Radiology in Oncology Trials: Critical Success Factors

Modern imaging technologies are being used with increasing frequency during evaluation of oncology drugs in clinical trials. Cross-sectional radiological imaging, particularly computed tomography and magnetic resonance imaging, has established a central role in the management of cancer patients.^1,2 Because imaging is very reliable for detection and characterization of cancers, it can serve to guide therapy by noninvasively monitoring treatment response.

Drug-regulating agencies such as the Food and Drug Administration (FDA) are willing to provide expedited approval of cancer drugs developed for treatment of cancers for which no adequate alternative is available.³ Because imaging can provide objective evidence of tumor shrinkage, it can help rapidly establish drug efficacy by supplementing subjective clinical endpoints such as quality of life. For example, FDA approved the anti-cancer drug capecitabine following an image-based oncology trial comprising only 162 subjects.⁴

This article identifies the critical success factors that allow effective use of radiology in the clinical evaluation of oncology drugs.

Role of radiology
Recent and ongoing technological advances in imaging technology and diagnostic contrast media have revolutionized the role of radiology in clinical practice. The technical advances include not only faster volumetric imaging (multislice computed tomography, or CT, and 3D magnetic resonance imaging, or MRI), but also functional imaging (positron emission tomography, or PET). Very accurate noninvasive monitoring of tumor burden is made possible by the ability to obtain high-quality, thin-slice, modern-day CT and MRI images in a single breath-hold using newer tissue-specific contrast media followed by superior image processing.⁵⁸ Radiologists can now evaluate tumor size with a high degree of precision, and can readily obtain tumor volumes. They can accurately quantify treatment response by comparing pretreatment and posttreatment measurements of tumor burden.

Radiological methods may not be appropriate, however, in certain cancers. For example, imaging has a very limited role in patients with leukemia and other hematological malignancies. Blood tests and other lab investigations may be more appropriate in such contexts. Thus oncologists need to continually evaluate the potential role of radiology in determining tumor burden, including determining which imaging method would be best used to estimate tumor burden in light of the rapid evolution of imaging technology.

A bewildering array of imaging modalities are available for potential use in drug trials. Plain radiography, sonography, CT, MRI, and radionuclide studies are now widely available. Recent advances in CT, MRI, and PET have completely revolutionized imaging, opening up many potential applications.^6,7,9 Faster imaging, thinner slice-profile, better 3-D reconstruction, and perfusion mapping are some of the advantages of the current helical CT technology.^5,6 Advances in MR technology such as spectroscopy, perfusion imaging, and diffusion imaging provide important physiological and biochemical data in addition to the anatomical information.¹⁰¹²

It is now possible to image changes in tumor perfusion and metabolism and hence better evaluate treatment response before a change in tumor size takes place. Diffusion MRI, MR spectroscopy, perfusion CT and MRI, and molecular imaging detect changes in tumor perfusion and metabolism earlier than the conventional radiological techniques.¹⁰¹² The principal investigator should select the appropriate imaging modality to be used before initiation of the study.For example, whereas tumor size surveillance is appropriate in evaluating the efficacy of cytotoxic drugs, perfusion studies may be ideally suited for evaluating tumors treated with anti-angiogenesis drugs.¹³¹⁶ The efficacy of tumor vaccines and gene-based drugs is best analyzed by molecular imaging strategies.^17,18 The radiologists input in this context is invaluable.

Uniform imaging protocolsA wide variation exists in the type of scanners, scanning parameters, and scanning techniques used, providing varying results. For example, radiology literature documents that imaging of the liver during the equilibrium phase may result in suboptimal visualization of lesions.¹⁹

Variation in imaging practices is a major disadvantage of multicenter clinical trials, potentially giving rise to confounding results. In a multicenter drug trial, it is imperative to choose sites that can support optimal and uniform imaging techniques. The adoption of a standard set of diagnostic imaging guidelines improves the value of imaging when used to measure tumor burden.

Electronic image data management
The use of Internet-based technologies in clinical trials has increased dramatically in a number of areas: electronic capture and transmission of clinical trials data; recruitment process by all trial participants, including subjects and investigators; and filing of new drug approval applications with government regulatory agencies.

