Use of AlignRT InBore in a Lung SABR Workflow on a Ring Gantry Linac

We’re going to be talking about how we started off our lung SABR workflow on our ring-based machines, or as you all know it, as SBRT. So we are based in the Royal Surrey County Hospital in Surrey. So we’ve got eight hospital sites. We’re based across all of South Surrey. We provide services to a population of about 330,000 patients. We’re a large teaching hospital. We have lots of students all over, and we treat most adult cancer patients. The population is about two million from the neighboring boroughs. Our cancer center is based over two sites, one in Redhill and the other one in Guildford, so they’re quite far apart. We offer surgery, chemotherapy, and radiotherapy across both sites. And we’re a seven linac department, so that’s three TrueBeams, four ring-based machines, which are two Halcyon and two Ethos, as well as brachytherapy and superficial units. And our ring-based machines have SGRT on them with AlignRT SGRT, and we’re aiming to continue our replacement program with more installations on our TrueBeams.

So when we introduced SGRT to our department, that was in 2024, we started off in our satellite center. We removed two TrueBeams and replaced them with the ring-based machines, a Halcyon and an Ethos. And I believe we were the first in the UK to be doing this, so we didn’t really have many other centers to speak to, so we did have to travel internationally to learn from other departments. Both machines were equipped with Hypersight and obviously AlignRT. Since then, we’ve been treating most cases: pelvis, chest, breasts, limbs, and our lung SABR cases. We are starting adaptive bladder and cervixes; I believe next week we’re going live. And so we’ll be getting more data for those. We are radiographer-led with our adaptive service, and hopefully in the near future, looking at tattoo-less radiotherapy and then introducing more of the head and neck treatments.

So when we introduced SABR into our satellite center, we were originally doing this at the satellite, but during the replacement program, we had to move all the patients to the main site. So obviously, these patients were traveling quite far, and it was really inconvenient for them. So to reintroduce this, we had to create an implementation working party, which is a multidisciplinary team of radiographers, dosimetrists, physicists, and our operations team to figure out this project. We created a SABR planning study, which looked at what the plans would be like because our planning team had not planned any of these patients before on ring-based machines. And we needed to look at the staff training. So the staff who were originally trained in SABR treatments had about 18 months to two years of a break in between starting these treatments again. So we needed to refresh their training, send them over to the main site for refreshers, and also do some training with VisionRT.

We had to look at our 4D CBCT workflow because the ring-based machines do not have 4D CBCT capabilities, and I’ll speak a little bit more about that shortly. And then we had to look at what our SGRT monitoring would look like for these patients. Did we have to explore different types of ROI? What sort of data were we wanting to get from this? And mainly because you can’t see your lasers when you’re inside of the bore, you’re not able to see if the patients are moving. So the SGRT was crucial for these patients. So we started off with an audit of 10 patients, and we’ll speak more about that as well.

So our planning study, we looked at three different types of plans. So those of you who don’t know the ring-based machines, they automate all of your fields, and it automatically decides which field it’s going to treat first in whatever order. You can’t change that on the system unless you’re changing it for every single patient. So we’ve allowed that for all patients, basically. So we had to look at if there would be any radiographer error with pausing the beams. Our workflow is take a CBCT, deliver the first arc, then take an interim CBCT, and then deliver the second arc, and so on. If you’re letting the machine automate your fields with the first plan, which was one isocenter and two arcs with the same gantry and collimator angles, only allowed a time lapse of two seconds for the radiographers to pause the beam, and same with plan two. We did one isocenter and two arcs with a slight collimator twist between the arcs, and that only allowed six seconds. That doesn’t account for if anyone walks into your treatment control room, if the phone rings. The slight distractions can cause the next arc to start treating before you’ve taken your interim one. So our planning team came up with plan three, which is two isocenters. They’ve got a 0.25 centimeter isocenter shift in between them, which actually stops the machine, does a hard pause, and it means that you don’t have that error.

So where the machine doesn’t have 4D capabilities, we actually send our patients to the main site for their 4D CBCT scan before their treatment. So we call it a day zero. This is because the UK SABR Consortium recommends 4D CBCT for all of the lung SABRs, and this is backed up by the ESTRO and ACCROP guidelines. So what we tend to do is have a patient have their CT scan at whichever site, then the planning team will plan them; whether they’re going to be having treatment at Guildford, it will be planned for a TrueBeam. If they’re going to have treatment at the satellite center, it would be planned for a ring gantry. And then they’ll have their day zero. So we make VAC bags for the patients. The VAC bag will be sent over to the main site where the TrueBeams are. They’ll have a day zero with their 4D scan, and if the tumor motion is stable, then they’ll continue having their treatment on the ring-based. But if there’s not, then they’ll be treated at Guildford on a TrueBeam. The challenges of this are obviously staffing. We have to make sure we’ve got the right number of staff at each site, and then also the patient-specific equipment. So we actually hire couriers to bring the VAC bags over to both sites. And it’s also increased travel for the patient. So obviously, we are an NHS institution. The patients, obviously, they don’t pay for their treatment, but it is an incurred cost of them traveling so far, and this does cause inconvenience for them. I’ll hand over to Gail.

