Abstract
Background/Aim: Dosimetric parameters in volumetric modulated arc therapy (VMAT), non-coplanar VMAT (NC-VMAT), and automated NC-VMAT (HyperArc, HA) were compared for patients with maxillary sinus carcinoma (MSC). Patients and Methods: Twenty HA plans were generated to deliver 70.4, 64, and 46 Gy for planning target volumes with high (PTV1), intermediate (PTV2), and low risk (PTV3), respectively. The VMAT and NC-VMAT plans were retrospectively generated using the same optimized parameters as those used in the HA plans. Results: For PTV1, the three treatment plans provided comparable target coverages. For PTV2, the D95%, D99%, and Dmin in the HA plans (64.7±1.2, 62.7±2.1 and 54.6±6.2 Gy, respectively) were significantly higher (p<0.05) than those in the VMAT plans (64.3±1.7, 61.9±2.4 and 52.9±6.4 Gy, respectively). The NC-VMAT and HA plans provided significantly higher (p<0.05) dosimetric parameters for PTV3 than those in the VMAT plans, and D99% in the HA was significantly higher than that in the NC-VMAT plans (52.5±3.0 vs. 51.8±2.7 Gy, p<0.05). The doses to the brain and brainstem were lowest in the HA plans (p<0.05). Moreover, dosimetric parameters of the contralateral organs (lens, optic nerve, retina, and parotid) were lower in the HA plans than in the VMAT and NC-VMAT plans. Conclusion: The HA plans provided the best target coverage and OAR sparing compared with VMAT and NC-VMAT plans for patients with MSC.
Maxillary sinus carcinomas (MSCs) are relatively rare neoplasms (approximately 80% of all paranasal sinus carcinomas) that account for only 3% of all head and neck carcinomas (1, 2). The incidence rate of paranasal sinus carcinomas in Japan was 4.2 times higher for males and 3.6 times higher for females compared to that in the United States (3). A high incidence of maxillary sinus carcinoma has been reported in Asian countries (4). Patients with MSC conditions are frequently diagnosed at an advanced stage due to the absence of symptoms in early stages (2, 5). The mainstream treatment for MSCs is surgery, followed by post-operative radiotherapy (6, 7), which results in significant disfigurement and function impairment (2). Recently, superselective intra-arterial infusion of high-dose cisplatin with concomitant radiotherapy (RADPLAT) has been performed for patients with MSC as the definitive treatment option, with feasible treatment outcomes and acceptable rates of acute and late toxicity (8, 9).
MSCs are often irregularly shaped, and tumors surround or involve various critical organs at risk (OARs), such as the lens, optic nerves, and retinas. Owing to advances in radiotherapy technology, volumetric modulated arc radiotherapy (VMAT) - in which multi-leaf collimators (MLC), gantry, and dose rate are continuously changing - has been widely introduced in clinical practice to minimize the radiation dose for OARs without compromising target coverage (10, 11). Jeong et al. demonstrated that VMAT plans provided better homogeneity and target conformity with equal or better OAR sparing compared to conventional intensity-modulated radiotherapy (IMRT) (12). Moreover, the use of non-coplanar beam arrangement (NC-VMAT) has the advantage of improving target coverage without giving up OAR sparing (13).
Automated NC-VMAT, so-called HyperArc (HA, Varian Medical System, Palo Alto, CA, USA), has been developed to generate a steep dose gradient with minimal workload (i.e., automatic determination of the collimator angle in the optimization process, movement of the treatment couch, and dose delivery during treatment). A number of investigators have reported that HA plans provided a higher conformity dose for the target and a lower radiation dose for surrounding tissues compared to VMAT plans for patients with brain metastases (14). Recently, HA has been applied to extra-cranial tumors, such as angiosarcoma of the scalp (15) and head and neck cancers (16, 17). Its superior dose distribution could form a new treatment approach. However, none of the studies on the performance of HA, investigated whether doses to the OARs could be reduced while maintaining target coverage for patients with MSC.
Our present study aimed at comparing dosimetric parameters for targets and OARs in the different VMAT, NC-VMAT, and HA plans for patients with MSC.
