Is a prospective trial necessary to suggest a clinical relevance?

Is a prospective trial necessary to suggest a clinical relevance? We would like to thank Dr Chamberlain for his comments on our article 1 and take this opportunity to respond to the issues he has raised. Radiological measurement of low-grade glioma (LGG) growth is a hot topic and clarification is required. Firstly, we wish to emphasize that although not validated by a prospective trial, our calculation method (velocity of diametric expansion [VDE]) has been successfully reproduced in other institutions. Four other studies (107, 64, 21, and 21 patients in each, independent databases) demonstrated similar ranges of VDE (median 3.5 – 5.5 mm/y) using the ellipsoid approximation used in our study. 2 – 5 Secondly, our study reflects everyday clinical practice, where routinely one compares interval MR images. We have observed that the MR characteristics of LGG do not vary significantly between 1.5T and 3.0T synchronous MRIs (personal observation). Furthermore, it has been demonstrated that areas of high signal corresponding to an LGG are equivalent on both T2-weighted and fluid attenuated inversion recovery sequences. 6 We acknowledge that a comparison study has never been performed to assess the equivalence of the 2 methods of volume measurements (seg-mentation vs ellipsoid approximation). However, measurements of VDE varied little between these techniques as reflected by: (i) a high coefficient of reliability at 0.824 (n ¼ 124, personal observations) and (ii) the similarity of VDE measurements performed before (ellipsoid approximation) and after (segmentation) partial surgical removals. 7 Thirdly, the time interval between MR examinations is indeed a crucial issue. One should bear in mind that the practical aim of our approach is the identification of " high-risk " LGG presenting with a VDE ≥8 mm/year (ie, an increase in the mean tumor diameter .1 mm over a 6-wk interval and .2 mm over a 3-mo period). We have previously proposed that the time interval between scans be determined on an individual basis, and according to the clinical condition of the patient. 8 Finally, the choice of an 8 mm/year cutoff is not arbitrary; it fits clinical practice and reflects our previous finding that patients harboring an LGG with a VDE ≥8 mm/year have a clinical course in keeping with a more malignant glioma. 9,10 We have demonstrated that VDE as a continuous predictor is an independent prognostic factor for overall survival, with a linear relationship between survival and VDE (hazard ratio, 1.09 per one unit …