The feasibility of maskless radiation therapy for head and neck cancer by using surface guided radiation therapy (SGRT)

Abbey Adams (00:03):

So good afternoon. My name is Abby and I’m a treatment unit leader or treatment senior depending on where you are at Genesis Care North Shore and Marta in Sydney, Australia. And I’ll be discussing a case study about a head and neck patient who was treated fully massless using AlignRT. So to give a bit of background about our sites, we operate between two sites across North Sydney in Australia, which is Marter and North Shore who share a staffing network. We have Varian TrueBeam at our Marter site and a Varian Edge machine at our North Shore site. It is important to note that both of our sites have AlignRT with the postural alignment system that we use to treat all patients with. So historically patients suffering from severe claustrophobia have presented as challenging cases when treating in the head and neck region due to being confined in a thermoplastic head and neck mask without a desirable treatment alternative. So considering this, the aim of this case study was to assess the feasibility of massless radiation therapy to treat head and neck cancer using surface guarded radiation therapy. This case study focuses on one case in particular and discusses the simulation planning and treatment workflow, the use of ROIs imaging challenges and limitations as well.

Abbey Adams (01:15):

So to give background on this case the patient is a 50-year-old male diagnosed with T2N1MX SCC of the left tonsil standardly. Additional scans are required to confirm diagnosis but this patient actually has a history of incomp completing scans due to his severe claustrophobia and anxiety. The patient was actually booked for a simulation appointment at another radiation therapy department in the area after a two hour long appointment and numerous attempts at making varying forms of head and neck mask. The patient actually refused simulation and could not receive treatment at this center. So this patient was also discussed extensively at numerous MDT meetings between radiation oncologists in the area and was then referred to Genesis Care North Shore to explore SGRT pathways.

Abbey Adams (02:02):

So once this patient was referred to the team, the case was discussed at length between the rts, ROS and physicists to ensure a smooth workflow. The patient was positioned standardly on an S-frame with a headrest A to allow for reproducible neck flexion. A custom mold care cushion was indexed using the headrest, ensuring good contact with the shoulders and neck. A knee fix was also indexed. A large non-slip mat was placed under the mold care cushion that extended to the knee fix to limit movement further as well. Along with ITN also being marked on the mold care. This man was clearly also extremely hairy and our past experiences with AlignRT has shown that large amounts of hair can affect the surface the camera actually picks up on. So this man’s facial hair was removed prior to his simulation was kept at the same length throughout treatment as well.

Abbey Adams (02:51):

So due to this being a new technique for the department, a day zero appointment was booked for this patient. This consisted of increased coaching for the patient to determine if the patient could actually withstand the treatment through a trial run. The patient was positioned on the treatment couch in their recorded simulation position while the AlignRT system was turned on. The patient was aligned and instructed to move their head in all directions to determine the sensitivity of the AlignRT cameras. The cameras were found to detect movement immediately and tight tolerances were set to ensure accuracy within a 0.5 PTV margin. It was actually decided to use two ROIs for setup. We found that by only using a singular ROI, including all relevant anatomy that was needed, produced two big of a surface for the system to work with and was less sensitive to small shifts of the skull. So a larger ROI was used first, which includes larger areas of the face and shoulders to ensure the patient’s body was in the correct position. This was called the setup ROI and will be used to treat during the first setup only. The second ROI used was a smaller region that focused on fine facial features close to the PTV and was used to manually adjust the face very slightly. This was known as the treatment ROI and was set to treat with beam control. We then rotated the gantry from 1 8 1 to 1 7 9 degrees with the imaging panels extended out while closely monitoring the AlignRT system to ensure the LINAC did not block the cameras at any point and no movement was actually detected from this session.

Abbey Adams (04:22):

So this is an image of the setup ROI that was used and includes all areas of the chest, shoulders and large areas of the phase to determine the body and shoulder position. The patient was instructed to move up or down the bed based on this and we could also adjust shoulder position with this ROI as well. So this image is the treatment ROI, which includes the forehead, nasal region, chin and ears. It was decided to exclude areas such as the eyes due to blinking and twitching not wanting to be a factor. So from the day zero appointment as well it was determined that the smaller area, the more sensitive to movement the AlignRT cameras were. This case was discussed extensively at peer review between the radiation oncologist and it was decided that no additional measures were necessary for this patient compared to a patient being treated with a mask. A larger PTV margin was discussed and considered to account for any error that may occur with this new technique. However, it was decided that a 0.5 standard expansion was used due to the tight tolerances set at the day zero appointment and this patient was planned to a radical dose of 70 Gy and 35 fractions to the oropharynx and lymph nodes.

Abbey Adams (05:32):

This patient was also planned using a vmat technique with three full arcs due to the volume being relatively large and treating bilaterally as well. After discussions with physics and the radiation oncologist, it was actually decided to use six FF energy for faster treatment delivery as we wanted to reduce the amount of time this patient was on the bed for due to his claustrophobia and anxiety as well. So for the patient’s treatment, they’re aligned using the setup ROI by manually shifting the bed and the patient. And then for final adjustments were made using the treatment ROI. The video function or postural alignment was also switched on to check shoulder, chin and nose position to see if they’re aligned with the projected SGRT line after day one. These couch values were required to allow auto shift couch couch function to allow quicker setup for future fractions as well. A pre-one beam was taken daily where the patient was closely monitored throughout the scan to interrupt the beam. If AlignRT noted any significant movements, the image match consisted of a bones match priority while also ensuring the soft tissue were within high and low dose PTV margins, chin and shoulder position as well as contour change were also considered daily.

