Significance of motion control
Respiratory movement leads to artifacts in the process of chest and upper abdomen imaging and limits the accuracy of magnetic resonance imaging, it also reduces the accuracy of positron emission tomography-CT images. With regard to radiotherapy, it increases radiation dose to the normal tissue adjacent to the irradiation volume. Consequently, the inhibition of respiratory movement during imaging capture and radiation therapy has been widely investigated to improve image quality and reduce exposure to the normal tissues, while maximizing radiation dose to the tumors  .
The largest organ displacement caused by respiratory movement occurs in the cephalic-caudal direction, and is smaller in other directions. Table 3  shows the mean, standard deviation and maximum displacement of some organs in these directions. According to the Report 76 of the American Association of Physicists in Medicine, the methods to reduce the influence of respiratory movement in radiotherapy can be divided into five categories: motion-encompassing, breath-holding, forced shallow breathing with abdominal compression, respiratory gating and real-time motion tracking. These techniques have been widely used in clinical practice, and each method has its own advantages and disadvantages. Because of the growing popularity of stereotactic ablative radiotherapy (SABR)/SBRT and particle radiotherapy which have higher requirement for precision, the search for a feasible and affordable method of respiratory motion control has clinical significance.
Application of invasive and non-invasive ventilator-assisted breathing management
Ventilator-assisted respiratory movement control has been applied in radiotherapy worldwide. High-frequency ventilation (HFV) is one of this type of technique and has been reported to suppress thoracic movement under general anesthesia in patients with invasive ventilation in SABR and percutaneous radiofrequency ablation of tumors [9,10]. Most reported studies of ventilator-assisted radiotherapy used high-frequency jet ventilation (HFJV) and high-frequency percussion ventilation (HFPV) where general anesthesia was needed which made the whole procedure more complex with accompanying anesthesia-related risk. There are no reports of HFOV-assisted radiotherapy in the literature.
Functional apnea was applied for proton therapy in the Rinecker Proton Therapy Center (RPTC), Munich, Germany . They reported that from September 2009 to November 2013, 61 patients received a total of 673 fractions of proton treatment. A total of 3025 apnea were performed: the anesthesia duration was 28-133 min (mean 57 mim), each apnea time was 2.3 min (2-9 min), the proton therapy time was 30 min on average, and no adverse events occurred. Anesthesia recovery time ranged from 45 minutes to 3 hours, with an average of 90 minutes. There were mild side effects such as throat irritation (1%), dysphagia (1%), hypotension (1%), nausea (1%) and drowsiness (3%). The mean target movement during apnea was 2 mm (range 0-4 mm). The technique uses endotracheal intubation under intravenous general anesthesia and mechanical ventilation. Mechanical ventilation was temporarily stopped during the proton beam-on period, while oxygen supply through the endotracheal intubation continued. During gantry rotation, the beam was off while mechanical ventilation was resumed. During the whole treatment, the patient's pulse oxygen saturation was kept normal and the level of partial pressure of carbon dioxide in endexpiratory gas (PETCO2 ) increased by 2-4 mmHg/min. Our center also applied functional apnea technique for radiotherapy in 2016. Although it was effective on the management of respiratory motion, there were shortcomes of safety and logistic consideration .
Goldstein JD et al reported their experience of movement management using continuous positive airway pressure (CPAP) ventilation in stereotactic lung radiotherapy. Under CPAP, the tumor motion in the superior-inferior, right-left and anterior-posterior plane decreases by 0.5~0.8 cm, 0.4~0.7 cm and 0.6~0.8 cm, respectively，and the lung capacity was significantly increased . Peguret N et al.  reported in 2015 percussion-assisted radiation therapy (PART) for managing respiratory movement. PART was initially tested in 10 volunteers and found to be well tolerated, allowing a median breath-hold time of 11.6 min (range 3.9 to 16.5 min). The technique was subsequently used in 3D conformal radiotherapy for breast cancer, SBRT for lung cancer and volumetric-modulated arc therapy (VMAT) for palliative pleural metastasis. The median breath-hold time was 11.6 min (3.9,16.5) and the mean apnea-like breath-hold time was 7.61 min (SD=2.3) without radiation interruption. Durham, A. D. reported the PART used in Hodgkin lymphoma allows a marked reduction in heart dose.
