Introduction
About 30% of non-small cell lung cancer (NSCLC) cases are evaluated as locally advanced at the time of diagnosis. Reference Chen, Ruiz and Hsieh1 This definition includes a broad spectrum. While defining T3–T4 and N2–N3 disease, TNMs (8th edition) Reference Rami-Porta, Asamura and Travis2 other than stage 3 A–B are also generally defined in stage 3C (T3–T4 and N3). Reference Rami-Porta, Asamura and Travis2 Chemoradiotherapy (CRT) is often used in the treatment of locally advanced NSCLC. Reference Postmus, Kerr and Oudkerk3
The lung itself is sensitive to radiation, and in applying radiotherapy to the thorax region, it is difficult to give the desired dose to planning target volume (PTV) without exceeding the dose to organs at risk (OARs) due to the close proximity of these organs, such as the spinal cord, oesophagus and heart. Reference Ozawa, Koiwai, Matsushita and Kadoya4 Therefore, modern techniques such as three-dimensional conformal radiation therapy (3DCRT), intensity-modulated radiation therapy (IMRT), volumetric-modulated arc therapy (VMAT) and helical tomotherapy (HT) are used to overcome these problems.
Compared to 3DCRT, IMRT was found superior in terms of dose conformity and OARs. Reference Murshed, Liu and Liao5 VMAT has better conformal target coverage than IMRT and provides lower doses in OARs, lower monitor units (MU) and shorter treatment time. Reference Pasler, Georg, Bartelt and Lutterbach6 HT, another novel approach of IMRT that can deliver radiotherapy (RT) with rotational fields, is increasingly used in lung cancer. Reference Zhu and Fu7
In our study, we wanted to undertake a dosimetric comparison of hybrid (3DCRT-VMAT) and HT techniques in terms of both PTV coverage and the doses received by the OARs in locally advanced NSCLC patients.
Patients and Methods
Patient selection
This was a retrospective study of 15 patients who received radiotherapy between January 2019 and December 2019 in our institution with thoracic radiotherapy for locally advanced NSCLC. Including criteria were having a large sized tumour and positive mediastinal lymph nodes.
Simulation and contouring of targets and OARs
Computed tomography (CT) datasets with a 3-mm slice thickness was taken from all patients in supine position and hands raised above the head at the T-board from C2–C3 level to L3–L4 level. The scans were performed under free breathing.
The images were sent to Velocity Contouring Station (version 2·8·1, the USA). Clinical target volume (CTV) and OARs were contoured by the same radiation oncologist in line with The Radiation Therapy Oncology Group (RTOG) Atlas. 8 The CTV was defined as the gross tumour volume +0·5 cm margin. OARs dose constrains were determined based on the RTOG 1306 study. Dose constrains were as follows: lungs: V 20 ≤ 37%, mean dose ≤ 20 Gy; spinal Cord: 0·03 cc ≤ 50·5 Gy; heart: V 60 < 1/3, V 45 < 2/3, V 40 < 100%; oesophagus: mean dose ≤ 34 Gy. 9
Treatment planning
PTV was obtained by adding 1 cm margin to the CTV. Prescription dose was determined as 60 Gy at 30 fractions. Dose calculation was normalised to cover 95% of the PTV. A volume of 0·03 cc within any PTV should not receive >110% of the prescribed dose. No more than 0·03 cc of PTV received <93% of its prescribed dose. None of the 0·03 cc or more volumes except for PTV were allowed to receive >110% of the prescribed dose of PTV.
Varian Eclipse planning system (version 13·7-Varian Medical Systems, Palo Alto, the USA) was used in hybrid planning, and the Tomotherapy Planning System (Hi-Art Tomotherapy, version 5·1·2, Accuray, Madison, WI, the USA) was used in HT planning. For hybrid plans, the structures contoured at Velocity Contouring Station were transferred to the Eclipse Planning System DICOM (Digital Imaging and Communication in Medicine) format.
The Isocenter was determined as the midpoint of the PTV volume. Anisotropic analytical algorithm photon dose calculation algorithm was used for VMAT plans. Maximum dose rate was 600 MU/min for the VMAT plan and 400 MU/min for the 3DCRT. The dose calculation grid was 2·5 mm.
In VMAT plans, two half 180° arcs or two full 360° arcs, depending on target localisation, were used. In all 3DCRT plans, two opposing fields were used. PO (Photon Optimizer version 13·7) algorithm was used for optimised leaf position, dose rate and gantry speed. The collimator was rotated 30°, 90° and 330° to reduce overlapping tongue and groove effects.
