Synopsis
This lecture discusses the heuristic and quantitative methods for DCE-MRI data analysis, the clinical applications of both approaches in cancer imaging, the major factors that cause variabilities in the estimated heuristic metrics and quantitative pharmacokinetic parameters, and the need for standardization of data acquisition and analysis to improve reproducibility and repeatability and for consensus/guideline on whether heuristic or quantitative data analysis is the best-practice approach for a particular cancer imaging problem or topic.
Highlights
There are three approaches for analysis of DCE-MRI time-course data: qualitative, semi-quantitative, and quantitative. The first two approaches are generally referred as heuristic methods, while the quantitative approach is generally referred as pharmacokinetic (PK) modeling of data.
Applications of DCE-MRI in cancer care and research using heuristic and quantitative methods.
Major factors in data acquisition and analysis that cause variations in estimated heuristic and quantitative DCE-MRI parameters need to be taken into consideration when planning DCE-MRI trial in a multicenter setting.
Investigations are needed to provide consensus and guidelines on whether heuristic or quantitative data analysis is the best-practice for a particular cancer imaging problem or topic.
Target Audience
Investigators who have been using or plan to use dynamic contrast-enhanced (DCE) MRI for cancer imaging research.Outcome/Objectives
In addition to the basics of DCE-MRI and examples of heuristic and quantitative DCE-MRI methods used in cancer care and research, this lecture will discuss major factors in data acquisition and analysis that cause variations in estimated heuristic and quantitative DCE-MRI parameters, and steps that can be taken to potentially reduce variations and improve reproducibility in multicenter studies. The need to reach consensus on whether heuristic or quantitative DCE-MRI data analysis is the best-practice method for a particular cancer imaging problem or topic will also be discussed.Introduction
DCE-MRI involves serial acquisition of T1-weighted MR images of a tissue of interest (e.g., a tumor) before, during, and after an intravenous injection of paramagnetic contrast agent (CA). As the CA enters into the tissue from the blood, it changes the tissue’s T1 value and therefore the measured signal intensity. Malignant tumors tend to have elevated angiogenesis and the neo-vasculature is usually quite permeable to the gadolinium-based small CA molecules. This makes DCE-MRI an attractive, noninvasive and functional imaging method for characterizing tumor microvasculature. There are commonly three approaches for analyzing the DCE-MRI signal intensity time-course data. Qualitative description of the curve pattern of the time-course (such as washout, plateau, and persistence) is often used in standard of care clinical practice. Semi-quantitative analysis of the time-course data (such as uptake slope, washout slope, percent signal change, time to peak, and area under the curve) is used in clinical care and research, and has been adopted by many commercial CAD (Computer Aided Diagnosis) systems. These two data analysis approaches are often referred as heuristic methods. By fitting the DCE time-course data with a pharmacokinetic (PK) model, quantitative analysis allows extraction of quantitative parameters of tissue biology. The estimated quantitative PK parameters are usually variants of: Ktrans, a rate constant for passive CA transfer between plasma and interstitium - a measure of microvascular flow and/or permeability; ve, the extravascular and extracellular space volume fraction (the putative CA distribution volume fraction); and vp, the plasma volume fraction. kep, the CA intravasation rate constant, can be calculated as Ktrans/ve. Though quantitative PK analysis has the advantage of directly quantifying tissue biological properties and the estimated parameters are in principle independent of data acquisition details, the precision and accuracy of the PK parameters can be affected by several factors in data acquisition and analysis. On the other hand, heuristic analysis is simpler and more straightforward without the need for sophisticated mathematical modeling, and is readily implementable in clinical settings. However, heuristic parameters are directly related to MRI signal change, not tissue biology, and the metric values are generally dependent on data acquisition details including scanner platform, field strength, pulse sequence and parameters, CA dose and injection rate, and even personal skills, which often vary from one institution to another, resulting in high variability and low reproducibility.Clinical Applications in Cancer Imaging
