Radiogenomics: Creating a link between molecular diagnostics and diagnostic imaging

doi:10.1016/j.ejrad.2009.01.050

European Journal of Radiology

Volume 70, Issue 2, May 2009, Pages 232-241

https://doi.org/10.1016/j.ejrad.2009.01.050 Get rights and content

Abstract

Studies employing high-throughput biological techniques have recently contributed to an improved characterization of human cancers, allowing for novel sub-classification, better diagnostic accuracy, and more precise prognostication. However, requirement of surgical procurement of tissue among other things limits the clinical application of such methods in everyday patient care. Radiographic imaging is routine in clinical practice but is currently histopathology based. The use of routine radiographic imaging provides a potential platform for linking specific imaging traits with specific gene expression patterns that inform the underlying cellular pathophysiology; imaging features could then serve as molecular surrogates that contribute to the diagnosis, prognosis, and likely gene-expression-associated treatment response of various forms of human cancer. This review focuses on high-throughput methods such as microarray analysis of gene expression, their role in cancer research, and in particular, on novel methods of associating gene expression patterns with radiographic imaging phenotypes, known as “radiogenomics.” These findings underline a potential future role of both diagnostic and interventional radiologists in genetic assessment of cancer patients with radiographic imaging studies.

Introduction

With the relatively recent advent of high-throughput biological methods, biomedical research has seen many changes. High-throughput tools are able to provide a global snapshot of cellular physiology, allowing massively parallel assessment of thousands of genes and gene products. Gene expression profiling of various human tissues has led to a better understanding of cellular pathways and various pathological conditions on a biomolecular level. Analysis of different cancerous tissues in relation to samples of normal organ tissue has permitted an enhanced understanding of tumorogenic processes and aided in improved staging and sub-classification of various malignancies. Numerous studies have also shown that gene expression signatures, each comprised of dozens to hundreds of genes, can significantly improve diagnostic classification, prognostication, and prediction of therapeutic response in cancer [1], [2], [3], [4], [5], [6].

However, gene expression profiling is dependent on the surgical procurement of tissue, yielding a host of risks and potential complications, and making it an unrealistic option for every cancer patient. In contrast to genetic profiling studies, radiographic imaging studies are part of routine clinical care. While it is accepted that imaging can provide important anatomical and morphological information, it is not perceived to imply much consequential molecular detail. However, several studies have recently introduced novel methods for correlating unique features of tumor morphology and physiology garnered through non-invasive imaging with specific patterns of gene expression on a genome-wide scale, thereby introducing the field of “radiogenomics” [7], [8], [9]. These researchers propose that specific radiological tumor phenotypes, or “radiophenotypes,” can serve as surrogates for gene expression signatures, informing a non-invasive yet accurate diagnosis of tumor subtype and molecular biology. Further, if imaging traits can be associated with previously determined treatment-response gene-expression patterns, then non-invasive, routine clinical studies can inform the likely response to specific chemotherapeutics and aid in the decision making process towards an optimal form and duration of treatment. Such methods will potentially allow diagnostic imaging to augment or supplant current techniques that utilize surgical biopsy and pathohistologic analysis, and contribute to a shift towards a more fundamental genetic-based and personalized medicine. The following discussion will center on the promise of high-throughput biological techniques and the emerging field of radiogenomics—an emerging technology that has the potential to link high-throughput molecular diagnostics and diagnostic imaging.

Section snippets

High-throughput biology and gene expression profiling of cancer

With the monumental undertaking of the Human Genome Project largely completed, the accessibility of genomic information has over the last decade brought a prodigious transformation to biomedical research. Investigators are now faced with an array of new tools that allow for the simultaneous measurement of thousands of biological molecules. These high-throughput methods (which are in various stages of maturity) include many novel platforms for profiling each fundamental stage in the flow of

Diagnostic imaging as a surrogate for gene expression profiling

Growth in the clinical utility of imaging modalities to non-invasively characterize disease and guide clinical management in parallel with the development of functional genomic tools such as microarrays provides powerful methods for dissecting both the clinical phenotype in vivo and the molecular basis of disease on a genome-wide level. While gene expression profiling can reveal important information related to underlying biology and its relationship to diagnosis, prognosis, and optimal

Conclusion

Over the last decade, high-throughput biological techniques have reshaped the perspective of biomedical research, allowing efficient and rapid assessment of the entire biological and molecular topography of a cell's physiology. These tools have contributed to a better understanding of cellular pathways and the pathophysiology that leads to tumorogenesis. Gene expression profiling of malignant tissue samples has permitted the elucidation of tumor gene expression signatures across a variety of

References (43)

