Hippocampus, spare or not to spare? Normal Tissue Complication Probability considerations

Purpose: The purpose of the study was to use the Lyman-Kutcher-Burman model to calculate and compare neurocognitive function (NCF) impairment in the two types of dose distributions in Volumetric Arc Whole-brain radiotherapy (VMAT WBRT).

Methods: The total dose prescribed to the Planning Target Volume (PTV) was 25 Gy and 30 Gy in 10 fractions. During the optimization of the Volumetric Arc Therapy Hippocampal-Avoidant WBRT (VMAT HA-WBRT) plan, the left and right hippocampus the D₁₀₀ < 9 Gy, and the point dose D_max < 16 Gy were minimalized based on RTOG0933 criteria. As an alternative 59 plans for non-hippocampus sparing, 25 Gy, and 59 plans for 30 Gy VMAT WBRT plans were prepared. To calculate the probability of NCF impairment, the Lyman-Kutcher-Burman (LKB) normal tissue complication probability model was used.

Results: The probability of NCF impairment in the HA-WBRT VMAT was equal to 38%, and it was significantly lower compared to 90% in the 25 Gy WBRT VMAT and 97% in the 30 Gy WBRT VMAT (p < 0.05).

Conclusions: The probability of NCF impairment was 51-59 percentage points lower in VMAT HA-WBRT compared to VMAT WBRT plans. To achieve a 5% probability of NCF impairment, the D₄₀ value should be reduced to a value of approximately 1 Gy. However, due to the complicated nature and types of NFC impairment, further analysis and strategies are needed to optimize dose distribution.