Biro P et al reported in 2009 a study of liver movement inhibition in dogs under total intravenous anesthesia (TIVA) using high frequency jet ventilation (HFJV) with or without muscle relaxants. The results showed that the liver movement of anesthetized dogs could be controlled within 3.0 mm after the introduction of HFJV, and the injection of muscle relaxants did not further reduce the liver movement. Animal studies have shown that the use of HFJV can limit liver movement to a certain extent, and it can be used in human to reduce the amplitude of liver movement under TIVA to perform radiofrequency ablation of liver tumors .
Non-invasive HFJV has also been reported. Ogna A et al carried out trials of respiratory inhibition under high frequency ventilation on animals and human volunteers. Prolonged apnea-like status could be achieved under NIHFJV with the respiratory frequency set at 250/min, resulting in a median duration of apnea of 20 min. And when the respiratory frequency was set at 500/min, the median duration of apnea was 5:16 min (3:57-6:48 min). The tidal volume of HFJV was dependent on respiratory rate and was 54 ml with a respiratory rate of 250/min and 26 ml with a rate of 500/min. PtcO2 was greater than or equal to 97%, and an increase of 6.2 mmHg of PtcCO2 was noted. Very small tidal volume results in a significant reduction in chest and abdominal movement compared to spontaneous breathing.
Management of respiratory movement with normal-frequency or high-frequency ventilator with or without sedation have been reported, but the procedures is were relatively complex and had a high risk of hypercapnia, and the ventilator-on duration is not long enough for most precise radiotherapy.
HFOV delivers small tidal volumes below the dead space volume with a high frequency of 300-900 breaths per minute, it can prevent ventilator-induced lung injury by decreasing the pressure of the airway, while providing adequate ventilation and oxygenation [17,18,19,20]. HFOV maintains alveolar inflation at a constant, less variable airway pressure with a sinusoidal air-flow oscillation to prevent the lung from the "inflate-deflate" cycle and provides improved oxygenation. Because the lung is kept in a constant inflated and well-oxygenated status, the patient can only breathe very shallowly while maintaining a normal partial O2 level, thus creating the possibility of a minimum breathing amplitude with subsequent very small respiratory motion of the lung. HFOV can be supply by Care Fusion Oscillator 3100A and 3100B ventilatory and the Drager Babylog VN500 ventilator and . In the clinical, HFOV is a ventilation mode that can be used to protect the lung in a full spectrum of patients with acute lung injury, from neonates to adults. HFOV is often used as a rescue strategy when conventional mechanical ventilation (CV) fails.
To our best knowledge, we are the first to report the application of NIHFOV for prolonged respiratory motion depression. The procedure, as tested by 23 healthy non-sedated adults, was smooth and feasible, and all of them had abundant blood oxygen without retention of carbon dioxide. NIHFOV may overcome some of the drawbacks of CPAP, PARP, and HFJV etc.
When we started the NIHFOV training, we chose the frequency of 250/min and the airway pressure of 15 cmH2O based on a literature review. Although this setting could be tolerated，once the subject swallowed，the procedure would be interrupted due to high airway pressure. Thus we titrate and optimize them. At the end of seven training sessions, all the subjects could easily underwent more than 30 minutes without discomfort.
In this study, we confirm the hypothesis that NIHFOV helps maintain normal blood oxygen and carbon dioxide levels, and allows a non-invasive and non-sedative approach of respiratory motion control. Compared with other means of methods, NIHFOV significantly prolonged apnea duration.
At the time of this manuscript preparation, our team has successfully titrated the oxygen concentration down to 50%, and the longest HFOV training time was over 60 minutes; the PtcCO2 was still maintained at around 36 mmHg and the PtcO2 at 120mmHg.
