2002 - Evaluating Tumor Growth and Dosimetric Under-Coverage Due to Delays in Brain Metastasis Radiosurgery

Purpose/Objective(s): The objective of this study was to evaluate the impact of time interval between planning imaging and stereotactic radiosurgery (SRS) delivery on brain metastases (BM).

Materials/Methods: We identified patients treated with SRS at a tertiary center for 1–4 intact BM. Inclusion criteria were no prior cranial radiotherapy and availability of both a thin-slice (1 mm) diagnostic and treatment planning magnetic resonance imaging (MRI) within a 31-day interval. To simulate the dosimetric impact of treatment delays, the targets were initially delineated on the diagnostic MRI. Then, the treatment planning (clinical) MRI was rigidly fused to the diagnostic scan. Lesions with geometrical shifts were excluded. A virtual SRS (dosimetric) plan was generated on the diagnostic MRI using the initial contours and the dosimetric coverage of the transferred clinical contours was evaluated to quantify the "coverage drop", simulating the target failure that would have occurred had the patient been treated based on diagnostic imaging. We also calculated the increase in effective diameter (ED, using the spherical equivalent formula), between the two scans to analyze impact of factors such as systemic therapy, primary site, extracranial disease, initial tumor volume, and time between scans on lesion growth. Lastly, we determined an theoretical isotropic margin (in 1-mm increments) required on the initial contour to cover the boundaries of the clinical lesion. Univariable/multivariable analysis was used to investigate factors associated with tumor growth.

Results: 100 patients with 189 lesions underwent SRS between 2018 and 2024. The median time between two MRI was 10 days (IQR: 5–14). Lung cancer (59%) was the most common primary and 67% of patients were not on any systemic therapy (65% of lesions). Of 189 lesions, 112 (59%) exhibited interval progression requiring =1 mm margin, forming the study cohort. Median ED on initial and clinical MRI, and percentage change in ED between scans was 6.2 mm (IQR: 3.4-10.5 mm), 7.1 mm (IQR: 4.6-11.5 mm), and 14% (IQR: 9-31%). Multivariable analysis demonstrated that each day of delay increased the ED by 1.34% (4.1% by equivalent volume; p<0.001), more pronounced in smaller lesions (as a continuous variable, p=0.045). Generating a radiosurgery plan using baseline imaging resulted in a median 17% (IQR: 8-21%) virtual target coverage loss. Applying an arbitrary margin (1 mm in 79%; 2 mm in 19% of lesions) to the initial lesion contour to capture the clinical lesion resulted in a median 0.11 cc (IQR: 0.04-0.36 cc) of normal brain (41% of the target volume) receiving the full prescription dose.

Conclusion: Within a median interval of only 10 days, nearly 60% of brain metastases required = 1mm margin, resulting in 17% dosimetric under-coverage. Agnostic isotropic margin expansion resulted in an increased volume of normal brain receiving the target dose (41% volume increase relative to target volume). We strongly recommend performing MRI proximate to SRS, whenever feasible.