RECIST revisited: a review of validation studies on tumour assessment

Abstract

The response evaluation criteria in solid tumours (RECIST) was developed in the late 1990s to replace the WHO criteria for response evaluation. The new criteria included important changes such as unidimensional tumour measurement, selection of target lesions with a minimum size, details concerning imaging modalities and a new threshold for assignment of objective progression. RECIST was published in February 2000 and very quickly came into operation first in clinical trials performed under the auspices of EORTC, US NCI or NCI Canada Clinical Trials Group but was adopted quickly thereafter by the entire cancer clinical research community. As several key features of RECIST were based on analysis of retrospective clinical data, it was felt important to carefully monitor the implementation of the guidelines and stimulate prospective validation studies. This paper reviews the literature that has been published on RECIST from 2000 up to November 2005. In total 60 papers and ASCO, abstracts directly refer to research studies or reviews related to RECIST and its implementation. Amongst the 60 references identified for this review, 11 papers refer to validation studies (seven prospective and four retrospective), six papers refer to the comparison of unidimensional measurements versus bi or tri-dimensional measurements, 12 papers address issues raised with the implementation of RECIST in Mesothelioma and Gastro-Intestinal Stromal Tumours and four papers report on an adaptation of RECIST for specific tumour types. In general, RECIST has been well received by the scientific community and most validation studies fully support the implementation of the new criteria. As expected, however, some issues have been identified. In keeping with the mathematical differences in definition of progression, RECIST delays the identification of progression as compared to WHO criteria in some instances. RECIST criteria are not easily applicable in some types of trials such as those in paediatric tumours and in mesothelioma. Furthermore, anatomical changes in the tumour as described by RECIST may be detected later than functional changes in some circumstances, as for example in Gastro-Intestinal Stromal Tumours treated with Imatinib. However, there is no other universal method of tumour assessment as yet and functional imaging methods have not been validated and will not be widely available for some time. The findings of this review, together with experience acquired thus far and the results of some ongoing research projects, have paved the way for RECIST 2.0 to be hopefully announced later this year.