The Inuence of Lobe-based Radiotherapy Plan Optimization Method on Different Lymph Node Irradiation Schemes for Operable LA-NSCLC Patients

Purpose Various target volume delineation schemes differ greatly for stage IIIa NSCLC radiotherapy. Although tightened target volume may give patients the opportunity to receive radiotherapy, it is not absolutely safe to narrow the irradiation area. For IIIa NSCLC patients who will undergo lobectomy, a new neoadjuvant radiotherapy based on sparing preserved lung lobes may improve the dose distribution of the preserved lobe, and provide freedom for physicians in optimizing treatment strategies. Materials and methods Computed tomography imaging data of 20 IIIA- p N2 NSCLC patients were used to produce conventional IMRT(IMRT) and Preserved Lobe based IMRT(P-IMRT) plan respectively according to two different target volume delineation schemes(OPT1 and OPT2). Dose results of target coverage, Total lung, Ipsilateral lung, Preserved Total Lung, Preserved Ipsilateral Lung, Contralateral Lung, Resected Lobe and other OARs in the four groups were analyzed. Results All plans met dose limits. Lobe-based IMRT signicantly reduce the irradiated dose of Lung lobes, especially Preserved Total Lung and Preserved Ipsilateral Lung, for both delineation schemes. Mean Dose of Preserved Total Lung decreased from 819.93 cGy to 690.98 cGy (OPT1) and 542.47 cGy to 469.62 cGy (OPT2), Mean Dose of Preserved Ipsilateral Lung decreased from 1282.95 cGy to 1068.55 cGy (OPT1) and 955.83 cGy to 795.97 cGy (OPT2), respectively. While the dose indices of Resected Lobe increased slightly for only about 1%. Comparing the four groups of plans, it’s more effective in optimizing the dose of lung lobes by this method for the delineation scheme with a large target volume. The lung dose metrics in P-IMRTOPT1 can be reduced to a value very close to that in IMRTOPT2, and some values are even lower than that in IMRTOPT2. Conclusion For IIIA-N2 NSCLC patients who will undergo lobectomy, no matter which target delineation scheme is chosen, preoperative neoadjuvant radiotherapy using a lobe -based planning can signicantly reduce the radiation dose that preserves the lobes. for the large-scale lymph node irradiation scheme, this can also reduce the dose of lung lobe to a level that is comparable to lower than that of the conventional IMRT small-area lymph node irradiation scheme, and reduce the obstacles for clinicians in selecting the optimal individualized scheme. Individualized precise radiotherapy is the future development direction. Our research results show that for IIIA-N2 NSCLC patients who are going to undergo lobectomy, no matter which target delineation scheme is selected, neoadjuvant radiotherapy based on lung lobe sparing can signicantly reduce the radiation dose to the preserved lobes. Even if the scheme of large-scale lymph node irradiation is chosen, the radiation dose to the preserved lobes can be reduced to the same or even lower level as that of IMRT in small-scale lymph node irradiation, which can reduce the constraints for clinicians in selecting the optimal individualized scheme for patients.


Introduction
Radiotherapy is one of the standard treatments for LA-NSCLC locally advanced non-small cell lung cancer, but the outcome is still unsatisfactory. Local recurrence and distant metastasis are the main causes of failure. For locally advanced NSCLC cases, the target area contains primary lesions and metastatic lymph nodes, and the irradiation range is too large, which will cause excessive irradiation of normal lung, heart, esophagus, spinal cord and other important tissues and organs. At the same time, the dose of normal tissues and organs is limited, and it is di cult to increase the dose of tumors.
At present, to balance target area control and reduction of normal tissue side effects, clinicians prefer Involved eld irradiation when designing target area delineation scheme, which is more effective and safer than Elective Node Irradiation (ENI). RTOG9311 conducted a dose escalation study of 3D CRT for non-operable NSCLC. Without selective mediastinal lymph node irradiation, the dose of radiotherapy could be safely raised to 77.4 Gy (25% < V20 < 36%) or 83.8 Gy (V20 < 25%). The 2-year local control rate was 50% − 78%.
ESTRO ACROP guidelines 1 provides two choices in de ning the target area of locally advanced NSCLC: 1 lymph node stations: inclusion of the whole pathologically affected lymph node station ( Fig. 2) including at least a 5-8 mm margin around the GTV to ensure that the recurrence rate of the mediastinum in the eld is minimized; 2 geometric expansion, which ensures the target dose and reduces the normal tissue irradiation through such an IFI target delineation strategy.
Literatures 2,3 reported that nearly 20% patients diagnosed as clinical N0 and N1 were upgraded to N2 stage after mediastinal lymph node dissection,and 20.4% patients diagnosed as pathological of N0 stage were con rmed to occur micro-metastasis after immune-histochemical staining 4 . In previous studies 5 3D-CRT technology was used, although some adjacent lymph nodes were not included in the target area, they still received a slightly lower radiation dose than the target area, thus ensuring a lower recurrence rate of lymph nodes in the eld. But now IMRT and VMAT technology are widely developed, which can achieve more conformal dose distribution and more. A steep dose drop, although it allows su cient dose coverage of the target area and reduces the dose of surrounding normal tissues, may result in a lower exposure dose to out-eld lymph nodes than in the past. This requires more careful selection of CTV ln in LA-NSCLC cases, so as to avoid increasing the recurrence rate in the eld.
Previous studies based on lobe protection to reduce preoperative neoadjuvant radiotherapy dose for IIIA-N2 NSCLS showed that this method could reduce the dose of preserved lobes while ensuring target dose coverage, and increase only about 1% for V 5 -V 20 in lobe that will be resected.
In this paper, 20 cases of neoadjuvant radiotherapy of IIIA-N2 NSCLC were studied. The design method of radiotherapy planning based on preserving lung lobe protection was applied to two different target delineation schemes. The effect of the two schemes on target dose coverage and normal lung tissue irradiation dose was studied. Whether this method can be used to control lung irradiation dose while giving clinicians free of choice for appropriate target delineation strategy.

