Small-cell lung cancer (SCLC) accounts for approximately 13% of the newly diagnosed lung cancers worldwide [1]. Brain metastases (BM) are a very common metastatic site in SCLC. That is to say, more than 10% of patients have BM at initial diagnosis; more than 50% develop BM within 2 years; and up to 80% are found with BM at autopsy [2-4]. Its frequent occurrence and the inadequacy of chemotherapy to prevent BM due to the blood–brain barrier has brought prophylactic cranial irradiation (PCI) to the agenda. After the meta-analysis of Aupérin et al. [5] showing that PCI decreased the BM risk by 25% and increased the 3-year survival by 5.4% in SCLC, PCI was accepted as the standard in limited-stage disease with a complete response to systemic treatment. In extensive-stage disease, PCI recommendation was initiated after the Phase 3 study of Slotman et al. [6]. However, the risk of developing brain metastasis after PCI remains, and reducing it requires the irradiation of the entire brain tissue with a homogeneous effective dose. With regard to BM development, if a homogeneous dose distribution on the target and high protection in critical organs are provided in PCI, the second serial irradiation becomes greatly advantageous as a treatment opportunity. While palliative doses were previously used in re-irradiation, stereotactic radiosurgery/radiotherapy (SRS/SRT) has been an important treatment option for more than 10 to 15 years now [6-10]. Kirakli et al. [10] reported a 10% re-irradiation rate in patients who underwent PCI for SCLC. Similarly, Bernhardt et al. [7] reported an 11.9% rate, while Slotman et al. [6] determined a 14.6% rate. The biggest limiting factor in re-irradiation after PCI irradiation is critical organ doses. In the past, the target area and the surrounding healthy tissues and organs received doses equal to or close to the prescribed dose in the three-dimensional conformal radiotherapy (3D CRT) technique for whole-brain irradiation. In the last 15 to 20 years, intensity-modulated radiotherapy (IMRT) and volumetric modulated arc therapy (VMAT) have been used for more homogeneous whole-brain irradiations with a higher critical organ protection, becoming more suitable treatment options for second-series irradiations. Instead of adjusting the intensity, the dynamic conformal arc therapy (DCAT) technique uses a multileaf collimator (MLC) to cover the target and block the critical organ when necessary, thereby offering a conformal treatment modality similar to intensity-modulated treatments with a larger and lower number of segments. This study investigated the DCAT technique frequently used in SRS/SRT treatments to determine the dose distribution and the advantages of DCAT in large-volume targets, such as the whole brain.

In the present work, we aim to make a dosimetric comparison of 3D CRT, which is a traditional PCI method, and the IMRT and VMAT techniques, which have been used for the last 15 to 20 years, with the DCAT technique.

The tomography data of 15 patients with SCLC who underwent radiotherapy for PCI were used. A thermoplastic head mask was used to immobilize our patients. Tomography images with slice intervals of 0.3 cm were scanned using Biograph mCT (Siemens Medical Solutions, Erlangen, Germany). The images were then transferred to the Prosoma 4.1 (MedCom, Darmstadt, Germany) contouring station. The same radiation oncologist contoured the whole brain and the normal tissues. Each patient underwent four different plans in the Monaco (Monaco Ver. 5.51; Elekta CMS, Stockholm, Sweden) treatment planning system. In all the planning, the prescription dose was set as 25 Gy in 10 fractions using X-ray with a 6 MV energy. The Monte Carlo dose calculation algorithm was used for the dose calculation. The grid space was 0.25 mm. The statistical uncertainty was 1%.

In the 3D CRT planning, a mutually parallel (gantry angles: 90° and 270°) MLC with the lower border passing through the second cervical vertebra and no collimator angle was used to cover the entire brain volume with lens protection as much as possible.

In the IMRT planning, seven fields were used in the plans prepared according to the sliding-window technique. The field gantry angles were 180°, 230°, 280°, 330°, 30°, 80°, and 130°. The minimum segment thickness was 0.75 cm. In the VMAT planning, a single 360° field was planned as dual-arc VMAT. The minimum segment thickness was 0.75 cm. In the DCAT plans, a single 360° field was used. The variable dose rate (VDR) was selected. The MLC margin for the target was 0–1 mm. The MLC blocking margin for the risky organs was 0–1 mm.