In hybrid-VMAT techniques, 120-leaf (central 20 cm of field uses 0·5-cm-wide leaves, outer field uses 1-cm-wide leaves) dynamic multi-leaf collimator was used. The maximum leaf speed was 2·5 cm/s. Then, 6 MV energy was used in all VMAT plans; 15 MV energy was used in all conformal plans.
For HT plans, a field width of 2·5 cm, pitch value of 0·287, modulation factor of 3 and fine dose calculation grid was used. Then, 6 MV energy was used in all HT plans.
Evaluation tools
Plan evaluation was performed by examining all CT slides one by one and by examining the dose-volume histogram (DVH) isodose curves. Sample isodose curve is shown in Figure 1.
Figure 1. Isodose curves of one patient in axial-sagittal and coronal section for (a) hybrid plan; (b) helical tomotehrapy plan.
The homogeneity index (HI) was calculated as HI = D 2-D 98/D p , where D 2 is the minimum dose to 2% of the target volume, D 98 is the minimum dose to the 98% of the target volume and D p is the prescribed dose. This is the most commonly used formula in the literature. Equation 1 shows that lower HI values exude a more homogeneous target dose. Reference Kataria, Sharma and Subramani10
The conformity index (CI) was calculated as:
$${{\rm{CI = }}\left( {{\rm{Cover \;Factor}}} \right){\rm{ \;X\; }}\left( {{\rm{Spill \;Factor}}} \right){\rm{ = }}{{V95PTV} \over {VPTV}}X{{V95PTV} \over {V95BODY}}}$$
where V 95PTV and V 95body are the volumes of the PTV and body, respectively, receiving at least 95% of the prescription dose, and VPTV is the volume of the PTV. The CI equal to 1 correlated with the ideal dose coverage or high conformity. Reference MacFarlane, Hoover and Wong11
The integral dose was defined as mean dose (Gy) × volume (L). Reference Aoyama, Westerly and Mackie12
Results
Of our patients, 14 were male and 1 was female. The mean age was found to be 66 years (range: 53–76). Mean PTV volume was 602·5 cc (range: 265–1461). Mean total lung volume was 4264 cc (range: 1885–6803). Patient characteristics are shown in Table 1.
Table 1. Patient characteristics
SCC: squamous cell carcinoma; AC: adenocarcinoma.
There was no difference between CI. While HI, D mean, D max, D 2 and V 105 were found to be lower in HT, V 100, total MU and total beam on time were found to be lower in the hybrid plan. PTV coverage and treatment parameters are shown in Table 2.
Table 2. Comparison of dosimetric parameters
The bold values’ significance level is p = 0.05.
Total lung
D mean was found to be 17 Gy in both techniques. V 10 value was 42·85 in the hybrid plan and 48·67 in HT (p = 0·037). The other parameters were found to be similar.
Ipsilateral lung
D mean was 27 Gy in the hybrid plan and 25 Gy in HT (p = 0·005), and V 20 doses were 53·8 and 49·3, respectively (p = 0·029). V 30 doses were 44·04 and 37·02, respectively (p = 0·004).
Contralateral lung
V 20 value was 9·3 in the hybrid plan and 14·4 in HT (p = 0·019). D mean, V 10 and V 30 values were statistically lower in the hybrid plan. V5 value was also lower in the hybrid plan, but no statistically significant difference was found.
Heart
D mean was 14·5 Gy in the hybrid plan and 18·7 in HT (p < 0·001), and V 30 values were 18·1 and 22·9, respectively (p = 0·009). No statistically significant difference was found in the other parameters.
Spinal cord
D max was 42·2 Gy in the hybrid plan and 41·4 (p = 0·680) in HT.
Oesophagus
D mean was 23 in the hybrid plan and 27·6 in HT (p < 0·001), and the V 35 values were 33·7 and 40·6, respectively (p < 0·001). V 40 and V 60 values were lower in the hybrid plan, while D max was similar in both techniques.
OARs doses are shown in Table 3.
Table 3. Comparison of OAR’s dose volume parameters
The bold values’ significance level is p = 0.05.
Statistical method
Data are represented by mean and standard deviation. Comparisons between two methods were performed by paired samples t-test. In all analyses, significance level was considered as 0·05. IBM SPSS Statistics for Windows version 22·0 (NY, USA) was used for the analysis.