A. Hsiao et al.
High-throughput biology in the postgenomic era
J Vasc Interv Radiol
(2006)
A.A. Alizadeh et al.
Genomic-scale gene expression profiling of normal and malignant immune cells
Curr Opin Immunol
(2000)
J.A. Johnson et al.
Molecular diagnostics as a predictive tool: genetics of drug efficacy and toxicity
Trends Mol Med
(2002)
M.E. Halatsch et al.
Epidermal growth factor receptor inhibition for the treatment of glioblastoma multiforme and other malignant brain tumours
Cancer Treat Rev
(2006)
S. Paik et al.
Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer
J Clin Oncol
(2006)
C.H. Chung et al.
Molecular portraits and the family tree of cancer
Nat Genet
(2002)
E. Segal et al.
From signatures to models: understanding cancer using microarrays
Nat Genet
(2005)
X. Chen et al.
Gene expression patterns in human liver cancers
Mol Biol Cell
(2002)
H.Y. Chang et al.
Robustness, scalability, and integration of a wound-response gene expression signature in predicting breast cancer survival
Proc Natl Acad Sci U S A
(2005)
S. Paik et al.
A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer
N Engl J Med
(2004)

M.D. Kuo et al.

Radiogenomic analysis to identify imaging phenotypes associated with drug response gene expression programs in hepatocellular carcinoma

J Vasc Interv Radiol

(2007)

E. Segal et al.

Decoding global gene expression programs in liver cancer by noninvasive imaging

Nat Biotechnol

(2007)

M. Diehn et al.

Identification of noninvasive imaging surrogates for brain tumor gene-expression modules

Proc Natl Acad Sci U S A

(2008)

S.P. Fodor et al.

Multiplexed biochemical assays with biological chips

Nature

(1993)

M. Schena et al.

Quantitative monitoring of gene expression patterns with a complementary DNA microarray

Science

(1995)

T.R. Golub et al.

Molecular classification of cancer: class discovery and class prediction by gene expression monitoring

Science

(1999)

A.A. Alizadeh et al.

Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling

Nature

(2000)

J.S. Lee et al.

Classification and prediction of survival in hepatocellular carcinoma by gene expression profiling

Hepatology

(2004)

T. Sorlie et al.

Repeated observation of breast tumor subtypes in independent gene expression data sets

Proc Natl Acad Sci U S A

(2003)

Y. Liang et al.

Gene expression profiling reveals molecularly and clinically distinct subtypes of glioblastoma multiforme

Proc Natl Acad Sci U S A

(2005)

C.M. Perou et al.

Molecular portraits of human breast tumours

Nature

(2000)

Cited by (255)