By taking advantage of the network architecture that exists for the Internet, many trials can benefit from image transfer methods and Internet topology already in place.The future standard of managing medical imaging is envisioned as a combination of several key technological components and processes in a well-designed image file transfer architecture. This architecture, through the extensive use of the Internet, will significantly address the traditional problem areas of applied clinical trials. Most important in trials is the collaboration of the equipment vendors, the pharmaceutical companies, and the image management companies to provide a technical solution that is cost-effective, technologically robust, and globally accessible.

With rapid developments in the field of telemedicine and picture archival and communication systems (PACS), many radiology departments are increasingly becoming filmless.²⁰ The advantages of reading directly from the monitor are numerous and include the ability to change the image characteristics by altering window settings, easy image transfer, and the ability to superimpose different imaging modalities of different data content. These abilities enhance the performance of the radiologist. Other advantages include accurate measurement using digital calipers, use of digital overlays, and the potential use of automated analysis by sophisticated image-analysis software tools.²¹

In addition, considerable time savings can be achieved through secure transmission of digital images via the Internet. Electronic transfer of data is also more economical than shipping hard copies.

Off-site image analysisConventionally, clinical investigators at the clinical site where subjects were imaged (that is, on-site) typically interpreted the images. Both drug regulatory agencies and pharmaceutical companies now often opt for independent evaluation of the images by a select group of expert physicians (off-site) who are blinded to the clinical data. This practice potentially reduces bias and interobserver variation in data analysis.²²

Involving subspecialty-trained radiologists in image interpretation is equally critical for optimal results. Precise tumor characterization and accurate tumor delineation are crucial factors in interpreting images. For example, surveillance of subjects with brain tumors is best served by neuroradiologists who have considerable experience and expertise in interpreting brain MRIs. On the other hand, gastrointestinal radiologists are better suited to interpret the CT images of subjects with colon cancer.

In addition, regulatory agencies mandate unbiased evaluation of images at randomized visits that are best done in off-site locations. Several studies indicate that off-site evaluation of studies improves consistency, reliability, accuracy, and reproducibility of readings.²² As a result, the image analyses in image-based oncology drug trials are increasingly performed in off-site core laboratories.

Novel tumor measurement techniques
Serial evaluation of tumor size following treatment provides objective evidence of drug efficacy. For more than two decades, the World Health Organization (WHO) criteria for categorizing patient response were used.²³ The WHO (bidimensional) technique consists of measuring each tumor twicetwo mutually perpendicular diametersand obtaining the product, or cross-product, of these diameters. The cross-products are then compared to categorize patient response.

Advances in imaging technology largely prompted an international working group to set new guidelines to evaluate patient response. The new recommendations, referred to as response evaluation criteria in solid tumors, or RECIST, simplified the measurement technique and advocated the use of unidimensional measurements.²⁴ The differences in the two techniques are summarized in Table 1.

Since most tumors are spherical, criteria based on diameter, cross-sectional area, and volume are interchangeable.²⁵ In a subset of tumors with asymmetrical tumor shrinkage or growth or non-ellipsoidal morphology, discrepancies between the two techniques are likely to set in. With advances in imaging technology that permit tumoral volumetric computations, criteria based on volume may be expected.

Interim data analysis
Investigators should continually review and update the methods used in data analysis, if needed. Continual review permits early detection of flaws in the methods and the opportunity to take corrective steps. Keeping the evaluation criteria as objective as possible can reduce interobserver variation.

Imaging can speed trials
Modern-day radiologic imaging can provide highly accurate, reproducible measures of viable tumor burden. Subspecialty radiologist analysis in an electronic environment provides a very thorough image analysis and a faster and more cost-effective means of evaluating drug efficacy.

References

1. M. Hasegawa, S. Sone, and S. Takashima, et al., Growth Rate of Small Lung Cancers Detected on Mass CT Screening, British Journal of Radiology, 73 (876) 12521259 (2000).

2. L.E. Hann, C.B. Winston, K.T. Brown, and T. Akhurst, Diagnostic Imaging Approaches and Relationship to Hepatobiliary Cancer Staging and Therapy, Seminars in Surgical Oncology, 19 (2) 94115 (2000).

3. Food and Drug Administration, Guidance for Industry, FDA Approval of New Cancer Treatment Uses for Marketed Drug and Biological Products (U.S. Department of Health and Human Services, FDA Center for Drug Evaluation and Research, Center for Biologics Evaluation and Research, December 1998, Clin 7).

4. J.L. Blum, S.E. Jones, and A.U. Buzdar, et al., Multicenter Phase II Study of Capecitabine in Paclitaxel-Refractory Metastatic Breast Cancer, Journal of Clinical Oncology, 17 (2) 485493 (1999).