Case information
Twenty patients with lung cancer who underwent surgical lobectomy in our institution from February 2017 to November 2017 were retrospectively analyzed. All patients were pathologically proved to be stage IIIA N2 non-small cell lung cancer. The case information was shown in Table 1.

ROI Contour
Preoperative CT of 20 patients were transfered to Pinnacle 3 9.10 through PACS system. Left Lung, Right Lung, Heart, Spinal Cord were automatic contoured and can be adjusted manually if necessary. An experienced radiotherapist delineates the PTV, esophagus and lung lobe(LUL/LLL/RUL/RML/RLL) to be resected for surgery, which is then approved by the superior physician.
Target Volumes: Two sets of target areas are delineated for each case according to two target delineation strategies. Figure 1 shows the PTV delineated in one case according to these two schemes. The speci c de ning rules are as follows:  For all plans, 95% of the volume of PTV is required to reach a prescription dose of 4000 cGy.

Treatment planning
For each case, two target delineation schemes were designed. And then two plans were optimized with 6 MV photon beams for a Elekta Synergy linear accelerator (Elekta Ltd, Crawley, UK) for each scheme. The prescription dose was 40 Gy in 20 fractions (daily dose = 2 Gy). Plans were normalized to deliver prescription dose to 95% of the PTV. IMRT plan based on conventional dose parameters and P-IMRT plan based on lobe protection optimization method. Re-optimized plan and conventional IMRT plan have the same beam directions, target and other normal tissue dose constraints. The P-IMRT plan is obtained by replacing the original total-lung and R/L lung parameters with the dose parameters of lung lobes that will be retained after operation.

Dose Comparison
Dose distribution of planning target volume (PTV) and organs at risk (OARs) was assessed by dosevolume histogram (DVH). Target area evaluation parameters included: Conformity Index(CI) 7  Other OARs: Spinal Cord D max , Esophagus D max , Heart D mean , Heart V 30 and Body D mean .

Statistical analysis:
Summary statistics for variables were expressed as mean ± standard deviation (SD). Differences of two groups in means for continuous variables were compared using Student's t-test. Statistical signi cance was de ned as P < 0.05 using a 2-tailed test. All analyses were performed using SPSS Statistics v22.0 (IBM Corp, Armonk, NY, USA). Figure 2 shows the volume change of PTV in each case based on two scenarios. The average volume of PTV-OPT1 is 444.25 ± 152.44 cm 3 , while the average volume of PTV-OPT2 is only 252.56 ± 138.97 cm 3 .

Results
This huge difference in PTV volume will greatly affect the di culty of planning and the irradiated dose of OARs.
PTV Table 2 showed the difference of dose in four groups of PTV. All four groups of plans met their prescription dose requirements. For PTV-OPT1, CI and HI of conventional IMRT plan were 0.67 ± 0.08 and 0.15 ± 0.01, respectively. CI and HI of P-IMRT plan which controls preserved lung lobes dose were 0.58 ± 0.09 and 0.2 ± 0.03, respectively, with statistical difference (p < 0.001); for PTV-OPT2, CI and HI of conventional IMRT plan and P-IMRT plan were 0.63 ± 0.12 vs 0.59 ± 0.13 (p = 0.056) and 0.14 ± 0.03 vs 0.15 ± 0.04 (p = 0.15), there was no statistical difference. For larger, more irregular target areas with more complex relationship with surrounding normal tissues, strengthening the control of preserved lung lobes dose will have a certain impact on the conformity and homogeneity of target.   For PTV-OPT2, the situation is almost similar. Table 4 shows the difference of dose reduction between IMRT plan and P-IMRT plan in two scenarios, and compares the dose parameters of IMRT OPT1 and P-IMRT OPT1 with that of IMRT OPT2 , respectively. For Dose decreased by 159.86 cGy. Obviously, for larger and more complex target volume, it is more satisfactory to control the preserved lobes dose in order to reduce the lung dose.  .650 Comparing IMRT and P-IMRT plans of PTV-OPT1 with IMRT plans of PTV-OPT2, although the lung dose of IMRT OPT1 is much higher than that of IMRT OPT2 , the lung dose metrics except V 5 of P-IMRT OPT1 decreases by controlling the preserved lobes dose to comparable values in IMRT OPT2 , especially V 13 , V 15 and V 20 of Preserved Ipsilateral Lung, they are even lower than those in IMRT OPT2 . Difference between those groups is shown in Fig. 3.

