Abstract

Raphael Bodensohn¹, Robert Forbrig², Stefanie Lietke³, Jonas Reis², Anne-Laure Boulesteix⁴, Ulrich Mansmann⁴, Indrawati Hadi¹, Daniel Felix Fleischmann¹, Kelly Hyeon Joo von Henning-Yoo¹, Johannes Mücke⁵, Adrien Holzgreve⁶, Nathalie Lisa Albert⁶, Viktoria Ruf⁷, Mario Dorostkar⁷, Stefanie Corradini¹, Claus Belka¹, Niklas Thon³, Maximilian Niyazi¹

¹University Hospital LMU Munich, Department of Radiation Oncology, Munich, Germany; ²University Hospital LMU Munich, Institute of Neuroradiology, Munich, Germany; ³University Hospital LMU Munich, Department of Neurosurgery, Munich, Germany; ⁴Faculty of Medicine LMU Munich, Institute for medical information-processing, biometry and epidemiology, Munich, Germany; ⁵Department of Radiation Oncology, University Hospital LMU Munich, Munich, Germany; ⁶University Hospital LMU Munich, Department of Nuclear Medicine, Munich, Germany; ⁷University Hospital LMU Munich, Center for Neuropathology and Prion Research, Munich, Germany

After cranial radiotherapy pseudoprogression (PsP) may frequently occur displaying a similar imaging pattern as progressive disease (PD) with contrast enhancement and perifocal edema. These equivocal findings make it challenging even for experienced radiologists to differentiate between both disease states. Contrast clearance analysis (CCA) could provide additional and important information by analyzing the contrast media (CM) washout; in theory, tumor tissue shows a higher CM washout than reactive tissue. This study evaluates the accuracy off CCA in distinguishing PsP from PD and tests its potential as a diagnostic tool.

For this prospective study 33 patients were to be included, who received cranial radiotherapy and required a stereotactic biopsy for further differentiation of an unclear progression on follow-up MRI. The planning-MRI with CM for the biopsy was complemented with the late phase T1-sequence one hour after the regular MRI T1-sequence with CM; for CCA both sequences are coregistered and a differential map is calculated by subtracting both imaging studies. Two blinded experienced neuroradiologists viewed the CCA independently and rated the progression between “real” tumor progression and PsP: The radiological assessment was then compared with the pathological results from the biopsy, and its accuracy calculated statistically.

The aimed number of 33 was reached; 16 (48.5%) patients were treated on a primary brain tumor, and 17 (51.1%) on a secondary brain tumor. 17 lesions (51.5%) were diagnosed by CCA as PD and 15 (45.5%) as PsP; for one patient with a non-contrast-enhancing low-grade glioma CCA did not deliver any output and therefore was not analyzable. CCA showed an accuracy in predicting the histological result of 0.844 (95%-CI 0.672-0.947) with a nearly significant p-value of 0.051 for N=32. If the one case, which was not analyzable by the CCA, had been predicted correctly, the accuracy would be 0.848 (95%-CI 0.681-0.949) and the p-value 0.041. Sensitivity and specificity of CCA was 0.929 (95%-CI 0.661-0.998) and 0.778 (95%-CI 0.524-0.936), respectively. 25 patients received due to regular diagnostics additionally a FET-PET. The accuracy hereby was 0.75 (95%-CI 0.533-0.902), the study was not powered for this examination.

Despite of slightly missing the threshold of statistical significance, CCA was highly accurate in this cohort, and could prove to be an easy and helpful method for distinguishing PsP from PD after radiotherapy. It seems comparable in accuracy to the FET-PET.