On the development of rational standards for nuclear response evaluation.

Aim: Small variations in test protocols can disproportionately affect the sensitivity and specificity of response evaluation in nuclear medicine. Although use of standardised methods can remedy this, standards must be shown to add value. We think a concept of "societal efficacy" is the benchmark criterion for value of medical interventions. This paper gives an overview of literature on nuclear response evaluation, and promotes a decision-analytic approach to the synthesis of standards.

Methods: A Medline search using the OVID database, 1966-January 2002. Reports were organised in relation to mode of treatment and timing of follow-up evaluation. Protocols of multimodality treatment were classified according to the treatment with the greatest tissue-inflammatory potential. The database will be made available on-line at the Website of the European Organisation for the Research and Treatment of Cancer (EORTC) Functional Imaging Group (http://www.eortc.be).

Results: Two hundred and twelve reports could be classified as primary studies in humans. 125 were formal "before-and-after" studies of response to anticancer therapy. More than 60 reported the use of serial positron emission tomography (PET) with fluorodeoxyglucose (FDG).

Conclusion: Descriptive reports of the accuracy and applications of new diagnostic technologies need to be linked to an expectation of improved research or clinical outcomes. To manage the large volume of information will require a trans-disciplinary perspective and use of advanced decision-analytic

Methods: At stake is the possibility of an "industrial" upscaling of one of nuclear oncology' strongest applications.