Identification of Brain Metastases Origins using Contrast Enhancement Patterns and Percentage Signal Recovery Values from Magnetic Resonance Imaging

Abstract

Background: Brain metastases (BMs) are common in systemic malignancies and can affect patient prognosis significantly. The most prevalent diagnostic imaging modalities for characterising BMs, specifically contrast-enhanced CT and MRI, provide minimal insight into the biology of a lesion or its origin. Newer imaging modalities may benefit from DSC-MRI, and with imaging markers such as PSR, function to differentiate BMs, based on vascular physiology, and integrity of the blood-brain barrier (BBB). This study aimed to correlate PSR values and contrast enhancement patterns observed in DSC-MRI images to determine the primary origin of BMs, specifically differentiating between lung cancer and breast cancer.

Materials and Methods: A single-centre, retrospective study was conducted with 70 patients with confirmed BMs using DSC-MRI (48 lung and 22 breast carcinomas). Standard formulas were used to calculate PSR values. Enhancement patterns were classified as ring-shaped, heterogeneous, homogeneous, or cystic. The data were analysed descriptively, and receiver operating characteristic (ROC) curves were used with SPSS. Statistical significance was set to p<0.05.

Results: There was noted variation in PSR and enhancement patterns alone by tumour origin. Cystic and ring-shaped patterns were more prevalent in breast carcinoma, whereas heterogeneous patterns were observed more frequently in lung carcinoma. The overall mean PSR was 23.94% ± 9.26. From the ROC analysis, the area under the curve of PSR alone demonstrated weak discriminatory ability (AUC = 0.577, p = 0.303). Combining augmentation with enhancement pattern data increased diagnostic accuracy, most notably in the cystic patterns (AUC = 1.0, p = 0.034).

Discussion: The findings suggest that PSR and contrast enhancement may play a role in non-invasively suggesting the origin of tumour type in brain metastases. The PSR on its own may not be singularly sufficient; however, when combined with other forms of morphological imaging biomarkers, it does positively support enhancement in diagnostic confidence with tumour types with previously unknown malignancy sources.

Conclusion: PSR values obtained from DSC-MRI, coupled with patterns of contrast enhancement, suggest a plausible, non-invasive method for differentiating the primary origins of BMs. These findings warrant further investigation, including larger sample sizes and longitudinal studies, to validate these findings and systematically integrate the results from the diagnostic process into the priority order of routine clinical workflows.