OR WAIT null SECS
The use of imaging in trials shows great promise, but proper application requires early dialogue.
Imaging is widely applied as a biomarker in therapeutic clinical trials. It is commonly used to assess the response to therapeutic interventions. However, the wealth of imaging modalities and techniques applicable for different diseases often make it difficult for the clinical trial manager to choose the technique best suited for the underlying question.
This article describes the opportunities and challenges for the use of imaging in clinical trials from an imager's perspective—one who has experience as a principal investigator, co-investigator, and core lab scientist.
The most common use of imaging in clinical trials today is in the evaluation of tumor size. The two most commonly used evaluation criteria are the World Health Organization (WHO)1,2 criteria and the Response Evaluation Criteria in Solid Tumors (RECIST)3 (see Figure 1).
Both assessments are based on the measurement of one or two diameters of selected target lesions. Up to a maximum of five lesions per organ and 10 lesions in total, representative of all involved organs, are identified as target lesions. The sum of the longest diameter for all target lesions is calculated. The response to treatment is classified by the following criteria:
These evaluation criteria have been developed to estimate volume changes of lesions in response to therapy. This was necessary because the assessment of the true lesion volume was very time consuming, since the lesions had to be outlined on every slice manually. Therefore, it was not applicable in clinical routine. However, the assessment of diameters also has several drawbacks.
Its biggest limitation is its inaccuracy. As a result, the assessed changes have to be large in order to be considered truly significant volume changes. For example, a 30% diameter decrease suggesting partial response is equal to a 65% volume decrease. Therefore, the assessment of the true three-dimensional volume of all lesions is a much more sensitive response criteria.
Computer algorithms have been developed to semi-automatically assess the tumor volume and are used in some clinical trials4,5 (see Figure 1). These algorithms have to be implemented in clinical routine. For clinical trials, the implementation of such semi-automatic algorithms would mean that tumor response may be detected in a much earlier stage and, therefore, success or failure of a treatment may be shown in a much shorter time frame. This would lower the costs of clinical trials significantly because the need for follow-up examinations would be reduced due to the earlier achievement of study endpoints.
Figure 1. Images of a male patient with liver metastases from colorectal cancer. The target lesion (arrow) (a) is measured according to WHO criteria by measuring its longest diameter and the according longest orthogonal diameter. A semi-automatic software segmented the tumor three-dimensionally (b), displaying three orthogonal planes.
Rapid scientific progress in imaging methodologies seems to be encouraging clinical researchers to use these methodologies to address their needs in clinical trials. Highly specialized techniques such as diffusion tensor imaging,6 neurofunctional magnetic resonance imaging (fMRI),7 and dynamic contrast enhanced magnetic resonance imaging (DCE MRI)8,9 seem to be the answers to researchers' unaddressed needs.
However, as the complexity of these imaging modalities increases, only highly specialized imaging experts will have the expertise to understand the underlying medical paradigms. Clinical trial sites having the desired patient populations frequently do not have that expertise on site. Moreover, other clinical realities prevent the application of such innovative techniques in routine clinical trials, including:
Therefore, clinical managers need to resist the urge to encourage imaging researchers to implement new, highly innovative methodologies in their therapeutic trials. Only in clinical routine should well-established, standardized, readily available methodologies be applied. Examples of such techniques include lesion size assessments using WHO or RECIST criteria and volume measurements, standardized DCE MRI10,11 or diffusion weighted MRI (DW MRI).12
Clinical trials should not be abused to assess innovative imaging techniques, such as diffusion tensor imaging or neurofunctional MRI. This introduces potential measurement inaccuracies into the trial results and may lead to confusion about the response to therapy due to conflicting results of the new techniques. Moreover, such new and innovative methodologies that have not been fully evaluated and standardized may introduce measurement errors that are not yet fully understood.
These techniques are, however, appealing to clinical investigators because they may be able to address the heterogeneity of disease, especially in the assessment of malignant diseases. While volume measurements only give you a global assessment of the disease, new and innovative techniques may assess therapy on a sublesion level.
Today's reality in clinical trials is that the pharmaceutical industry develops study protocols in conjunction with visible experts in the clinical field of the study. Together, they devise a study protocol that entails imaging as a biomarker as the clinical experts know them in their clinical routine. Regularly, none of the involved parties has the expertise in imaging that would determine the needs of a desired imaging technique.
Is MRI better than CT for size measurements of brain metastases? Is CT better than MRI for the assessment of lung tumors? Is there a possibility to do dynamic contrast enhanced MRIs on liver metastases? What are the technical requirements to use PET/CT as an imaging biomarker? These are only a few questions that may need to be addressed in the planing of a clinical trial involving imaging biomarkers. Usually, by the time an imager sees the protocol, it is already at a stage when it is too late to address such questions.
In the local setting, the local coprincipal investigator often does not involve its local radiologist until the study protocol is approved and the first patients are scheduled. Here, imaging is usually considered a utility that can be ordered like in clinical routine. However, it usually has more complexity than a routine clinical scan. Consider the following examples:
Reimbursement. This may be the most unaddressed issue. Most clinical investigators do not even consider reimbursement for imaging biomarkers. But while health insurance pays for the costs of imaging in clinical routine exams, trial exams need to be paid by the sponsor. This might not be a big concern of the imager, but it sure is for the hospital administration. However, even if the clinical investigator plans for imaging in their budget, he or she commonly does not know about additional costs for trial related procedures, such as data transmittal to the sponsor, documentation, data storage, and other trial-related requirements.