Patients and Methods
Patients and simulations. This retrospective study was approved by our institutional review board and included 20 patients with MSC who underwent definitive radiotherapy (RADPLAT). Table I lists patient characteristics. In the simulation, patients were immobilized using a thermoplastic mask in the supine position, and image acquisition was performed using a dual-energy computed tomography (CT) scanner (Revolution HD; GE Medical Systems, Milwaukee, WI, USA). The scanning parameters of 80/140 kVp (tube voltage), 375 mA (tube current), 0.7 s/rotation (gantry rotation time), and 0.984:1 (helical pitch) were used (18). Virtual monochromatic images were reconstructed at 77 keV using a slice thickness of 2 mm.
Patient characteristics.
Treatment planning. Treatment plans were generated using the Eclipse treatment planning system (Varian Medical Systems, Palo Alto, CA, USA) based on the acquired images. Three types of planning target volumes (PTVs) were generated: high (PTV1), intermediate (PTV2), and low (PTV3) risk. PTV3 was not detected in five patients. A sequential simultaneous integrated boost technique (19) was used to deliver 70.4 Gy/32, 64 Gy/32, and 46 Gy/23 fractions of PTV1, PTV2, and PTV3, respectively. In our clinical practice, HA plans were generated using three to four arcs, and the collimator angle was optimized for each beam angle (Figure 1). A linear accelerator equipped with an MLC of 2.5 mm leaf width was used for treatment and a photon beam energy of flattening-filtered 6 MV or flattening-filter-free 6 MV was used. The dose constraints for OARs were as follows: maximum dose (Dmax) of the brainstem < 54 Gy, lens (at least contralateral) < 5 Gy, mandible < 75 Gy, optic chiasm < 50 Gy, optic nerves < 50 Gy, retinas < 50 Gy, spinal cord < 45 Gy, and mean dose (Dmean) of parotid < 26 Gy. In the optimization process, sparing the brainstem, spinal cord, optic chiasm, optic nerves (at least contralateral), and retinas (at least contralateral) were considered to be the highest priority, and the target coverage was compromised. In cases where the PTV involved the OARs and radiation oncologists did not allow under-dosage of the PTV, exceeding the dose limit was tolerated by the ipsilateral optic nerves, retinas, and lens.
Comparison of the beam arrangement and dose distribution between volumetric modulated arc therapy (VMAT), non-coplanar VMAT (NC-VMAT), and automated NC-VMAT (HyperArc, HA) plans.
VMAT and NC-VMAT were retrospectively generated to assess the performance of the HA plans (Figure 1). Dual full arcs were used for the VMAT plans, and dual non-coplanar half arcs (couch angle of 90° or 270°) were used for the NC-VMAT plans. The collimator angle was determined to be either 30° or 330° (default setting) for both plans, and the normal tissue objective (NTO) was used in the optimization process, instead of SRS NTO (applied for HA plans). The other optimization parameters for the PTVs and OARs were the same as those in the HA plans.
Data analysis. Dosimetric parameters were calculated for PTVs and OARs. such as DX% (Gy), Dmin (Gy), and VXGy (cm3) represent the dose delivered to X% of the given volume, minimum dose, and absolute volume receiving more than X Gy, respectively. For PTV1, the homogeneity index (HI) and conformity index were calculated as follows: HI=Dmax/Dpres and CI=Vpres/TV, where Dpres, Vpres, and TV indicate the prescription dose, isodose volume of the prescription dose, and the target volume, respectively (20, 21). Doses for PTV2 and PTV3 were evaluated by subtracting the higher dose PTV structures from the lower dose PTV structures. The dosimetric parameters of the ipsilateral organs (lens, optic nerve, retina) were not evaluated when these organs were involved in the PTVs that were not allowed to be under-dosed. Subsequently, the monitor units (MUs) required for VMAT, NC-VMAT, and HA were recorded.
Results are expressed as the mean±standard deviation. The normality of the dosimetric parameters and MUs was tested using the Shapiro-Wilk test. To measure the statistical difference in the VMAT, NC-VMAT, and HA plans, the paired Student t-test and Wilcoxon signed-rank test were performed for normally distributed and non-normally distributed data, respectively. All statistical analyses were performed using the SPSS software (IBM, Armonk, NY, USA), and statistical significance was set at p<0.05.