The VMAT technique was based on the inverse optimization principle [11]. In addition, the dose distributions were obtained by dynamically and simultaneously optimizing the MLC position and speed, gantry speed, and dose rate. Thus, planning was done by adjusting the beam intensity with hundreds of small segments. In contrast to VMAT, the sliding-window IMRT technique was applied only at fixed gantry angles. Meanwhile, the DCAT treatment was performed by adjusting the MLC position, speed, and field aperture to cover the tumor during the gantry rotation. Unlike VMAT, instead of the MLCs modulating the beam, blocking was used for the target coverage and the protection of organs around the target volume (Fig. 1). To improve the DCAT efficacy, the Monaco (Elekta CMS) treatment planning system allows planning as modified DCAT by integrating segment shape optimization (SSO) and VDR with inverse optimization. In other words, modified DCAT is a delivery technique based on inverse planning, where the MLC is shaped according to the target projection and allows the optimization of the gantry speed, dose rate, and segment shape [12].

Figure 1. Field aperture and multileaf collimator (MLC) blocking representation from 14 different projections in the dynamic conformal arc therapy (DCAT) technique.

In all prepared plans, 95% of the planned target volume of the brain (PTV_Brain) was normalized to receive the prescribed 25 Gy dose. For each treatment plan, the dose volume histograms were used to compare the PTV_Brain minimum (D₉₉), maximum (D₁), and heterogeneity index (HI) dose values and critical organ doses, including the right and left lens D_mean dose values and the right and left optic nerve D_maximum and D_{0.35 cc} dose values. The monitor unit (MU), beam-on-time, and V₂, V₅, and V₁₀ dose values of the body within the tomography field of view were also compared. The formula below was used for the HI in PTV_Brain, where D_5% and D_95% are the dose values received at 5% and 95% of the target volume, respectively. The ideal HI value was 1. This value moved away from 1 as the plan became less homogeneous [13]:

All results were analyzed using Friedman’s two-way statistical analysis. A statistical analysis of the four planning techniques in pairs was also performed using the Wilcoxon signed-rank test. A P<0.05 value was considered significant in all the statistical analysis results.

The highest values at the D₉₉ doses in PTV_Brain were obtained using the VMAT technique (P<0.001). The similar results were acquired using the IMRT and VMAT techniques. The lowest result for the D₁ dose value was obtained with the IMRT (P=0.211) planning. The lowest results for the HI values were obtained using the VMAT technique (P<0.001). The most homogeneous results in PTV were obtained by VMAT (P=0.049). Plans with a homogeneity close to VMAT were obtained with DCAT (Table 1).

Table 1 . Target and critical organ doses.

		3D CRT	IMRT	VMAT	DCAT	P-value
PTVBrain	D99	2,344.38±22.14	2,389.6±32.63	2,384.6±30.8	2,360.59±31.82	<0.001
	D1	2,762.56±47.60	2,716.6±186.1	2,748.3±25.5	2,753.88±26.38	0.211
	HI	1.10±0.011	1.10±0.0053	1.092±0.009	1.096±0.009	0.049
L optic nervous	Dmaximum	2,634.37±53.10	2,507.88±80.246	2,463.8±55.9	2,629.72±40.52	<0.001
	D0.35 cc	2,352.22±335.088	1,962.2±240.88	1,947.4±267.5	2,419.4±183.34	<0.001
R optic nervous	Dmaximum	2,641.093±54.94	2,512.19±78.35	2,469.02±69.12	2,623.58±44.79	<0.001
	D0.35 cc	2,380.78±275.89	2,025.33±172.74	1,979.88±324.4	2,437.38±140.73	<0.001
L lens	Dmean	233.58±32.08	423.11±47.34	327.32±14.89	446.43±23.80	<0.001
R lens	Dmean	249.54±65.88	427.08±50.83	328.06±16.92	442.90±22.69	<0.001

Data are presented as mean±standard deviation..

3D CRT, three-dimensional conformal radiotherapy; IMRT, intensity-modulated radiotherapy; VMAT, volumetric modulated arc therapy; DCAT, dynamic conformal arc therapy; PTV_Brain, planned target volume of the brain; D, dose; HI, heterogeneity index; L, left; R, right..

