Inter-subject cerebrovascular variability: a source of uncertainty for dose calculation to circulating blood cells for glioblastoma patients treated with modern radiotherapy techniques.
Abdelkhalek HammiNadya ShusharinaSonya DjuffouoPublished in: Physics in medicine and biology (2024)

To assess how inter-subject variations in brain vasculature among glioblastoma patients affects the calculated dose received by circulating blood cells (CBC) during radiotherapy and its subsequent impact on CBC depletion.
Methods:
Ten glioblastoma patients treated with either IMRT or VMAT were selected. For each patient, 23 cerebrovascular models were developed based on 23 healthy subject MRangiography data to simulate intra- and inter-subject blood vessel diversity. Based on the corresponding treatment plan of the patient, the dose to CBC was calculated for all the 230 scenarios. The impact of inter-subject variation on fraction of irradiated blood volume (VD>0cGy) and lymphocyte kill rates as a function of the CTV size and treatment technique were analysed.
Results:
The dose fluctuation to CBC was higher in IMRT plans compared to VMAT plans. The uncertainty in the VD>0cGy was 18.3% for IMRT and 2.0% (CI95%) for VMAT per treatment fraction. The dispersion of the D2% index was 6 cGy for IMRT and 1 cGy for VMAT (CI95%) for one single treatment fraction of 2 Gy. The uncertainty in killed CBC due to inter-subject diversity in brain blood vessel increased with increasing CTV size. The depletion uncertainty of CBC was σ = 11.2%.
Conclusions:
VMAT showed greater robustness against inter-subject variation in blood vessels compared to IMRT. We recommend considering the uncertainty in depleting CBC resulting from the use of less patient-specific and generic blood vessel phantoms to improve the RiL assessments.
.
Keyphrases
- induced apoptosis
- early stage
- radiation therapy
- finite element
- case report
- machine learning
- health insurance
- combination therapy
- white matter
- newly diagnosed
- end stage renal disease
- resting state
- deep learning
- ejection fraction
- climate change
- cell proliferation
- signaling pathway
- cell death
- functional connectivity
- electronic health record
- peripheral blood
- rectal cancer
- smoking cessation