Discussion
In our study, 2 different treatment plans (hybrid-VMAT and HT) of 15 patients diagnosed with locally advanced NSCLC were compared. The plans made with both techniques are suitable for treatment. HT achieved better HI, higher V 100 (%) and lower V 105 (%). However, total MU and beam on time were found to be higher in HT.
Total lung V 20 and the mean lung dose are the best known parameters for radiation pneumonia. Reference Graham, Purdy and Emami13 When the V 20 value was kept below 20%, symptomatic pneumonia risk was 18%, and fatal pneumonia risk was 0%, and when the V 20 value exceeded 30%, symptomatic pneumonia risk and fatal pneumonia risk increased to 32·6% and 2·9%, respectively. Reference Palma, Senan, Tsujino, Barriger, Rengan and Moreno14 In the study of Kristensen et al., 10 Gy total lung volume was found significant in terms of radiation pneumonia. Median V 10 value was 60·3% in patients who developed pneumonia and 52·6% in those who did not develop pneumonia (p = 0·02). Reference Kristensen, Nottrup and Berthelsen15 In our study, lung parameters were suitable for dose constrains, and total lung V 10 value was 42·8 in the hybrid plan and 48·6 in HT (p = 0·037). The other parameters were similar. In the hybrid plans, the opposite lung D mean, V 10, V 20 and V 30 values were lower than HT. Due to the mutually conformal areas, the opposite lung could be better protected. In a similar dosimetric study, HT was compared with VMAT and IMRT. Compared to IMRT, mean lung V 20 and V 30 values were lower in HT, but no difference was found between HT and VMAT (V 20:21·8%, 22·2%, 24·2%, respectively; V 30:15%, 15·7%, 16·7%, respectively). Since the mean PTV volume in the study was quite low compared to our PTV volume, the lung values were higher in our study (312 cc versus 602·5 cc). Reference Xu, Deng and Yang16
Cardiac toxicity due to RT is a major problem, and cardiac doses should be kept as low as possible. According to the available evidence-based data, it seems that the most relevant heart dose constraints are V 30 < 50% and V45 < 35%. Reference Vojtíšek17 Speirs et al. found heart V 50 dose to be the strongest predictor (V 50: volume receiving ≥ 50 Gy). When stratified by heart V 50 less than 25% versus 25% or greater, the 1-year overall survival (OS) rates were 70·2% versus 46·8%, and the 2-year OS rates were 45·9% versus 26·7%, respectively (p < 0·0001). Reference Speirs, DeWees and Rehman18 In our study, the doses given to the heart were below the limits in both techniques. D mean and V 30 values were lower in the hybrid plan. In the dosimetric study of Xu et al., the mean heart D mean was found to be 18·2 Gy in HT, 12·7 Gy in VMAT and 12·5 Gy in IMRT in patients with PTV ≥ 312 cc (p = 0·04 for HT versus IMRT). Reference Xu, Deng and Yang16 In our study, D mean was found to be 14·5 Gy in the hybrid plan and 18·7 Gy in HT (p < 0·001).
Spinal cord D max doses are generally tried to be kept ≤50 Gy. It can be difficult to provide these doses with 3DCRT. This problem can be overcome when 3DCRT and VMAT are combined. Reference Kim, Lee and Kang19 In our study, spinal cord D max was found to be suitable in both the hybrid plan and the tomo plan.
Radiation oesophagitis (RE) is an acute and chronic toxicity that significantly impairs quality of life and increases even more when RT is applied simultaneously with chemotherapy. Various DVH parameters are used to predict RE. Mean oesophageal dose (MED), maximal oesophageal dose, V 20, V 35 and V 60 are some of them. Reference Huang, Bradley and El Naqa20 In a metanalysis, V 60 emerged as the best predictor for both moderate and severe RE. Three groups were determined according to V 60 doses: low risk (V 60 < 0·07%), intermediate risk (V 60 between 0·07% and 16·99%) and high risk (V 60 ≥ 17%) risk. Reference Palma, Senan and Oberije21 While the mean oesophagus dose was achieved as ≤34 Gy in both techniques, V 60 dose was found to be 17·6% in tomotherapy. D mean, V 35, V 40 and V 60 doses were found to be lower in the hybrid plan.
The main limitation of this study was the variety of tumour size and localisation. This could be eliminated by initiating a long-term project with a higher number of patients included.
Conclusion
Suitable dose coverage and OAR doses can be provided with both techniques. Especially the opposite lung, heart and oesophagus doses can be kept lower with the hybrid plan, and lower MU and shorter beam on time can be provided.
Acknowledgements
None.