Imaging Genomics and Multiomics: A Guide for Beginners Starting Radiomics-Based Research
2024, Academic Radiology
Radiomics uses advanced mathematical analysis of pixel-level information from radiologic images to extract existing information in traditional imaging algorithms. It is intended to find imaging biomarkers related to the genomics of tumors or disease patterns that improve medical care by advanced detection of tumor response patterns in tumors and to assess prognosis. Radiomics expands the paradigm of medical imaging to help with diagnosis, management of diseases and prognostication, leveraging image features by extracting information that can be used as imaging biomarkers to predict prognosis and response to treatment. Radiogenomics is an emerging area in radiomics that investigates the association between imaging characteristics and gene expression profiles. There are an increasing number of research publications using different radiomics approaches without a clear consensus on which method works best. We aim to describe the workflow of radiomics along with a guide of what to expect when starting a radiomics-based research project.
Accurate non-invasive grading of nonfunctional pancreatic neuroendocrine tumors with a CT derived radiomics signature
2024, Diagnostic and Interventional Imaging
The purpose of this study was to develop a radiomics-signature using computed tomography (CT) data for the preoperative prediction of grade of nonfunctional pancreatic neuroendocrine tumors (NF-PNETs).
A retrospective study was performed on patients undergoing resection for NF-PNETs between 2010 and 2019. A total of 2436 radiomic features were extracted from arterial and venous phases of pancreas-protocol CT examinations. Radiomic features that were associated with final pathologic grade observed in the surgical specimens were subjected to joint mutual information maximization for hierarchical feature selection and the development of the radiomic-signature. Youden-index was used to identify optimal cutoff for determining tumor grade. A random forest prediction model was trained and validated internally. The performance of this tool in predicting tumor grade was compared to that of EUS-FNA sampling that was used as the standard of reference.
A total of 270 patients were included and a fusion radiomic-signature based on 10 selected features was developed using the development cohort (n = 201). There were 149 men and 121 women with a mean age of 59.4 ± 12.3 (standard deviation) years (range: 23.3–85.0 years). Upon internal validation in a new set of 69 patients, a strong discrimination was observed with an area under the curve (AUC) of 0.80 (95% confidence interval [CI]: 0.71–0.90) with corresponding sensitivity and specificity of 87.5% (95% CI: 79.7–95.3) and 73.3% (95% CI: 62.9–83.8) respectively. Of the study population, 143 patients (52.9%) underwent EUS-FNA. Biopsies were non-diagnostic in 26 patients (18.2%) and could not be graded due to insufficient sample in 42 patients (29.4%). In the cohort of 75 patients (52.4%) in whom biopsies were graded the radiomic-signature demonstrated not different AUC as compared to EUS-FNA (AUC: 0.69 vs. 0.67; P = 0.723), however greater sensitivity (i.e., ability to accurately identify G2/3 lesion was observed (80.8% vs. 42.3%; P < 0.001).
Non-invasive assessment of tumor grade in patients with PNETs using the proposed radiomic-signature demonstrated high accuracy. Prospective validation and optimization could overcome the commonly experienced diagnostic uncertainty in the assessment of tumor grade in patients with PNETs and could facilitate clinical decision-making.
The application of machine learning and deep learning radiomics in the treatment of esophageal cancer
2023, Radiation Medicine and Protection
Esophageal cancer (EC) is a very aggressive disease with most cases diagnosed at advanced stages. Early detection and prognosis prediction are of clinical significance in the optimal management of EC. Genomic and proteomic technologies demonstrated limited efficacy due to the invasive nature and the inherent tumor heterogeneity. Non-invasive radiomics has achieved significant results in tumor characterization, treatment response and survival prediction for various cancers. In this article, the current application of both machine learning and deep learning based radiomics in the diagnosis, prognostic prediction and treatment outcome prediction for patients with EC were reviewed. The current challenges and prospects for the future application of radiomics in EC were also discussed.
Prediction of Parkinson's disease pathogenic variants using hybrid Machine learning systems and radiomic features
2023, Physica Medica
In Parkinson’s disease (PD), 5–10% of cases are of genetic origin with mutations identified in several genes such as leucine-rich repeat kinase 2 (LRRK2) and glucocerebrosidase (GBA). We aim to predict these two gene mutations using hybrid machine learning systems (HMLS), via imaging and non-imaging data, with the long-term goal to predict conversion to active disease.
We studied 264 and 129 patients with known LRRK2 and GBA mutations status from PPMI database. Each dataset includes 513 features such as clinical features (CFs), conventional imaging features (CIFs) and radiomic features (RFs) extracted from DAT-SPECT images. Features, normalized by Z-score, were univariately analyzed for statistical significance by the t-test and chi-square test, adjusted by Benjamini-Hochberg correction. Multiple HMLSs, including 11 features extraction (FEA) or 10 features selection algorithms (FSA) linked with 21 classifiers were utilized. We also employed Ensemble Voting (EV) to classify the genes.
For prediction of LRRK2 mutation status, a number of HMLSs resulted in accuracies of 0.98 ± 0.02 and 1.00 in 5-fold cross-validation (80% out of total data points) and external testing (remaining 20%), respectively. For predicting GBA mutation status, multiple HMLSs resulted in high accuracies of 0.90 ± 0.08 and 0.96 in 5-fold cross-validation and external testing, respectively. We additionally showed that SPECT-based RFs added value to the specific prediction of of GBA mutation status.
We demonstrated that combining medical information with SPECT-based imaging features, and optimal utilization of HMLS can produce excellent prediction of the mutations status in PD patients.
Preoperative CT-based deep learning radiomics model to predict lymph node metastasis and patient prognosis in bladder cancer: a two-center study
2024, Insights into Imaging
Methodology for Good Machine Learning with Multi-Omics Data
2024, Clinical Pharmacology and Therapeutics

View all citing articles on Scopus

View full text

Radiogenomics: Creating a link between molecular diagnostics and diagnostic imaging

Abstract

Introduction

Section snippets

High-throughput biology and gene expression profiling of cancer

Diagnostic imaging as a surrogate for gene expression profiling

Conclusion

J Vasc Interv Radiol

Curr Opin Immunol

Trends Mol Med

Cancer Treat Rev

Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer

J Clin Oncol

Molecular portraits and the family tree of cancer

Nat Genet

From signatures to models: understanding cancer using microarrays

Nat Genet

Gene expression patterns in human liver cancers

Mol Biol Cell

Robustness, scalability, and integration of a wound-response gene expression signature in predicting breast cancer survival

Proc Natl Acad Sci U S A

A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer

N Engl J Med

Radiogenomic analysis to identify imaging phenotypes associated with drug response gene expression programs in hepatocellular carcinoma

J Vasc Interv Radiol

Decoding global gene expression programs in liver cancer by noninvasive imaging

Nat Biotechnol

Identification of noninvasive imaging surrogates for brain tumor gene-expression modules

Proc Natl Acad Sci U S A

Multiplexed biochemical assays with biological chips

Nature

Quantitative monitoring of gene expression patterns with a complementary DNA microarray

Science

Molecular classification of cancer: class discovery and class prediction by gene expression monitoring

Science

Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling

Nature

Classification and prediction of survival in hepatocellular carcinoma by gene expression profiling

Hepatology

Repeated observation of breast tumor subtypes in independent gene expression data sets

Proc Natl Acad Sci U S A

Gene expression profiling reveals molecularly and clinically distinct subtypes of glioblastoma multiforme

Proc Natl Acad Sci U S A

Molecular portraits of human breast tumours

Nature