Abbey Adams (06:45):

All shifts were applied and recorded. It was decided to use 4DOF only due to the increased risk of the patient movement. When shifting the bed using 6DoF a reference capture was taken for AlignRT that would be used as a reference to detect any movement extending beyond 1.5 millimeters and 1.5 degrees to which this beam was set to beam hold. If this were to occur, the process was to re-image and start the workflow again. So once the treatment fields were delivered, a post KV MV pair was delivered to ensure the AlignRT system was accurate and there was no interfraction motion that the AlignRT cameras did not detect. The KV MV pair was matched to bones and all shifts were recorded daily.

Abbey Adams (07:28):

So there were actually a few complications noted from the duration of this patient’s treatment. However, one noted in particular was a jaw positioning error on fractions. Three. The patient’s jaw opened very slightly throughout the treatment and was picked up on the post MVKV image. This was put down to the patient being more relaxed compared to their CT scan and the patient got used to the procedure as well. This was a concern however as AlignRT didn’t actually detect this change but the movement was in 0.5 centimeters and within the PTV margin to combat this in future fractions. More ROI region was drawn around the chin and jaw as well as reminders over the microphone for the patient to keep their jaw closed, which was successful in treating this patient in future fractions.

Abbey Adams (08:12):

Another issue encountered due to the patient being more relaxed was a neck flexion issue At the patient’s CT scan he was extremely tense due to being stressed and anxious and had quite an extended neck. A trend noted from fraction eight to fraction 12, the RT noticed that the patient’s neck was more flexed in the purple due to the patient being more relaxed on the bed. This made positioning day to day very difficult to reproduce and caused image match issues due to the bones and soft tissue no longer lining. A rescanned occurred after fraction 12 with a patient being more relaxed, which was successful in correcting these issues for future fractions.

Abbey Adams (08:49):

Another anticipated issue for head and neck patients is weight loss. So from fraction 15 to fraction 17 the treatment team noticed an approximate one centimeter contour decrease. This was sent back to planning to do a dose calculation where the plan was found to be acceptable but calculated to be 1.8% hotter in the PTV area with OAR still being with intolerance. This however caused issues with setup daily as AlignRT was having difficulties comparing the patient setup with simulation setup due to being one centimeter different. The treatment staff had to take an SGRT reference capture to be used daily for future setups where AlignRT takes a new capture from a setup and compares to future setups to which was successful in resolving setup issues for the future fractions.

Abbey Adams (09:36):

So the above graph compares the results of the treatment time for the massless head and neck patients to another patient who has completed treatment at the same time, was actually referred by the same oncologist, had the same diagnosis and was also treated with three arcs but was treated with a mask instead. So while understanding this is a limited cohort due to being a relatively new technique at our center, it gives representation of comparative results as variables were kept the same such as machine staffing and this treatment center. It was found that the average treatment time for the masks was patient was approximately 17 minutes, which took about four minutes longer than the head and neck patient that was treated with a mask. The additional time can be put down to additional in-room setup with the ROIs in comparison to moving directly to Marks on cast. It being a new technique at the center and staff training as well as more patients come through it is time is likely to decrease. It is also important to note that the two spikes in treatment time for fraction 25 and 29 were due to a reset up of the patient which was the only times this occurred and AlignRT effectively detected the movement both times.

Abbey Adams (10:42):

So the above table compares the results for the shift supplied off the cone beam for the same two patients. They both actually produce similar results with them both being very minimal. So by comparing these results it shows that our AlignRT produces an accurate in-room setup and all the ROIs determined gave a good result as it resulted in small applied shifts.

Abbey Adams (11:03):

So the above graph shows results from the matched post MVKV image that were matched independently by the same staff member to reduce bias. The red line depicts a 0.5 shift which would indicate the patient moved beyond the PTV margin throughout treatment and was not detected by AlignRT. But no post shifts ever recorded were above the 0.5 centimeter line and the spike on fraction three was due to the chin drop. As discussed earlier, the average of all shifts in all directions equated to less than one millimeter proving that all AlignRT was very accurate in monitoring this patient. Throughout treatment,

Abbey Adams (11:38):

This patient was also surveyed post-radiation therapy at a follow-up nursing consult and the positive comments were overwhelming. The main takeaways were his appreciation to be given the opportunity to participate in this new technique as if feared he had very poor outcome without having AlignRT. He liked the extra measures the staff also went to to make him feel more relaxed such as a custom Spotify playlist and consistent staffing from simulation to treatment. He also experienced no abnormal side effects which would be expected. These results were the most vital for us as our main aim for this technique was to provide more support for these patients.

Abbey Adams (12:15):

So in future the main goal is to collaborate with more oncologists to trial this technique further and gather data from a larger patient cohort. Future applications to be considered would include the use of Sydney imaging, which can take MV images throughout treatment with no additional dose to the patient. The images could also be assessed throughout this treatment to give further confidence with this technique. This would require additional testing though as the panels would actually be required to be extended throughout the treatment, which could potentially affect monitoring, but is an application we’re trialing currently. We also currently have a very well developed SRS workflow, which uses an open face mask. Currently we are also looking into doing a hybrid between these two techniques for standard head and neck patients as well as another alternative for these patients as well. And with a larger patient cohort, we would also be working towards removing the post KV MV image, which would reduce treatment time further as well. But that’s all from me. Thank you very much for your time.