Biomarkers initiate major shift in imaging research

Successful radiological research requires correlation with a gold standard that is accepted as the best diagnostic test or optimum method for treatment assessment. This may involve analysis of a pathologic specimen or another invasive procedure. Such standards are, however, associated with long observation times or a need for surgical exploration. Using the five-year survival rate, a relatively long time in the setting of a cancer therapy, would prevent many patients from benefiting from a potentially successful treatment option.

It typically takes up to 12 years to develop a new drug and can cost hundreds of millions of euros to gain marketing approval in the U.S., Europe, and elsewhere. Research may be further limited by ethical restrictions on the use of highly invasive tests, such as the painful iliac crest biopsy for histologic workup of osteopathies.

The desire to implement research results more effectively and efficiently in clinical practice has driven attempts to overcome these limitations. The concept of using biomarkers from mainly noninvasive tests as surrogate standards, or endpoints, for translational research is an important milestone for medical science.^1,2 This method was recognized by regulatory bodies, including the U.S. Food and Drug Administration and National Institutes of Health, in 1999.

Radiologists are increasingly confronted with imaging requests related to treatment control. Clinical colleagues seek standardized feedback that is more specific than merely "worsening" or "improving." Imaging signs are being used more in post-treatment assessment of musculoskeletal diseases, alongside laboratory tests and clinical biometry tools.

Anatomic changes have traditionally served as biomarkers in the Response Evaluation Criteria in Solid Tumors (RECIST). A reduction in tumor diameter, measured with CT, corresponds well with complete, partial, stable, and progressive responses to cancer therapies.

Another important bio-marker is the T-score, which is calculated from bone densitometry. This provides accurate, noninvasive measurements of the bone's calcium content. Changes in bone mineral density in re-sponse to therapy can thus be identified reliably. The T-score is used routinely for diagnosing osteoporosis.

Conventional x-rays offer a superior imaging biomarker for fractures.^3,4 Structural analysis "microtechniques," especially microCT, may in the near future help gain more insight into architectural changes in the bone.

Radiological depiction of the slowing of arthritis progression has been accepted as a surrogate biomarker when examining the effect of anti-rheumatic drugs on synovial inflammation to hand and finger joints. Erosions and joint space narrowing visible on conventional radiographs still form part of the various scoring systems quantifying the severity of joint damage.^5,6

Ultrasound and/or MRI have been used to assess arthritis and provide damage scores based on healing erosions and other signs of tissue repair. Imaging results from these modalities, however, are not yet acceptable as surrogate biomarkers in arthritis research.^7,8 Com-puter-aided diagnosis schemes based on new forms of artificial neural networks may help replace the cumbersome task of scoring the degree of damage from several dozen hand and finger joints.^9,10

Imaging can assist research outside the sphere of drug development as well. Creation of new devices for orthopedic procedures will be driven by virtual reality techniques based on multislice CT data sets. Refined segmentation techniques, such as edge detection of vertebral bodies, should be available in the near fu-ture.¹¹ Volumetric imaging may be a more powerful predictive instrument than 2D measurements of tumor diameter when as-sessing tumor response.

Radiologists should be aware that quantitative imaging will progressively enhance routine clinical work. Such imaging techniques need to be standardized and approved, especially when a contrast agent is to be administered, as is likely for the assessment of many musculoskeletal disorders. Imaging signs that have been recognized as biomarkers must be interpreted according to strict rules. Investigators should document their interpretation skills with certificates or interobserver agreement with reference centers.

Development of new high-quality biomarkers requires cooperation among academia, clinical practice, governmental bodies, and industry. This should also facilitate improvements to existing biomarkers though implementation of new imaging modalities, better investigation techniques, and refined interpretation skills.

DR. KAINBERGER is an associate professor of radiology at the Medical University of Vienna.

References

1. Peterfy CG. Role of MR imaging in clinical research studies. Semin Musculoskelet Radiol 2001;5:365-378.

2. Smith JJ, Sorensen AG, Thrall JH. Biomarkers in imaging: realizing radiology's future. Radiology 2003;227:633-638.

3. Genant HK. Assessment of vertebral fractures in osteoporosis research. J Rheumatol 1997; 24:1212-1214.

4. Wang YX. Medical imaging in pharmaceutical clinical trials: what radiologists should know. Clin Radiol 2005;60:1051-1057.

5. Sharp JT. Measurement of structural abnormalities in arthritis using radiographic images. Radiol Clin North Am 2004; 42:109-119.

6. van der Heijde DM. Radiographic imaging: the 'gold standard' for assessment of disease progression in rheumatoid arthritis. Rheumatology (Oxford) 2000;39 Suppl 1:9-16.

7. Ostergaard M, Edmonds J, McQueen F, et al. An introduction to the EULAR-OMERACT rheumatoid arthritis MRI reference image atlas. Ann Rheum Dis 2005;64 Suppl 1:i3-7.

8. Wakefield RJ, Balint PV, Szkudlarek M, et al. Musculoskeletal ultrasound including definitions for ultrasonographic pathology. J Rheumatol 2005;32:2485-2487.

9. Kainberger F, Peloschek P, Imhof H. [Quantitative imaging of rheumatoid arthritis: from scoring to measurement.] Radiologe 2006;46:(in press). German.

10. Wick M, Peloschek P, Bogl K, et al. The X-Ray RheumaCoach" software: a novel tool for enhancing the efficacy and accelerating radiological quantification in rheumatoid arthritis. Ann Rheum Dis 2003;62:579-582.

11. Smyth PP, Taylor CJ, Adams JE. Vertebral shape: automatic measurement with active shape models. Radiology 1999;211:571-578.