Susceptibility Weighted and Perfusion Imaging for Differentiating Recurrence and Radiation Necrosis in Post-Operative Malignant Brain Tumors: A Retrospective Cohort Study

Brain tumors present a challenge in oncology, necessitating comprehensive treatment strategies. Despite treatment, recurrence rates, especially in malignant tumors, remain high. Novel therapies have emerged, but treatment-related effects such as pseudo-progression (PsP) and radiation necrosis (RN) complicate diagnosis. Imaging techniques such as perfusion imaging and susceptibility-weighted imaging (SWI) show promise in distinguishing tumor progression from treatment effects. This retrospective study evaluated the effectiveness of these modalities in detecting brain tumours, their progression, and necrosis.

This study was carried out at the Aga Khan University Hospital over a period of 11 years using non-probability retrospective sampling. The inclusion criteria were histopathologically proven brain tumours, complete imaging records, and available medical histories. Data were collected from the hospital's database. Statistical analyses were performed using SPSS v26, employing logistic regression models to assess accuracy.

The study included 37 patients with brain tumours, with a near-equal sex distribution. The most affected lobe was the parietal lobe, and glioblastoma was the most prevalent tumour type. Perfusion imaging identified radiation effects in 62.2% of the cases, while SWI showed radiation necrosis in 64.9%. Logistic regression models indicated significant accuracy for both modalities, with combined perfusion and SWI showing higher specificity.

Our study highlights the effectiveness of novel MRI techniques, specifically perfusion MRI and SWI MRI, in distinguishing between tumour recurrence and therapy-induced remission in brain tumour patients. These imaging methods offer crucial insights for accurate diagnosis and treatment planning. Despite the study's limitations, including its single-centre design and small sample size, the findings underscore the need for further multi-centre research. Future studies should combine various imaging techniques with machine learning to improve diagnostic accuracy and include longitudinal analyses to assess long-term benefits. Future studies should integrate imaging techniques with machine learning and include longitudinal analyses.