Brain tumor imaging in clinical trials

Abstract

There are substantial challenges in the radiologic evaluation of tumor size during clinical trials, and it is important for neuroradiologists to have a firm understanding of these issues. This review will examine measurement approaches, response criteria, selection of lesions for measurement, technical imaging considerations, interval between tumor measurements and response confirmation, and validity of imaging as a measure of efficacy.

Fig 1.

Three enhancing foci in a patient with glioblastoma illustrate issues with lesion measurement during clinical trials. Lesion A is homogeneously enhancing and exceeds 10 mm in diameter and thus is ideal for serial measurement by RECIST or 1D (lower left), Macdonald or 2D (lower right), and volumetric (upper right) approaches. Lesion B is predominantly necrotic and is amenable to volumetric measurement (upper right) because the enhancing and nonenhancing components can be segmented. Lesion C is too small in diameter (8 mm) for accurate serial measurement and should be followed as a nonmeasurable lesion (see text). Images are postgadolinium contrast-enhanced axial T1-weighted.

Fig 2.

Infarcts on immediate postoperative MR images are common. These infarcts often demonstrate nodular gadolinium enhancement on subsequent studies, a finding that could be easily confused with tumor. Immediate postoperative DWI (upper left) and apparent diffusion coefficient (upper right) show restricted diffusion (arrows), followed by a 3-month postoperative T2-weighted (lower left) image and a gadolinium-enhanced T1-weighted (lower right) image showing encephalomalacia and enhancement of the infarct. (Reprinted by permission of Lippincott Williams & Wilkins; Clinical and radiographic features of peritumoral infarction following resection of glioblastoma. Neurology 2006;67:1668–70).

Fig 3.

Novel therapeutic agents in clinical trials may require use of imaging techniques other than gadolinium-enhancing tumor. Shown here is decreased tumor enhancement but not diameter in a patient with glioblastoma after initiation of a therapy with an inhibitor of VEGF and irinotecan. Note the increase in extent of the infiltrative component of the lesion (lower right). Axial post-gadolinium contrast T1-weighted images (left-hand column) and axial T2-weighted/fluid-attenuated inversion recovery images (right-hand column) were acquired before (upper row) and 7 weeks after (lower row) institution of therapy.