Volumetric modulated arc therapy treatment planning based on virtual monochromatic images for head and neck cancer: effect of the contrast‐enhanced agent on dose distribution

Abstract Virtual monochromatic images (VMIs) at a lower energy level can improve image quality but the computed tomography (CT) number of iodine contained in the contrast‐enhanced agent is dramatically increased. We assessed the effect of the use of contrast‐enhanced agent on the dose distributions in volumetric modulated arc therapy (VMAT) planning for head and neck cancer (HNC). Based on the VMIs at 40 keV (VMI40keV), 60 keV(VMI60keV), and 77 keV (VMI77keV) of a tissue characterization phantom, lookup tables (LUTs) were created. VMAT plans were generated for 15 HNC patients based on contrast‐enhanced‐ (CE‐) VMIs at 40‐, 60‐, and 77 keV using the corresponding LUTs, and the doses were recalculated based on the noncontrast‐enhanced‐ (nCE‐) VMIs. For all structures, the difference in CT numbers owing to the contrast‐enhanced agent was prominent as the energy level of the VMI decreased, and the mean differences in CT number between CE‐ and nCE‐VMI was the largest for the clinical target volume (CTV) (125.3, 55.9, and 33.1 HU for VMI40keV, VMI60keV, and VMI77keV, respectively). The mean difference of the dosimetric parameters (D99%, D50%, D1%, Dmean, and D0.1cc) for CTV and OARs was <1% in the treatment plans based on all VMIs. The maximum difference was observed for CTV in VMI40keV (2.4%), VMI60keV (1.9%), and VMI77keV (1.5%) plans. The effect of the contrast‐enhanced agent was larger in the VMAT plans based on the VMI at a lower energy level for HNC patients. This effect is not desirable in a treatment planning procedure.


2.A | Electron density conversion table
The electron density relative to water (ED) lookup table (LUT) was generated using a tissue characterization phantom (GAMMEX467, Gammex RMI, Middleton, WI), which is used in conjunction with a CT scanner to establish the relationship between the ED of various tissues and their corresponding CT numbers in HU. The reference materials mimicked human body organs with known EDs, and the specifications are listed in Table 1. To extend the usable range of the LUT, reference material made of aluminum was used in this study. The arrangement of the reference materials of the tissue characterization phantom are shown in Fig. 1. The DECT scans (Revolution HD, GE Healthcare, Milwaukee, WI) were performed using 80/ 140 kVp photon beam energies, and the measurement was repeated five times to minimize random variations of the HU measurement.
Based on the acquired data, VMIs at 40, 60, and 77 keV (VMI 40keV, VMI 60keV , and VMI 77keV, respectively) were reconstructed, and the scanning parameters were as follows: helical pitch: 0.984:1, field of view (FOV): 500 mm, slice thickness: 2 mm, and volume CT dose index: 15.02 mGy. The theoretical ED was plotted as a function of the mean CT numbers in VMI 40keV, VMI 60keV , and VMI 77keV , and the LUT for each VMI was registered in a treatment planning system (TPS) (Eclipse version 13.7, Varian Medical Systems, Palo Alto, CA).