Other OARs
For the two scenarios, dose to some OARs in IMRT and P-IMRT plans will increase. For PTV-OPT2, there is no statistical difference except Spinal Cord D max and Heart V 30 . However, Heart V 30 in P-IMRT OPT1 decreased slightly, possibly because these OARs did not exceed the dose limit of OAR constraints in conventional IMRT plans, and P-IMRT plans only changed the lung dose limitations and did not further control the dose of OARs. However, the results also showed that the dose of Body D mean in P-IMRT decreased for both PTV-OPT1 and PTV-OPT2. For PTV-OPT1, Body D mean decreased from 676.8 ± 124.0 cGy to 652.96 ± 122.85 cGy(p < 0.001), indicating that the overall dose of the patient outside the target volume was reduced.

Discussion
In this study, we evaluated the impact on target dose coverage and normal lung tissue by the application of lobe-preserving radiotherapy planning in two different target delineation schemes for neoadjuvant radiotherapy patients with IIIA-N2 NSCLS. The results show that this method can signi cantly reduce lung dose, especially the dose to preserved lung lobes after operation, without compromising target coverage.
The target volume of LA-NSCLC radiotherapy includes primary tumors and metastatic lymph nodes. When treating a large target volume, it can be challenging to deliver desired prescription dose to the target due to the need to spare organs at risk (OARs) 9 . In clinic, in order to improve loco-regional control and survival rates without increasing normal tissue toxicity, target delineation strategy was adjusted by optimizing the irradiation area of mediastinal lymph nodes and reducing the eld size 10 . Although previous studies 11,12 showed that limited target volume of IFI did not increase the risk of local recurrence, randomized controlled studies showed that the prescribed dose of IFI could be escalated for a small eld, with a 5-year LC rate (51%:36%, P = 0.032) and a 2-year OS rate (39.4%:25.6%, P = 0.048). The 5-year OS rate (25.1%:18.3%, P > 0.05) were better than ENI, but these studies were based on 3D-CRT technology.
With the development of technology, the performance of radiotherapy equipment has been improved continuously. The radiation delivery system has also progressed from conventional 2D-RT, 3D-CRT to IMRT. Image-guided IMRT can deliver radiation dose to target more accurately and protect normal tissue better. Dosimetric studies and clinical data of LA-NSCLC patients treated with IMRT radiotherapy have been reported [13][14][15] . An analysis of an RTOG 0617 study shows that IMRT was associated with lower rates of severe pneumonitis and cardiac doses, which proves the rationality of applying IMRT to LA-NSCLC, which improves target coverage while minimizing radiation to surrounding tissues 16 . However, at present, image-guided IMRT has become a conventional radiotherapy method, it remains to be seen whether highly conformal dose distribution and tightened margin may adversely affect the control of microscopic lesions of in-eld lymph node stations. In addition, IMRT technology may lead to an increase in the incidence of severe and fatal RP. It is necessary to increase the dose-volume limits of low-dose lung to reduce the incidence of fatal pneumonia 17 .
Our results show that, for any target delineation scheme, the doses of all lung lobes and total lung except resected lobe can be signi cantly reduced by changing the optimization parameters and ensuring that the target dose coverage meets the clinical requirements while keep the dose of Resected Lobe within acceptable range. However, for different target delineation schemes, the reduction range of the radiation dose of the lung lobes is different by controlling the radiation dose to the preserved lobes. For larger and more complex target volume, controlling the preserved lung lobes dose can signi cantly reduce the lung dose.
Moreover, by controlling the preserved lobes dose, the lung dose metrics in P-IMRT OPT1 was reduced very close to the value in IMRT OPT2 except V 5 , especially V 13 , V 15 and V 20 of Preserved Ipsilateral Lung were even lower than that in IMRT OPT2 , which made it possible for clinicians to select target delineation scheme for such bene ciary cases.
There are still some limitations in this study: the number of cases in this study is small, and no further indepth study has been carried out. There were few studies on LA-NSCLC radiotherapy for lung lobe protection before, and there was not much experience to refer to. We are still in the exploratory stage.
HL performed the manually contouring of lung lobes , and was a major contributor in writing the manuscript.
XF and ZX contributed to the conception of the study and helped perform the analysis with constructive discussions.