Imaging equipment. In a multicenter clinical trial, the available equipment varies between sites. Especially for nonroutine imaging biomarkers, it is very likely that not all sites can perform these imaging studies. Therefore, it is mandatory that local imagers are involved before the beginning of the trial.
Training. Local imaging personnel have to be trained in order to perform the imaging according to protocol and to acquire highly standardized images that allow a comparison throughout the study. This is not only true for highly specialized imaging procedures, but also for routine imaging.
A mechanism to ensure the appropriate communication with the local imagers is to contract with them directly instead of placing imaging costs into the budget of the local principal investigator (PI). This would prevent almost all of the issues previously described and would enable the imagers to provide the best services possible. In addition, it would save the trial sponsor money. First of all, the imaging biomarkers would be acquired in a standardized way that is suitable for the trials purpose from all sites. Second, nonqualified sites could be excluded prior to the beginning of the trial. Third, proper documentation and image data transfer is ensured.
Appropriate training and documentation of the trial procedures and results are an integral part of a clinical trial. This is especially true when it comes to imaging biomarkers.
An imaging site manual is mandatory for the success of the trial. This site manual is not just a copy of the trial protocol, but it is tailored toward imaging. Its main purpose is to describe the image acquisition and all related procedures. It needs to have a list of all important onsite and offsite contacts, including the sponsor, core lab, local PI, and imaging personnel involved. Moreover, the imaging procedure needs to detailed.
In addition to the acquisition protocol, it describes patient handling, data storage and transfer, the necessary equipment, and any additional procedures. An imaging site manual allows all involved imaging personnel to be informed about the trial-related imaging procedures, and therefore ensures the quality of the image acquisition. It needs to be to the point, and should not be overburdened with nonimaging information.
All involved personnel need to be trained in the imaging procedures. The best way to accomplish this is a pretrial site visit that is attended by all imagers. During the visit, a volunteer or a phantom (especially if imaging involves ionizing radiation) should be scanned according to the imaging protocol. The acquired data should be stored and transferred to the imaging core lab, and the appropriate documentation completed.
From the perspective of an imager, clinical therapeutic trials are currently often a burden instead of an opportunity. However, this does not have to be the case. Early involvement, sufficient documentation and training, adequate reimbursement, and the use of appropriate techniques will allow the imager to be an integral part of the study team, and will therefore ensure high-quality results, the cost-effective use of imaging, and the best possible outcomes for all involved parties.
Johannes T. Heverhagen,* MD, PhD, works in the department of diagnostic radiology at Philipps University Marburg, University Hospital Giessen and Marburg GmbH, Baldinger Strasse, 35033 Marburg, Germany, email: firstname.lastname@example.orgMichael V. Knopp, MD, PhD, works in the department of radiology at Ohio State University.
*To whom all correspondence should be addressed.
1. WHO handbook for reporting results of cancer treatment. Geneva (Switzerland): World Health Organization Offset Publication, 1979.
2. A.B. Miller, B. Hoogstraten, M. Staquet, A. Winkler, "Reporting Results of Cancer Treatment," Cancer, 47, 207–214 (1981).
3. P. Therasse, S.G. Arbuck, E.A. Eisenhauer et al., "New Guidelines to Evaluate the Response to Treatment in Solid Tumors," Journal of the National Cancer Institute, 92, 205–216 (2000).
4. J.M. Kuhnigk, V. Dicken, L. Bornemann et al., "Morphological Segmentation and Partial Volume Analysis for Volumetry of Solid Pulmonary Lesions in Thoracic CT Scans," IEEE Trans Med Imaging, 25, 417–434 (2006).
5. L. Bornemann, J.M. Kuhnigk, V. Dicken et al., "Informatics in Radiology (infoRAD): New Tools for Computer Assistance in Thoracic CT Part 2. Therapy Monitoring of Pulmonary Metastases," Radiographics, 25, 841–848 (2005).
6. V. Gulani and P.C. Sundgren, "Diffusion Tensor Magnetic Resonance Imaging," Journal of Neuroophthalmology, 26, 51–60 (2006).
7. W.T. Yuh, G.A. Christoforidis, R.M. Koch et al., "Clinical Magnetic Resonance Imaging of Brain Tumors at Ultrahigh Field: A State-of-the-Art Review," Top Magn Reson Imaging, 17, 53–61 (2006).
8. G. Jia, J.T. Heverhagen, H. Henry et al., "Pharmacokinetic Parameters as a Potential Predictor of Response to Pharmacotherapy in Benign Prostatic Hyperplasia: A Preclinical Trial using Dynamic Contrast-Enhanced MRI," Journal of Magnetic Resonance Imaging, 24, 721–725 (2006).
9. B.J. Woods, B.D. Clymer, T. Kurc et al., "Malignant-lesion Segmentation Using 4D Co-occurrence Texture Analysis Applied to Dynamic Contrast-enhanced Magnetic Resonance Breast Image Data," Journal of Magnetic Resonance Imaging, 25, 495–501 (2007).
10. P.L. Choyke, A.J. Dwyer, M.V. Knopp, "Functional Tumor Imaging with Dynamic Contrast-enhanced Magnetic Resonance Imaging," Journal of Magnetic Resonance Imaging, 17, 509–520 (2003).
11. M.V. Knopp, F.L. Giesel, H. Marcos et al., "Dynamic Contrast-enhanced Magnetic Resonance Imaging in Oncology," Top Magn Reson Imaging, 12, 301–308 (2001).
12. H.B. Newton, A. Ray-Chaudhury, R. Cavaliere, "Brain Tumor Imaging and Cancer Management: The Neuro-oncologists Perspective," Top Magn Reson Imaging, 17, 127–136 (2006).