Changing paradigms for tumor response spark debate

Yes, size does matter, but it is certainly not the only way to monitor the efficacy of cancer treatment. The way in which lesions are measured is extremely important as well, ECR delegates were told at a special focus session on the assessment of tumor response.

For the past 50 years, imaging has been regarded as a substitute or surrogate marker to assess therapy response, with the most widely used criterion being a change in size, said Prof. Rodney Reznek, a professor of diagnostic imaging at St. Bartholomew's Hospital in London. The development of chemotherapy drugs that act on cancers in novel ways, such as agents that are cytostatic rather than cytotoxic, has led to questions about the universal suitability of this strategy.

"It has become apparent that size may not be the most sensible or reliable criteria to assess response. For this reason, other biomarkers are being used, one of which is functional imaging," he said.

Of course, size is not being abandoned altogether. The problem is that the correlation between tumor shrinkage as measured on CT and patient survival is not quite as strong as might be expected, said Dr. Larry Schwartz, chief of MRI at New York City's Memorial Sloan-Kettering Cancer Center.

"So while we are doing okay, we really haven't yet optimized tumor response assessment," he said.

The size-based criteria that are used to decide how cancer patients in clinical trials have responded completely to their treatment have recently been changed. Version 1.1 of RECIST (Response Evaluation Criteria of Solid Tumours), published in the European Journal of Cancer at the start of 2009, still requires that unidimensional measurements be made of target lesions. The overall tumor burden will also still be defined according to the sum of lesion diameters.

What is different, however, is the selection of target lesions. RECIST 1.1 states that up to five lesions should be measured, with a limit of two per organ. This replaces the previous RECIST guidance to measure 10 lesions (five per organ). Another change is that the criteria for "progressive disease" require not only a 20% increase in size but that this change should be greater than 5 mm.

All ongoing clinical trials should stick with the original version of RECIST, Schwartz said. This means that the revised criteria will not actually be used for several more months yet.

PET/CT is the main modality in the move to functional imaging for therapy response. For some cancers, such as lymphoma, FDG-PET/CT has become a standard way of measuring tumor response at the end of treatment, said Prof. Wolfgang Weber, chair of nuclear medicine at the University of Freiburg in Germany. This is achieved by simply comparing a pretherapy baseline scan against the results of post-therapy imaging.

FDG-PET/CT is also being investigated as a predictive tool. With so many different drugs becoming available, it would be helpful for oncologists to know sooner rather than later whether their chosen agent is likely to work. Research is under way to see whether the results of an FDG-PET/CT scan after the first cycle of chemotherapy could be a good indicator of patient outcome.

FDG-PET/CT is not the only functional imaging option. Many MR-based methods could equally be used to evaluate tumor response to therapy, according to Dr. Anwar Padhani, a consultant radiologist and director of clinical research at the Mount Vernon Cancer Centre in London.

Padhani outlined a range of MRI techniques that can support decision making in clinical trials of new cancer drugs. These included MR spectroscopy to monitor altered metabolism, blood oxygen level-dependent MRI to identify hypoxia, diffusion MRI to assess cell proliferation, and dynamic contrast-enhanced MRI to monitor angiogenesis through tissue perfusion.

He highlighted a number of key questions that still need to be addressed as this role for imaging proceeds: How do we take this information and integrate it into a patient study? How do we correlate changes we are seeing with patient outcomes? How do we develop common measurement and analysis methods so we can standardize these and use them in multicenter studies?

The Quantitative Imaging Biomarker Alliance, formed by the RSNA, is looking at this latter issue, Schwartz said. Questions from ECR delegates to the speakers on the potential role of yet more functional imaging techniques stressed the need for clarity on this topic. For example, radiology researchers in France are investigating the role of ultrasound as a tool for tissue perfusion measurements. Dynamic contrast-enhanced imaging may also be performed on CT, rather than MRI, and the results combined with PET.

"What we need is some single-center trials to demonstrate proof of concept, to work out the reproducibility of the techniques, and then to move that in a very rigorous and rapid way into Phase III trials," Schwartz said.