few months, a subset of patients (10%) survives more than 5 years after diagnosis. Identification of these patients may benefit treatment decision by e.g. offering short-term survivors best-supportive care.
The investigators hypothesize that a combination of CT/MRI features, molecular biology and clinical data can provide an accurate prediction of medical outcome. The long term objective is to build a Decision Support System based on the predictive models established in this study.
An extensive dataset consisting of imaging, clinical, treatment data and outcomes of 360 patients treated in Maastricht since 2004 has been retrospectively collected. This includes 128 patients diagnosed with a biopsy only, with a tumor in situ on the planning CT. This dataset will be used to build predictive models of outcome (survival at 6- and 12 months). This analysis will be complemented by a radiomics study, analyzing both CT and MRI radiomics features.
In order to prove its value the signature will be validated on external datasets.
Study Design
Outcome Measures
Primary Outcome Measures
- Sensitive value of 6 months survival after radiotherapy with radiomics [6 months]
Secondary Outcome Measures
- Sensitive value of 12 months survival after radiotherapy with radiomics [12 months]
- Specificity value of 6 months survival after radiotherapy with radiomics [6 months]
- Specificity value of 12 months survival after radiotherapy with radiomics [12 months]
Eligibility Criteria
Criteria
Inclusion Criteria:
-
Histologically proven glioblastoma
-
Diagnosed with a biopsy only
-
Treated with curative intent
-
Required data available (clinical/radiological/radiotherapy structure set)
Exclusion Criteria:
Contacts and Locations
Locations
| Site | City | State | Country | Postal Code | |
|---|---|---|---|---|---|
| 1 | Maastricht Radiation Oncology (MAASTRO clinic) | Maastricht | Limburg | Netherlands | 6229ET |
Sponsors and Collaborators
- Maastricht Radiation Oncology
Investigators
None specified.Study Documents (Full-Text)
None provided.More Information
Additional Information:
Publications
- Aerts HJ, Velazquez ER, Leijenaar RT, Parmar C, Grossmann P, Carvalho S, Bussink J, Monshouwer R, Haibe-Kains B, Rietveld D, Hoebers F, Rietbergen MM, Leemans CR, Dekker A, Quackenbush J, Gillies RJ, Lambin P. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat Commun. 2014 Jun 3;5:4006. doi: 10.1038/ncomms5006. Erratum in: Nat Commun. 2014;5:4644. Cavalho, Sara [corrected to Carvalho, Sara].
- Carvalho S, Leijenaar RT, Velazquez ER, Oberije C, Parmar C, van Elmpt W, Reymen B, Troost EG, Oellers M, Dekker A, Gillies R, Aerts HJ, Lambin P. Prognostic value of metabolic metrics extracted from baseline positron emission tomography images in non-small cell lung cancer. Acta Oncol. 2013 Oct;52(7):1398-404. doi: 10.3109/0284186X.2013.812795. Epub 2013 Sep 9.
- Lambin P, Rios-Velazquez E, Leijenaar R, Carvalho S, van Stiphout RG, Granton P, Zegers CM, Gillies R, Boellard R, Dekker A, Aerts HJ. Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer. 2012 Mar;48(4):441-6. doi: 10.1016/j.ejca.2011.11.036. Epub 2012 Jan 16. Review.
- Leijenaar RT, Carvalho S, Velazquez ER, van Elmpt WJ, Parmar C, Hoekstra OS, Hoekstra CJ, Boellaard R, Dekker AL, Gillies RJ, Aerts HJ, Lambin P. Stability of FDG-PET Radiomics features: an integrated analysis of test-retest and inter-observer variability. Acta Oncol. 2013 Oct;52(7):1391-7. doi: 10.3109/0284186X.2013.812798. Epub 2013 Sep 9.
- Panth KM, Leijenaar RT, Carvalho S, Lieuwes NG, Yaromina A, Dubois L, Lambin P. Is there a causal relationship between genetic changes and radiomics-based image features? An in vivo preclinical experiment with doxycycline inducible GADD34 tumor cells. Radiother Oncol. 2015 Sep;116(3):462-6. doi: 10.1016/j.radonc.2015.06.013. Epub 2015 Jul 7.
- 15-43-05/08-intern