Thank you. So as Hannah said, we had quite a lot of experience treating SBRT for our lung patients at our main site on TrueBeams, probably about over 10 years. But all our experience in terms of our confidence with immobilization, setup monitoring was on a TrueBeam. So when we moved over to treating at the satellite center with ring-based and AlignRT on board, we wanted to ensure we had the same confidence with that going forwards. And we’d already looked at the end-to-end testing with the out of board cameras versus in board during our commissioning, and we were confident that they really aligned. So we were happy with that part. We just wanted to see how it would work on a SABR sort of, or SBRT workflow.

So we thought we audited our first 10 patients to do this. And again, it was to ensure our safe delivery for SABR on ring-based with AlignRT. We evaluated our in-board SGRT with our ancillary equipment, so abdominal belt compression, for example. We wanted to compare the use of in-board SGRT monitoring data to the cone beam CT data, our soft tissue match prior to our arc deliveries. And we wanted to evaluate the reliability of in-board SGRT for continuous motion monitoring, because as Hannah said, when they’re in-board, we can’t see the lasers, so we can’t be monitoring using the lasers.

So our method was, as Hannah explained, we planned our patients with these two isocenters with a 2.5-millimeter shift, and there were no issues with the planning. It’s a really small shift. It just allows a hard pause, so it stops the ring-based linear accelerator from automatically loading the next beam. In terms of immobilization, our standard immobilization for these patients are arms up on a wing board with a vacuum bag for arm support. And patients with middle to lower lobe tumors are typically treated with an abdominal compression belt. And as was discussed this morning, it’s really important to select an appropriate region of interest.

So for each patient, real-time delta shifts in six degrees of freedom—vert, long, lat, yaw, roll, and pitch—from the in-board SGRT system were noted at specific time points. We took them at just before our first cone beam CT acquisition. We then noted them after applying our first cone beam CT moves, after our first arc delivery, before applying our interim cone beam CT moves, and after our second arc delivery. And for consistency, we noted these deltas always at exhale, so you can see the trace. So we noted them at exhale.

So the different time points were used as follows. The change whilst acquiring our cone beam CTs and matching was deemed to be the change between time points one and two, and three and four. And any change during arc delivery was taken as the deltas at time point three and time point five, and this is because immediately after doing our match, we then acquire a new sort of surface and in effect, reset our deltas to zero. And then finally, the deltas from our online cone beam CT soft tissue match were compared to time point one.

So our results. So 10 patients were audited, including eight upper lobes and two middle lobes. The ROI always excluded the abdominal belt. And a standard thorax ROI was not always appropriate for all our SABR patients. So in the images you can see there, the one on your far left is what we might use as a standard thorax ROI, and then in the middle, if it was an upper lobe, and on the far right, if it was perhaps a lower lobe, or perhaps when we have several different lesions which we’re treating. And our tolerances are generally two millimeters for vert, long, and lat, and three degrees for your roll and pitch.

So looking at this, the results in a graph, across all our 10 patients, the mean change in deltas during the time the cone beam CTs were acquired and matched and during arc delivery was less than one millimeter for vert, long, and lat, and less than 0.3 degrees for yaw, roll, and pitch. So on that graph, you’re seeing on the Y-axis, it’s the mean delta change in centimeters or degrees, and on the X-axis, they’re sort of paired. So we have the change in vert during the cone beam CT across all 10 patients, and the change in vert during the arc delivery across all 10 patients, and that sort of goes along there for long, lat, yaw, roll, and pitch.

When we compared the AlignRT deltas to our soft tissue match, again, there was no significant difference really to note, and that was really reassuring for us. When AlignRT flagged a move, we found that it justified an additional image because a soft tissue positional change was confirmed. And overall, we found that patients experienced shorter treatment appointment times by an average of nine minutes. So I have to qualify this. We were comparing to our patients being treated on TrueBeam without AlignRT to patients on a ring-based with AlignRT. So the shorter treatment time is attributed to both being on a ring-based in combination with AlignRT.

So in conclusion, we feel that onboard SGRT for lung SABR or lung SBRT was adopted easily with decreased time on the treatment couch, and to date, we have treated more than 50 patients. The sub-millimeter change in deltas during arc delivery makes us confident that we have a safe treatment approach. And as I mentioned, we saw no significant difference between AlignRT deltas immediately prior to matching and our agreed online soft tissue match deltas. And this means that moving forwards, we would consider omitting our interim cone beam CT and only acquiring if the AlignRT deltas are out of tolerance. And I think this perhaps is in line with one of the talks earlier this morning, the SRS talk, where they were talking of perhaps dropping their mid or interim imaging.

So we’d like to acknowledge all the staff at our center for enabling this working party to take this project forward and establishing it to a standard of practice, which is really important for our patients at our satellite center because it means that they don’t have to travel. And also, the decreased time means that we can have shorter appointment times and have more throughput, which is a really big thing for us in the NHS. And yeah, thanks for listening, and we’re happy to take any questions. Thank you.

*This transcript has been AI-generated. Contact us at secretary@sgrt.org if there are any issues.