A comparison of the critical organ doses showed that lower values were obtained using the VMAT technique (P<0.001) at the left optic nerve D_maximum and D_{0.35 cc} dose values. Similarly, lower values were acquired by VMAT (P<0.001) in the right optic nerve D_maximum and D_{0.35 cc} dose values. Lower values were determined with the 3D CRT planning (P<0.001) in left and right lens D_mean dose values.

Lower values were obtained with the 3D CRT planning (P<0.001) in the V₂, V₅, and V₁₀ dose values of the body. In addition, similarly high values were acquired by the VMAT and IMRT techniques, while lower ones were obtained by DCAT and 3D CRT when compared to other planning techniques (Fig. 2).

Figure 2. V₂, V₅, and V₁₀ dose values of the body according to the four different planning techniques. 3D CRT, three-dimensional conformal radiotherapy; IMRT, intensity-modulated radiotherapy; VMAT, volumetric modulated arc therapy; DCAT, dynamic conformal arc therapy.

The lowest MU values were obtained with DCAT (P<0.001). Similar results were obtained by DCAT and 3D CRT (P=0.256). The MU values were 66% (P=0.001) lower than that obtained by IMRT, and 72% (P=0.001) lower than that determined by VMAT. The lowest results were obtained using the 3D CRT (P<0.001) planning in the beam-on-time values. A 3.2% difference was found between 3D CRT and DCAT (P=0.008). The beam on-time values obtained by the 3D CRT planning were 76% (P=0.001) lower than those acquired with IMRT planning and 65% (P=0.001) lower than those determined with VMAT planning (Fig. 3, Table 2).

Table 2 . Wilcoxon signed-rank statistical analysis of the pairwise comparisons of the four planning techniques.

		VMAT vs. DCAT	VMAT vs. 3D CRT	VMAT vs. IMRT	DCAT vs. 3D CRT	DCAT vs. IMRT	3D CRT vs. IMRT
MU		0.001	0.001	0.001	0.256	0.001	0.001
Beam-on-time		0.001	0.001	0.001	0.008	0.001	0.001
Body	V2	0.003	0.001	0.011	0.001	0.008	0.001
	V5	0.001	0.001	0.820	0.001	0.001	0.001
	V10	0.009	0.001	0.011	0.001	0.001	0.001
PTVBrain	D99	0.015	0.005	0.609	0.140	0.041	0.002
	D1	0.307	0.334	0.112	0.650	0.307	0.233
	HI	0.302	0.061	0.021	0.083	0.124	1.000
L optic N.	Dmaximum	0.001	0.001	0.023	0.865	0.002	0.002
	D0.35 cc	0.001	0.001	0.017	0.496	0.001	0.001
R optic N.	Dmaximum	0.001	0.001	0.064	0.427	0.001	0.001
	D0.35 cc	0.001	0.001	0.099	0.551	0.001	0.001
L lens	Dmean	0.001	0.001	0.001	0.001	0.233	0.001
R lens	Dmean	0.001	0.005	0.001	0.001	0.609	0.001

VMAT, volumetric modulated arc therapy; DCAT, dynamic conformal arc therapy; 3D CRT, three-dimensional conformal radiotherapy; IMRT, intensity-modulated radiotherapy; MU, monitor unit; PTV_Brain, planned target volume of the brain; D, dose; HI, heterogeneity index; L, left; R, right..

Figure 3. Monitor unit (MU) (a) and beam-on-time (b) values according to the four different planning techniques. 3D CRT, three-dimensional conformal radiotherapy; IMRT, intensity-modulated radiotherapy; VMAT, volumetric modulated arc therapy; DCAT, dynamic conformal arc therapy.

Our study investigated the applicability of the DCAT technique generally used for radiosurgery treatments only in small and uniformly shaped target volumes in large-sized PCI patients. VDR, critical organ MLC blockade, and SSO were used to modify the DCAT technique that approximates the dose distributions obtained with the intensity-modulated treatments.