5. K. Klingenbeck-Regn, S. Schaller, T. Flohr, B. Ohnesorge, A.F. Kopp, and U. Baum, Subsecond Multi-slice Computed Tomography: Basics and Applications, European Journal of Radiology, 31 (2) 110124 (1999).

6. H. Hu, H.D. He, W.D. Foley, and S.H. Fox, Four Multidetector-row Helical CT: Image Quality and Volume Coverage Speed, Radiology, 215 (1) 5562 (2000).

7. M. Poustchi-Amin and J.J. Brown, Echo-planar MRI Allows Study of Fast-changing Physiologic Processes, Diagnostic Imaging (San Francisco), 22 (4) 3539 (2000).

8. J.P. Earls and D.A. Bluemke, New MR Imaging Contrast Agents, Magnetic Resonance Imaging Clinics of North America, 7 (2) 255273 (1999).

9. U. Roelcke and K.L. Leenders, Positron Emission Tomography in Patients with Primary CNS Lymphomas, Journal of Neurooncology, 43 (3) 231236 (1999).

10. J.C. Wong, J.M. Provenzale, and J.R. Petrella, Perfusion MR Imaging of Brain Neoplasms, American Journal of Roentgenology, 174: 11471157 (2000).

11. R. Brasch and K. Turetschek, MRI Characterization of Tumors and Grading Angiogenesis Using Macromolecular Contrast Media: Status Report, European Journal of Radiology, 34 (3) 148155 (2000).

12. T.L. Chenevert, L.D. Stegman, and J.M. Taylor, et al., Diffusion Magnetic Resonance Imaging: An Early Surrogate Marker of Therapeutic Efficacy in Brain Tumors, Journal of National Cancer Institute, 92 (24) 20292036 (2000).

13. C. Harvey, A. Dooher, J. Morgan, M. Blomley, and P. Dawson, Imaging of Tumour Therapy Responses by Dynamic CT, European Journal of Radiology, 30 (3) 221226 (1999).

14. J.M. Cherrington, L.M. Strawn, and L.K. Shawver, New Paradigms for the Treatment of Cancer: The Role of Anti-Angiogenesis Agents, Advances in Cancer Research, 79: 138 (2000).

15. T. Mikkelsen, Cytostatic Agents in the Management of Malignant Gliomas, Cancer Control, 5 (2) 150162 (1998).

16. J.L. Blum, Xeloda in the Treatment of Metastatic Breast Cancer, Oncology, 57 (Supp. 1): 1620 (1999).

17. C. Bremer and R. Weissleder, In Vivo Imaging of Gene Expression, Academic Radiology, 8: 1523 (2001).

18. P. Wunderbaldinger, A. Bogdanov, and R. Weissleder, New Approaches for Imaging in Gene Therapy, European Journal of Radiology, 34 (3) 156165 (2000).

19. P. Dawson and J. Morgan, The Meaning and Significance of the Equilibrium Phase in Enhanced Computed Tomography of the Liver, British Journal of Radiology, 72 (857) 438442 (1999).

20. R.L. Arenson, K.P. Andriole, D.E. Avrin, and R.G. Gould, Computers in Imaging and Health Care: Now and in the Future, Journal of Digital Imaging, 13 (4) 145156 (2000).

21. L.H. Schwartz, M.S. Ginsberg, and D. DeCorato, et al., Evaluation of Tumor Measurements in Oncology: Use of Film-based and Electronic Techniques, Journal of Clinical Oncology, 18 (10) 21792184 (2000).

22. D. Sahani, S. Saini, and G.A. Fatuga, et al., Quantitative Measurements of Medical Images for Pharmaceutical Clinical Trials: Comparison between On-site and Off-site Assessments, American Journal of Roentgenology, 174 (4) 11591162 (2000).

23. A.B. Miller, B. Hoogstraten, M. Staquet, and A. Winkler, Reporting Results of Cancer Treatment, Cancer, 47: 207214 (1981).

24. P. Therasse, S.G. Arbuk, and E.A. Eisenhauer, et al., New Guidelines to Evaluate Response to Treatment in Solid Tumors, Journal of National Cancer Institute, 92: 205216 (2000).

25. S. Saini, Radiologic Measurement of Tumor Size in Clinical Trials: Past, Present, and Future, American Journal of Roentgenology, 176: 333334 (2001).