Results
Table II shows the dosimetric parameters for the PTVs and entire body region. For PTV1, three treatment plans provided comparable dosimetric parameters except for Dmax (75.7±1.1 vs. 75.3±1.1 Gy for VMAT and HA plans, respectively, p<0.05) and HI (1.08±0.02 vs. 1.07±0.02 Gy for VMAT and HA plans, respectively, p<0.05). For PTV2, the D95%, D99% and Dmin in the HA plans (64.7±1.2, 62.7±2.1 and 54.6±6.2 Gy, respectively) were significantly higher (p<0.05) than those in the VMAT plans (64.3±1.7, 61.9±2.4 and 52.9±6.4 Gy, respectively). The NC-VMAT and HA plans provided significantly higher (p<0.05) dosimetric parameters for PTV3 than those in the VMAT plans. D99% in the HA was significantly higher than that in the NC-VMAT plans (52.5±3.0 vs. 51.8±2.7 Gy, respectively, p<0.05). In contrast, the volumes receiving 20-40 Gy in the entire body were smaller in the HA plans (295±126, 398±167, and 630±257 cm3 for V40Gy, V30Gy and V20Gy, respectively) compared to those in the VMAT (340±152, 494±213, and 814±323 cm3) and NC-VMAT plans (314±139, 426±184, and 679±274 cm3).
Dosimetric parameters for targets and body region.
Table III shows the dosimetric parameters for OARs. For the brain, the HA plans provided significantly lower V40Gy, V30Gy, and V20Gy than the VMAT and NC-VMAT plans (p<0.05). In addition, the Dmax of the brainstem and mandible was significantly lower (p<0.05) in the HA plans (23.4±8.7 and 58.0±15.3 Gy) compared to that in VMAT (27.1±8.7 and 59.9±13.5 Gy) and NC-VMAT plans (24.7±8.2 and 60.2±13.4 Gy). Comparable values between the three treatment plans were obtained for the ipsilateral organs (lens, optic nerve, retina, and parotid), optic chiasm, and spinal cord. For contralateral organs, the Dmax of the lens, optic nerve, and retina, as well as Dmean of the parotid in the HA plans (5.4±4.0, 19.6±11.4, 14.3±7.7, and 3.9±1.9 Gy, respectively) were significantly lower (p<0.05) than those of the VMAT plans (6.9±5.3, 24.0±13.4, 18.7±11.8, and 5.6±3.1 Gy, respectively). Moreover, the Dmax of the optic nerve was significantly lower (p<0.05) than that of the NC-VMAT plans (21.2±12.1 Gy).
Dosimetric parameters for organs at risk (OARs).
Figure 2 shows a comparison of the MUs required for the VMAT, NC-VMAT, and HA plans. The VMAT (658±104 MU) and HA plans (665±144 MU) required significantly higher (p<0.05) MUs than the NC-VMAT plans (611±93 MU).
Comparison of the MUs required in the volumetric modulated arc therapy (VMAT), non-coplanar VMAT (NC-VMAT), and automated NC-VMAT (HyperArc, HA) plans. *Statistically significant at p<0.05.
Discussion
This study compared the dosimetric parameters of the VMAT, NC-VMAT, and HA plans for patients with MSC. Because of the rarity of cases, most hospitals lack experience in treating patients with MSCs, which results in missing solid evidence to determine the optimal treatment strategy. Surgical resection followed by postoperative radiotherapy has been widely implemented, with a 5-year overall rate of 48-62% (6, 22, 23). Jang et al. reported that the treatment outcome of definitive radiotherapy with or without chemotherapy for patients with MSCs was insufficient, with a 5-year overall rate of 34% (5). In Japan, RADPLAT has been applied to patients with MSCs, while western countries applied RADPLAT for patients with head and neck cancers other than MSCs (24). A favorable 5-year overall survival of 55-69% has been reported by several institutions (2, 8, 9), and RADPLAT has the potential to be a new, less invasive alternative to surgery followed by postoperative radiotherapy. Thus, sophisticated radiotherapy techniques are required to maximize the treatment effect and minimize the risk of radiation-induced side effects.
