TL;DR
This study investigates how the iodinated rectal hydrogel spacer, used to improve prostate cancer treatment planning, can lead to inaccuracies in proton therapy (PT) dose calculations due to its high CT intensity. The researchers found that replacing the iodinated spacer with water in treatment planning systems significantly reduces these errors, suggesting water as a practical and clinically viable alternative for material override.
What Problem Does This Paper Address?
Radiation therapy (RT) for prostate cancer is limited by the need to spare nearby organs-at-risk (OARs) like the rectum and bladder. Proton therapy (PT) offers advantages in dose deposition, but precise targeting is still crucial. Hydrogel spacers are used to create a buffer between the prostate and rectum, increasing the distance to OARs and allowing for dose escalation. However, standard hydrogel spacers are often invisible on CT scans, necessitating alternative localization methods. Iodinating the spacer improves its visibility on CT, but this high iodine content can lead to treatment planning systems (TPS) misinterpreting its material properties. Specifically, the high CT intensity (Hounsfield Units or HU) of the iodinated spacer can cause the TPS to assign it a high-density material property, potentially leading to inaccurate proton stopping power and range calculations, and thus affecting dose distribution accuracy. This paper aims to systematically quantify this discrepancy and propose a solution.
Methods
The researchers conducted a systematic study using CT scans from 20 prostate cancer patients who received a hydrogel spacer. They simulated three scenarios for material properties within the TPS:
- True Material (TM): This served as the ground truth, using the actual elemental composition of the iodinated hydrogel spacer, verified by experimental measurements of its proton stopping power ratio (SPR).
- No Material (NM) Override: In this scenario, the TPS used the CT intensity (HU) of the iodinated spacer directly to determine its material properties. Due to the iodine, this resulted in the spacer being treated as a high-density material.
- Water Material (WM) Override: Here, the iodinated spacer's material property was manually overridden to that of water.
Proton therapy plans (IMPT) were generated for each patient using the TM scenario. The dose distributions were then recalculated for the NM and WM scenarios using the same plans and CT data. The accuracy of the NM and WM scenarios compared to the TM ground truth was evaluated using:
- 3D Gamma Index (GI): This metric assesses the spatial agreement between dose distributions, with criteria of 1%/1mm, 2%/2mm, and 3%/3mm. A passing ratio (PR) above 90% was considered clinically acceptable.
- Dosimetric Indices: Various dose-volume histogram (DVH) metrics were calculated for critical structures, including the clinical target volume (CTV), prostate, seminal vesicles, bladder, rectum, and neurovascular bundles (NVB). These included D02, D05, D20, D50, D95, D98, mean dose, and equivalent uniform dose (EUD).
Datasets
The study utilized CT scans from 20 prostate cancer patients who had received a hydrogel spacer. No new datasets were introduced; rather, existing patient data was used to simulate different material overrides within the TPS.
Key Insights and Ideas
- Iodine's Impact on CT Visibility is a Double-Edged Sword: While iodine makes the spacer visible on CT, its high density can mislead PT planning systems.
- "No Material" Override is Problematic: Treating the iodinated spacer based solely on its high CT intensity (NM scenario) leads to the most significant dose discrepancies, particularly for OARs like the neurovascular bundles (NVB).
- Water as a Practical Proxy: Overriding the iodinated spacer's material to water (WM scenario) provides a remarkably accurate representation of the dose distribution compared to the true material. This is a significant finding because water is a standard material in all TPSs.
- NVBs are Particularly Sensitive: The neurovascular bundles, critical for erectile function, showed the largest dose deviations in the NM scenario, highlighting the clinical importance of accurate material assignment.
Key Results
- The NM scenario showed the most deviation from the ground truth (TM).
- The iodinated spacer, when treated as a high-density material (NM), led to dose differences of up to 350 cGy (approximately 5% of the prescribed dose) for the NVB structure.
- Gamma index passing ratios were notably lower in the NM scenario compared to the WM scenario, especially for the NVB. For example, PR11 dropped as low as 0.7 for NVB in the NM-TM scenario, far below the clinical threshold of 0.9.
- The WM scenario demonstrated high agreement with the TM scenario, with dose differences for NVB being less than 60 cGy and not statistically significant.
Limitations and Future Work
- Limited Beam Arrangement: The study used only two lateral beams. The authors suggest further studies are needed to determine the effect of different beam arrangements on dose discrepancy and to find optimized beam arrangements and spacer locations.
- Patient Anatomy Variability: The authors acknowledge that the magnitude of dose discrepancy can be highly variable depending on patient anatomy and spacer location.
- Future Work: The paper strongly advocates for the use of water density as a clinically accessible and convenient alternative material override for iodinated spacers in PT planning. Further research could explore the impact of different spacer volumes and materials, as well as investigate the optimal placement strategies for these spacers in conjunction with various beam arrangements.
Summary updated