Page 108 - Read Online

Thonglert et al. Hepatoma Res 2023;9:40 https://dx.doi.org/10.20517/2394-5079.2023.47 Page 11 of 23

maintaining the dose to the OARs below the constraint can be difficult to achieve.

The rationale for MRgRT
MRgRT is one of the most promising innovations in RT. These systems have an MRI unit integrated into a
linear accelerator (Linac). MRgRT possesses three main advantages: improved soft-tissue visualization,
online adaptive radiotherapy (ART), and real-time synchronous tracking and respiratory gating [57-59] . These
characteristics offer more precise radiation delivery, which may translate into better clinical outcomes with
regard to improved LC and reduced toxicity.

MRgRT enables the implementation of more hypofractionated dose escalation, while minimizing
intentional underdosing to the tumor near the OAR by providing precise radiation with a smaller
[63]
uncertainty margin [60-62] . MRgRT provides better visualization of tumors and OARs than CBCT imaging .
Real-time MRI scans can capture intrafractional motion and facilitate tumor tracking. Consequently, an
invasive procedure such as fiducial implantation is not required [61,62,64] . Certain MRgRT devices have the
[65]
capacity to automatically gate delivery, thereby limiting the margin of ITV . As surrounding GI structures
have interfraction variation, high-quality images and online ART systems allow the ability to recontour, re-
[66]
optimize, and adapt treatment to adjust for anatomical changes . Furthermore, MRgRT may also allow for
smaller treatment margins to be applied when accounting for motion and setup uncertainties. Together,
these features may lead to the potential for improved tumor coverage and reduced dose to normal organs,
which may translate to improved tumor local control and decreased normal tissue toxicities.

Clinical outcomes and applications of MRgRT for iCCA
The feasibility and effectiveness of MRgRT in treating iCCA are based on retrospective studies as
summarized in Table 2 [60,67-69] . Nearly all studies included patients with primary and metastatic liver cancer,
except one study that focused only on patients with CCA. Luterstein et al. conducted a retrospective review
of 17 patients with CCA (5 with iCCA) treated with MR-guided liver SBRT . The margin of the planning
[68]
target volume (PTV), which accounts for setup uncertainties, was only 3 mm from the gross tumor volume
(GTV), smaller than historically used for treatment planning. The patients were treated with breath-hold
respiratory gating and real-time tumor tracking, and adaptive planning was routinely performed. The
median RT dose was 40 Gy, delivered in 5 fractions, to tumors with a median tumor size of 25.5 mm (range
14-37 mm). At a median follow-up time of 15.8 months (range 4.5-29.9 months), the LC rates for 1 year and
2 years were 85.6% and 73.3%, respectively. The median OS was 18.5 months, with 1-year and 2-year OS
rates of 76% and 46.1%, respectively. Only one patient (6%) who was treated before using adaptive planning
experienced grade 3 acute toxicity of duodenal ulcer and perforation. No grade 3 late toxicities were
observed. A multi-institutional retrospective study by Rosenberg et al. included 26 patients with primary
liver cancer (2 with iCCA) or liver metastasis treated who underwent MR-guided liver SBRT . The median
[69]
dose delivery was 50 Gy (range 30-60 Gy) in 5 fractions. A 2- to 5-mm expansion was performed to create
the PTV. Inspiratory breath hold was used in 16 patients, while 10 patients used a modified shallow internal
target volume or exhale-based setup. All of them were monitored with real-time tumor tracking. Daily
adaptive treatment was not administered. The median PTV was 98.3 cm (range 13-2,034 cm ). At a median
3
3
follow-up of 21.2 months, the rate of FFLP was 80.4%. Two patients (7.9%) experienced Grade 3 GI toxicity:
1 with hilar stricture and 1 with portal hypertension. There were no reported GI toxicities of grade 4 or
higher. Two patients had a decrease in Child-Pugh class. The authors mentioned that both patients had
large-volume tumors that led to high mean liver doses.

Rogowski et al. reported 11 patients with primary and secondary liver tumors, 2 of whom had iCCA, treated
[60]
with online adaptive MRgRT . The PTV expansion was 3-5 mm from the GTV. All patients were treated