Bristol et al. compared the outcomes of patients with MSCs who underwent postoperative radiotherapy between 1969-1991 and 1991-2002. Various technologies, such as staging (X-ray image, CT, and magnetic resonance imaging) and radiotherapy (60Co, electron, and 6-MV photons) were updated in the later groups (22). Nevertheless, there have been many improvements in the radiotherapy technique; however, the treatment outcome for this disease remains almost unchanged. The authors concluded that newer strategies, including dose escalation with IMRT, are required to improve treatment outcomes. Dirix et al. compared the treatment outcomes of sinonasal cancer between patients treated with postoperative three-dimensional conformal radiotherapy (1992-2002) and IMRT (2003-2008) (7). They found that IMRT significantly improved disease-free survival and reduced acute and late toxicity, and concluded that IMRT should be considered as a standard treatment technique. Orlandi et al. demonstrated that VMAT outperformed IMRT with respect to target coverage, and NC-VMAT may be advantageous for patients with critical extensions of disease with large volumes and/or irregular shapes (13). In this study, we first demonstrated that HA provided further improvement of the target coverage, reduction of the dose spread (V20Gy-V40Gy for the entire body), and OAR sparing compared to VMAT and NC-VMAT (Table II). Therefore, HA has the potential to be a standard treatment plan in definitive radiotherapy for patients with MSCs, and its clinical benefits should be established in future studies. For head and neck cancers, HA has been demonstrated to be an effective treatment with a highly conformal dose distribution and excellent OAR sparing. Additionally, radiation therapists did not need to enter the treatment room to rotate the treatment couch even in the non-coplanar beam arrangement (16, 25, 26).
MSCs are surrounded by several OARs, where the doses for these organs should be minimized. Tumor dose coverage is often constrained by the radiation tolerance of the optic apparatus, and the risk of radiation-induced side effects, particularly visual impairment, increases markedly at doses ≥60 Gy (27). However, Hoppe et al. reported that a biologically equivalent dose of radiation ≥65 Gy was a significant factor for improving overall survival in patients with paranasal sinus carcinoma (28). A reduction in salivary function is one of the major radiation-induced side effects that reduces the patient’s quality of life (QOL). Poor dental hygiene, oral pain, and difficulty in chewing and swallowing can be caused by insufficient salivary function (29). According to an investigation by Buus, a lower dose of in the parotid gland resulted in better parotid function (30), and thus, doses to the parotid should be as low as possible. Patients frequently complain of fatigue during and after radiotherapy (31), and doses to the central nervous system structures are related to these symptoms. Gulliford et al. reported that the maximum and mean doses to the brainstem and cerebellum were significantly higher in patients with acute fatigue of grade 2 or higher than in those without such symptoms (32). In this study, we demonstrated that HA plans reduced doses to the vision system, salivary glands, and central nervous system compared with VMAT and NC-VMAT (Table III), and that HA plans have the potential to reduce various radiation side effects in patients with MSC.
There are limitations to this study. First, the number of patients was limited, and our data could not support an in-depth analysis, such as the effect of PTV volume and width of the MLC. Second, patients without neck lymph node metastases were investigated. However, Homma et al. showed that 21.9% of patients with T4 MSC had lymph node metastasis at diagnosis (33), and a larger volume was irradiated for these patients (34). Third, the same optimization parameters, except for the SRS NTO, were used in the VMAT, NC-VMAT, and HA plans, while better plans might be generated by using different parameters. Despite these limitations, our quantitative data can provide useful information for improving target coverage and reducing OAR doses in definitive radiotherapy for MSC patients.
In conclusion, our results demonstrated that HA plans provided better target coverage (D95%, D99% and Dmin) compared to VMAT and NC-VMAT plans. Moreover, the doses in HA plans for the contralateral vision system, contralateral salivary gland, and central nervous systems were the smallest. We therefore believe that HA plans can form a new choice for dose delivery to patients with MSC.
Acknowledgements
This study was supported by JSPS KAKENHI Grant (Grant-in-Aid for Scientific Research (C) 21K07742).
Footnotes
Authors’ Contributions
All the Authors participated in the writing of this article and take responsibility for its content. The Authors confirm that the content of the manuscript has not been published, or submitted for publication elsewhere.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
- Received September 28, 2022.
- Revision received October 19, 2022.
- Accepted October 24, 2022.
- Copyright © 2023, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).