The biggest advantages of the conventional 3D CRT technique are low MU, treatment time, and scattered radiation. Compared with intensity-modulated techniques such as VMAT and IMRT, the MU and treatment time are three to four times lower. For DCAT, which is used in SRS/SRT treatments, the MU and beam-on-time values almost similar to those in the 3D CRT technique were obtained in patients with PCI (Fig. 3).

The first treatment (i.e., whole-brain irradiation) homogeneity is critical in the second series of irradiations. A more homogeneous dose distribution is achieved in the first irradiated plan. The lower the critical organ doses, the higher the doses that can be safely administered in the second series of irradiations. Therefore, obtaining a homogeneous dose distribution in PCI irradiation is desirable. Accordingly, VMAT and DCAT (P=0.302) provide more homogeneous plans and advantages for second-series irradiations (Fig. 4).

Figure 4. Display of the dose distributions in volumetric modulated arc therapy (VMAT), dynamic conformal arc therapy (DCAT), three-dimensional conformal radiotherapy (3D CRT), and intensity-modulated radiotherapy (IMRT) planning: while the prescription dose is shown with a light blue isodose line, the maximum dose points of 2,770 and 2,850 cGy are shown with red isodose lines.

The VMAT and IMRT techniques can obtain more homogeneous and conformal plans by modulating more intensities, which is more advantageous in critical organ maximum doses. Treatments, such as hippocampus sparing, can also be easily performed [14]. The hippocampus is the part of the brain mainly responsible for learning and memory [15,16]. The problems of IMRT and VMAT are their long treatment times (Fig. 3), high dosage output, low-dose irradiation of a large normal tissue volume (Fig. 4), high doses of leakage and transmission radiation, scattered radiation, and risk of a secondary cancer [17,18]. Similar to that of the 3D CRT logic, DCAT does not aim to change the beam intensity, but apply the treatment with a field aperture that surrounds the target. When the critical organ blockage is optimized using SSO and VDR features in the DCAT technique, treatment with low MU and beam-on-time values, such as 3D CRT technique is, applied.

The advantages of the DCAT technique in radiosurgery implementations in small and concave-shaped target volumes and its applicability in large and concave-shaped targets, such as the whole brain, were investigated in this work. While DCAT provides low treatment time, MU, and integral dose, such as 3D CRT in PCI patients, it could also be advantageously applied with a homogeneous dose distribution close to that in VMAT and low critical organ doses. Without sacrificing the homogeneous dose distribution and the critical organ doses in intensity-modulated therapies, three to four times lower treatment time and MU similar to 3D CRT can be achieved by DCAT. Our results show that DCAT, which can be applied in small target volumes, can also be successfully planned in large target volumes, such as the whole brain. Scalp or hippocampus spared treatments can easily be performed using the IMRT and VMAT techniques. Great difficulties are faced when planning these treatments using DCAT. In our study, DCAT was more feasible because of its VDR and MLC blockade features. Similarly, these plans are thought to be made more easily by further developing and modifying DCAT.

None.

All relevant data are within the paper and its Supporting Information files.

The authors have nothing to disclose.

Conceptualization: Ismail Faruk Durmuş. Data curation: Ismail Faruk Durmuş Dursun Esitmez. Formal analysis: Ismail Faruk Durmuş Dursun Esitmez, Guner Ipek Arslan, Ayse Okumus. Investigation: Ismail Faruk Durmuş Dursun Esitmez. Methodology: Ismail Faruk Durmuş Dursun Esitmez, Guner Ipek Arslan, Ayse Okumus. Project administration: Ismail Faruk Durmuş. Resources: Ismail Faruk Durmuş Dursun Esitmez, Guner Ipek Arslan, Ayse Okumus. Software: Ismail Faruk Durmuş Dursun Esitmez. Supervision: Ismail Faruk Durmuş. Validation: Ismail Faruk Durmuş Dursun Esitmez, Guner Ipek Arslan, Ayse Okumus. Visualization: Ismail Faruk Durmuş Dursun Esitmez, Guner Ipek Arslan, Ayse Okumus. Writing – original draft: Ismail Faruk Durmuş Dursun Esitmez, Guner Ipek Arslan, Ayse Okumus. Writing – review & editing: Ismail Faruk Durmuş Dursun Esitmez, Guner Ipek Arslan, Ayse Okumus.