There are several limitations of this study. Pure oxygen inhalation was used in this experiment, and PtcO2 of the subjects reached more than 300 mmHg. This is equivalent to radiotherapy under hyperbaric oxygen which may affect the outcome of radiotherapy, and needs to be further studied. We chose pure oxygen based on the data from high-frequency ventilation (HFV) in the literature, which using a noninvasive interface (HF-NIV), the HF-NIV was performed using a Monsoon III ventilator. But from our experience, pure oxygen may not be necessary for NIHFOV. Interestingly, the first author forgot to connect oxygen to the ventilator during one of the training. Although only the room air was pumped, blood oxygen level was still within normal limit, and the training was also progressed normally, hinting that a reduced oxygen concentrations may still be workable with NIHFOV. We have started to examine the NIHFOV procedure with a lower oxygen concentration. We found when the subject was training with sitting position, saliva often appeared, some subjects experienced saliva accumulation after about 10 minutes in a sitting positive, which requiring inducing swallowing and causing uncomfortable and coughing feeling. This scenario did not occur with supine position. So the authors suggest volunteers to use supine position. There have been previous reports of complications such as dryness of upper respiratory tract and sore throat with using a noninvasive interface (HF-NIV). In this study, we kept the humidifier open and humidified the inhaled gas, avoided severe dryness and pain of upper respiratory tract.
Because 4-DCT has a 6-time higher exposure dose than that of conventional CT scanning [24,25], we limited the application of 4DCT to only eight subjects. Because there were no tumor or metal markers,nor iodine oil as reference in the scanning areas, we used the diaphragm mobility in the cephalic-caudal direction as the endpoint of organ motion. However, it cannot measure the mobility of the anterior-posterior and left-to-right direction. As this technique matures and is practiced in clinical, further study needs to be carried out in this respect.
This study systematically evaluated the effectiveness and safety of this technology, which enriched our team's experience. We confirm that NIHFOV has good gas exchange function and low airway pressure with small tidal volume, no carbon dioxide retention occurs. At first, we also feel NIHFOV use is time-consuming, need a experienced technical team, but with most of our team members accepted the training as a healthy volunteers, we made a training video, summarize and exchange experiences, quickly make NIHFOV easy and acceptable. The subjective feelings of 23 volunteers were good, among whom the oldest volunteer was 69 years old, and all volunteers had fulfilled all training with good tolerance. The NIHFOV duration could easily exceed 30 min, and no CO2 retention ever occurred. In addition, in the implementation of this technique, as HPJV and CPVP technique, the lung volume increases. In this study, it was found that the variation coefficient of the left lung was larger than that of the right lung. After analysis, it was found that volume of air content of the gastric bubble was the reason why the variation of the left lung volume is obvious. It will affect the mobility and reproducibility of the left lung. This should be emphasized and avoided in practice.
This novel approach to managing respiratory movement using NIHFOV has never been previously reported. Our system is not a commercial system, and many parts are put together by our team. It took half a year for our team to learn how to use it harmoniously, because its efficacy is clear and its apply in patients is easily recognized and quickly mastered. In the initial stage of applying HFOV technology, we closely monitor all vital signs and various parameters, then we found that the technology is easy to operate, only a slight increase PtcCO2, there is no need to initial hyperventilation, for heart rate and blood pressure, there is no obvious fluctuations, the entire team mastered the technique, we simplify the process. All indexes and parameters of the subjects were monitored only during the training initial phase. Now we don't think it's necessary to measure blood pressure, or assess arrhythmias. Because it is an open system, the patient is not sedated, is sufficient oxygenated, and the training can be discontinued at any time during treatment. No side effects were observed. In the future, it will be applied in carbon ion therapy. At finished carbon ion treatment and waiting for radiation attenuation session, patients can be informed through the microphone, and the connection can be unplugged and spontaneous ventilation can be resumed.
In conclusion, we demonstrate the safety, feasibility, and good tolerability of NIHFOV in suppressing respiratory movement in healthy volunteers. This study lays the foundation for further research on the potential benefits of this technique in improving the accuracy of carbon ion radiotherapy as well as photon radiotherapy, ensuring the accuracy of radiotherapy delivery, and further reducing the dose of normal tissues.