Repeatability questions complicate search for imaging biomarkers

Would you drink a half-liter of cream? Not the most appealing prospect, but Prof. David Lomas did just that to check the robustness of liver fat as an imaging biomarker and self-experiment.

“Apart from paralyzing my stomach for the first 12 hours, the following day my liver fat went up 4%,” he told delegates at an ECR New Horizons session on biomarkers. “So there is an issue about dynamic range of fat in the liver, and elsewhere in the body. How rapidly does it change? We don’t really know. In fact, we don’t know what the normal level of fat in the liver is, either.”

Measurement of hepatic steatosis, using MR spectroscopy or chemical shift imaging, is one of a number of MRI biomarkers that could be used to investigate abdominal disease, according to Lomas, a professor of clinical MRI at the University of Cambridge in the U.K. He presented four such candidates during his lecture, outlining the biological basis for each and the main methods of analysis, as well as the weight of clinical evidence.

In the case of liver fat measurements, for example, both MRS and chemical shift imaging are robust, well-understood techniques, Lomas said. However, as with many up-and-coming imaging biomarkers, their repeatability is largely unknown. Such measurements are nonetheless likely to become more important, given the redefinition of fatty liver by drug companies as a treatable medical condition, not a normal variant that can be left alone.

MR-based iron measurements, which can indicate the severity of liver and/or heart disease, are now being used in fields outside of radiology. This is a good indication of a technique’s acceptance, he said. Researchers have two methods to choose from when measuring iron concentration: they can either look at the ratio of T2 shortening in the liver and muscle, or measure R2 and R2* directly. Both methods have been validated, but again, repeatability is uncertain.

Diffusion-weighted imaging is another promising MRI biomarker, this time as an early indicator of treatment response. The apparent diffusion coefficient is typically low in solid tumors but is elevated following necrosis, a change that could be measured using diffusion-weighted MRI. The technique is not necessarily straightforward, though. In the liver, for example, respiratory motion makes it difficult to track the tiny movements of water molecules, requiring breath-holds. Eating has also been shown to influence diffusion MRI results, and certain ADC measurements may be sensitive to perfusion.

Lomas described his fourth candidate biomarker, MR elastography, as being challenging. The concept is relatively straightforward: mechanical waves directed through the organ of interest are used to assess the elasticity or stiffness of tissue, and hence identify the presence of fibrotic changes. Early results are promising, but the technique is quite complicated and, at the moment, the technology is not being used at many sites.

“This is not an easy application because you need special hardware and analysis tools,” he said. “And how do you validate this? There aren’t easy ways you can measure stiffness of the liver in vivo except by this technique.”

Prof. Eric Aboagye, director of the Comprehensive Cancer Imaging Centre at Imperial College in London, described a selection of PET molecular biomarkers that are of interest for drug development work and may also have clinical applications. However, as with all new biomarkers, the questions are: Do we measure the imaging endpoint reliably, and is the imaging endpoint reproducible?

He outlined four key biological mechanisms that PET tracers are being developed to target: angiogenesis, hypoxia, cell proliferation, and apoptosis. Imaging biomarkers for these biological targets are advancing, but more validation is needed.

The clinical value of PET biomarkers must also be established, Aboagye said. F-18 MISO and copper-60 ATSM, for example, have both shown promise in hypoxia imaging and could potentially be used to facilitate “dose painting” in radiotherapy. This emerging method of treatment would see hypoxic areas of tissue being blasted with higher doses of radiation in a bid to improve efficacy.

A version of this article appeared in the 2010 ECR Today newspaper.