At present, 3DCBCT technology is still widely used for location verification in clinical practice. However, a complete 3DCBCT scan include multiple respiratory phases, resulting in motion artifacts (11). However, 4DCBCT takes into account the time weighted average factor on the basis of 3DCBCT, which can display the position of tumor and surrounding normal organs throughout the respiratory cycle, track the longest dwell stage of tumor in the complete respiratory phase, and more accurately guide the setup error correction (12). 4DCBCT can dynamically observe 10 groups of setup errors under 10 respiratory phases in the complete respiratory cycle, and obtain the final setup errors through time weighted average, so as to improve the accuracy and ensure the efficacy of radiotherapy.
Many studies have shown that 4DCBCT has potential advantages in stereotactic radiotherapy of lung cancer. It can monitor the spatial position information and motion range of tumors, guide setup error correction, and support smaller CTV-PTV margin (13). Previous studies have reported that the setup errors measured based on 4DCBCT scan were smaller than that on 3DCBCT scan during lung cancer irradiation. Similarly, the location of liver tumors is easily affected by diaphragm movement, respiratory movement, gastrointestinal peristalsis, etc., and 4DCBCT also has potential advantages in monitoring the target motion in the radiotherapy of liver cancer. However, few studies have reported the application of 4DCBCT in the radiotherapy of liver cancer. Vergalasova et al. found that even with implanted markers, 3DCBCT scanning under free breathing state might not be accurate enough in guiding liver stereotactic radiotherapy (14). Our study first explored the feasibility and potential advantages of 4DCBCT in IGRT for liver cancer. This study found that the overall setup errors based on 4DCBCT measurement were smaller than 3DCBCT, especially in S-I direction, A-P direction, transverse plane, and coronal plane. In addition, the setup errors in S-I direction were more diverse in 4DCBCT group and 3DCBCT group, with 3.1% and 8.4% of patients with setup errors more than 10mm, respectively. Thus, 4DCBCT is essential for monitoring the setup errors in S-I direction. And we believe that 4DCBCT is superior to 3DCBCT in monitoring the setup errors during liver cancer irradiation.
We also found that the CTV-PTV margin guided by 4DCBCT could be smaller than 3DCBCT, which should be beneficial to normal liver tissue sparing while ensured the target dose coverage, and thus better ensure the safety and efficacy of radiotherapy for patients with liver cancer. The error in liver cancer radiotherapy has multiple sources, such as irregular respiratory motion, poor posture repetition, and other motion during treatment. These variances may affect the dose distribution in the target area and endanger normal tissues. Most of the selected patients in this study were patients with stage III liver cancer. The median size of the tumor was 7.6 cm. 87.3% of the patients were infected with chronic hepatitis B virus. Therefore, it was more necessary to reduce the radiation dose exposing on the normal liver tissue to avoid the occurrence of radiation liver disease. Although the CTV-PTV margins may be insignificant comparing to “supermassive” liver tumors, it may play an important role in SBRT for liver cancer under free breathing and its value should be further explored.
In terms of time cost, it took 4 minutes for acquisition and high-resolution reconstruction of 4DCBCT images, and 1 min for image registration. Although the scanning time of 4DCBCT was averagely 62s longer than that of 3DCBCT, it could online evaluate the range of tumor motion and OARs to avoid the occurrence of target missing and unexpected OARs exposure. Compared with the overall treatment time, the increased position verification time is acceptable. Currently, several studies have focused on reducing 4DCBCT scanning time(15).
Our study has two advantages in design. First, the patients were scanned consecutively with 3DCBCT and 4DCBCT after setup, which minimized the interference of other unrelated factors, and could more accurately evaluate the difference between the setup errors measured by the two image guidance methods. Second, the 3DCBCT and 4DCBCT used in this study are widely used image guidance systems for Elekta linear accelerator, making our findings valuable in guiding the clinical practice of liver cancer radiotherapy.
This study also has limitations. First, this study is a single center study, and the accuracy of treatment setup depends on the skill and experience of the stuff. Therefore, the results of this study (such as setup errors and CTV-PTV margins) might not be well generalized to other centers without an extension multi-center study. Second, the number of enrolled cases was small, and no hierarchical analysis was conducted on tumor size and location, etc. Third, this study did not compare the differences in dosimetry distribution, local control and adverse events between the two groups. Fourth, the motion track of the lesions and surrounding OARs in six dimensional directions should be further explored.