In applications for cancer imaging, both the heuristic and quantitative DCE-MRI methods have been used for screening/detection, diagnosis, staging, therapy response evaluation, and assessment of residual disease and recurrence. There has been substantial literature evidence showing that both approaches can improve accuracy in cancer diagnosis, and provide earlier and better evaluation of cancer response to therapy than the standard of care approach of measuring imaging tumor size. Texture analysis of heuristic and quantitative DCE-MRI parameter spatial distributions has recently gained significant interest for its utility in characterizing tumor heterogeneity.Variability of Heuristic and Quantitative DCE-MRI Parameters
Heuristic DCE-MRI metrics are calculated directly from the signal time-course data without converting the signal time-course into T1 and CA concentration time-courses. The variability of these metrics is mostly caused by data acquisition details, such as flip angle, temporal resolution, etc. For quantitative PK analysis of DCE-MRI data, the signal time-course needs to be converted to T1 and/or CA concentration time-courses. The accuracy and precision of the derived PK parameters can be affected by several major factors in both data acquisition and analysis, including temporal resolution, acquisition time, quantifications of arterial input function (AIF) and native tissue T1, as well as selection of PK model and software package for data analysis. The variations in estimated PK parameters caused by uncertainties of these factors are, however, often systematic. Therefore, the utility of quantitative DCE-MRI for certain cancer imaging applications can be robust despite errors or variations in estimated PK parameter values. The major challenge in translating heuristic or quantitative DCE-MRI methods into multicenter trials or clinical practice is to find solutions to standardize data acquisition and data analysis and improve reproducibility and repeatability of the heuristic or quantitative DCE-MRI imaging biomarkers. Conclusion
DCE-MRI is a valuable noninvasive functional imaging method for tumor characterization. Both heuristic and quantitative approaches in data analysis have been shown useful in improving cancer detection and evaluation of therapeutic response. Both methods have advantages and disadvantages in data analysis, parameter variations, and results interpretation. Standardization in data acquisition and analysis is needed for either method to be adopted in multicenter trials or general clinical practice. Further investigations are needed to reach consensus or provide guidelines on whether heuristic or quantitative DCE-MRI is the best-practice approach for a particular cancer imaging problem or topic.Acknowledgements
NIH grants U01 CA154602, R44 CA180425References
Sourbron SP, Buckley DL. Tracer kinetic modelling in MRI: estimating perfusion and capillary permeability. Phys Med Biol 2012;57:R1-R33.
Khalifa F, Soliman A, El-Baz A, Abou El-Ghar M, El-Diasty T, Gimel'farb G, Ouseph R, Dwyer AC. Models and methods for analyzing DCE-MRI: a review. Med Phys 2014;41:124301.
Leach MO, Morgan B, Tofts PS, Buckley DL, Huang W, Horsfield MA, Chenevert TL, Collins DJ, Jackson A, Lomas D, Whitcher B, Clarke L, Plummer R, Judson I, Jones R, Alonzi R, Brunner T, Koh DM, Murphy P, Waterton JC, Parker G, Graves MJ, Scheenen T, Redpath T, Orton M, Karczmar G, Huisman H, Barentsz J, Padhani A. Imaging vascular function for early stage clinical trials using dynamic contrast-enhanced magnetic resonance imaging. Eur Radiol 2012;22:1451-1464.
Abramson RG, Li X, Hoyt TL, Su PF, Arlinghaus LR, Wilson KJ, Abramson VG, Chakravarthy AB, Yankeelov TE. Early assessment of breast cancer response to neoadjuvant chemotherapy by semi-quantitative analysis of high-temporal resolution DCE-MRI: preliminary results. Magn Reson Imaging 2013;31:1457-1464.
Huang W, Li X, Chen Y, Li X, Chang MC, Oborski MJ, Malyarenko DI, Muzi M, Jajamovich GH, Fedorov A, Tudorica A, Gupta SN, Laymon CM, Marro KI, Dyvorne HA, Miller JV, Barbodiak DP, Chenevert TL, Yankeelov TE, Mountz JM, Kinahan PE, Kikinis R, Taouli B, Fennessy F, Kalpathy-Cramer J. Variations of dynamic contrast-enhanced magnetic resonance imaging in evaluation of breast cancer therapy response: a multicenter data analysis challenge. Transl Oncol 2014;7:153-166.
Huang W, Chen Y, Fedorov A, Li X, Jajamovich GH, Malyarenko DI, Aryal MP, LaViolette PS, Oborski MJ, O’Sullivan F, Abramson RG, Jafari-Khouzani K, Afzal A, Tudorica A, Moloney B, Gupta SN, Besa C, Kalpathy-Cramer J, Mountz JM, Laymon CM, Muzi M, Kinahan PE, Schmainda K, Cao Y, Chenevert TL, Taouli B, Yankeelov TE, Fennessy FMM, Li X. The impact of arterial input function determination variations on prostate dynamic contrast-enhanced magnetic resonance imaging pharmacokinetic modeling: a multicenter data analysis challenge. Tomography 2016;2:56-66.