2.B | CT scans
This retrospective study included 15 patients (median (range) age 62 (50-76) yr; 11 male and 4 female) treated using VMAT technique at our institution. The primary site was oropharynx in six patients, hypopharynx in three, larynx in three, nasopharynx in one, maxillary sinus in one, and oral cavity in one. The study was approved by our ethics committee with written informed consent provided by the patients. The patients were immobilized with a thermo plastic mask in a supine position. Two consecutive DECT scans were performed for each patient. The first DECT set was scanned before the contrast-enhanced agent was injected (nCE-VMI). The second DECT acquisition was performed with the injection of the contrast-enhanced agent to achieve 450 mgI/kg with an injection time of 50 s (1.5 to 1.9 ml/s), and the patients were scanned for 70 s after the T A B L E 1 Specifications of the reference materials of the phantom.  3 | RESULT Figure 3 shows the respective LUTs generated from the VMI 40keV , VMI 60keV , and VMI 77keV (LUT 40keV , LUT 60keV , and LUT 77keV ). The equivalent CT numbers were obtained for the low-density materials in LUT 40keV , LUT 60keV , and LUT 77keV (ED < 0). The CT number changed considerably for high-density material rods, and the CT number of aluminum varied widely in the range from 1796 ± 55.9 HU (VMI 77keV ) to 3018 ± 49.3 HU (VMI 40keV ).
The CNR at VMI 77 keV was significantly lower than those of the other energy levels (40 and 60 keV). and MD than those in the nCE-VMI (P < 0.01) for the targets and OARs (except for brain stem), as the energy level of the VMI decreased. The maximum difference of the ED between the CE-VMI and nCE-VMI was 0.8 for the CTVe(VMI 40keV ), and that of MD was 0.9 g/cm 3 for the CTVe and parotid ipsilateral (VMI 40keV ). Figure 5 summarizes the relative differences in the dosimetric parameters using the AAA between the CE-and nCE-TP for the target and OARs. Regarding the target, the CE-TP had higher values of the dosimetric parameters than the nCE-TP for almost all patients, and the mean difference in the dosimetric parameters was <1%. As the energy level of the VMI decreased, statistically significant differences in the dosimetric parameters (P < 0.05) were observed (except for D 99% in the CTVe). The maximum differences in D 1% for the CTVb were 0.8% and 0.7% in the VMI 40keV and VMI 60keV, respectively. For the CTVe, the maximum differences in the VMI 40keV , VMI 60keV , and VMI 77keV were 2.4%, 1.9%, and 1.5% in D 99% , respectively. For the OAR, a similar trend of differences in the dosimetric parameters was observed as that for the targets. The maximum difference was −1.7% for the brainstem (D 0.1cc ) in the VMI 40keV , −1.1% in the VMI 60keV , and −0.9% in the VMI 77keV .
The relative differences between the CE-and nCE-TP for the target and OARs in the dosimetric parameters using the AXB are shown in Fig. 6 | 149 highest at VMI 40keV , and they concluded that tumor conspicuity is greatest at VMI 40 keV and is useful for tumor detection. 8 In our study, the energy level of the VMI for highest CNR was the same as their study. Although the VMI at low-energy levels with a high image quality may improve the accuracy of target delineation in radiotherapy treatment planning, the iodine does not exist in the patients during dose delivery.
Alan et al. evaluated the effect of contrast in the dosimetry of the VMAT for the HNC using a conventional 120 kVp image. 12 In their study, treatment planning on contrasted images generally showed a lower dose administered to both organs and target than planning on noncontrasted images, and the difference is generally <2%. To the best of our knowledge, this is the first study to compare the VMAT plans generated using the VMIs between with and without the contrast-enhanced agent on the dose calculations. The effect was significant as the energy level of the VMI decreased ( Fig. 5) because the difference in the CT number between the CEand nCE-VMI was large in the VMI at a low-energy level. In this study, the mean effect on the targets and OARs in the VMI 40keV was <1%, and the effect did not differ more than 2.5% in any patient.
The main advantage of the AXB algorithm is its accuracy in practice of radiotherapy. 19,20 However, the impact of the contrast-enhanced agent was larger in the AXB than AAA, as shown in Figs. 5, 6. We comparing the before-IMRT delivery and postfraction CBCT scans, and the mean of the difference between before and after treatment was <1 mm in three degrees (medial-lateral, cranial-caudal, and anterior-posterior). In our study, the time taken between the nCE and CE scan was approximately 3 min, which is equal or shorter than that of treatment. Thus, the effect of the intrafraction error is considered to be small, and a medical physicist validated the registration and checked that there was no considerable anatomical change and rotation error between the CE and nCE scan. There also were no large streaking artifacts in the CE scans. Therefore, deformable image registration may be an effective tool for decreasing the dose difference by changing the patient position between the CE-VMI and nCE-VMI scans. Second, we used a single-source DECT, which uses a single x-ray tube with a fast kilovolt switching system and a single detector. However, there are several types of DECT acquisition systems: dual-source DECT, which utilizes two x-ray tubes and two detectors, and detector-based spectral CT, which uses a single x-ray tube and a detector made of two layers. 26 Our previous study demonstrated that the CT numbers in the case of VMIs at low energy levels could be considerably inaccurate, especially for high-density materials. 15 Because the CT numbers are converted into ED and MD values by using LUTs for dose calculation in the radiotherapy treatment planning processes, the inaccurate CT numbers could possibly affect the dose distributions. Finally, the effect of the contrast-enhanced agent on the accuracy of the target and/or OAR contouring was not investigated in this study. The difference of contouring with and without the contrast-enhanced agent, which has not been investigated in this study, may be the effect of dose distribution.

| CONCLUSION
With the decrease of energy levels in the VMIs, the difference in the CT number between CE-and nCE-VMI increased. The deviation of the CT number affected the dose calculation more significantly in the treatment plan using the VMI 40keV than that using the VMI 77keV . Although the mean effect of the contrast-enhanced agent was <1%, the maximum difference was 2.4% for evaluating the target dose in the treatment plan based on the VMI 40keV .
F I G . 6. Difference in the dosimetric parameters between the treatment plans using Acuros XB based on the contrastenhanced-(CE) and non-CE (nCE)-virtual monochromatic image for the (a) planning target volume and (b) organs at risks.