112IN - Radiologist - How to perform correct RECIST assessment

Date 29 September 2012
Event ESMO Congress 2012
Session ESMO-EANM-ESR Joint symposium: Imaging biomarkers in the era of targeted therapies
Topics Imaging, Diagnosis and Staging
Presenter Yves Menu
Authors Y. Menu
  • Department Of Radiology, Saint Antoine Hospital, 75012 - Paris/FR


RECIST is a universal standard in tumour response evaluation. Therefore, it is of utmost importance that imaging technique and radiological reports are compliant with the recommendation. Advantages of standardization comprise the comparability of different examinations in the same patient, and responses in a cohort. Precise definition of response or progression makes possible to evaluate the accuracy of a specific treatment. Another advantage is that structured reporting is easier to perform, improving efficiency and communication. Technical requirements are important though easily achievable: slice thickness of 5 mm or less reconstructed in the axial plane, contiguous slices (on CT, minimal gap on MRI), robust and reproducible contrast injection with similar delay and injection rate, similar coverage ensure proper comparison of images and clear basis for decision making. Thinner slices, alternative planes, specific contrast media and/or customized MRI sequences are optional, and should not replace basic techniques. Interpretation of baseline images requires a perfect knowledge of the method for measurement of the largest diameter, identification of targets and non targets. Interpretation of evaluation examinations needs iterative comparison with Baseline and Nadir examinations, and identification of unequivocal new lesions. Obviously, a universal standard does not fit all oncology cases. It appears that some tumours like GIST or HCC may need adapted criteria in addition to RECIST. PET is only mildly included among the criteria, although it proves to be an important method for evaluation. Finally, morphology does not summarize tumour biology. Besides RECIST, the addition of structural, metabolic and/or functional information would be desirable.


All authors have declared no conflicts of interest.