Table 1.
Overview of challenges for the implementation of MR-guided radiotherapy on a hybrid machine for breast cancer patients.
| Challenge | Effect | Potential solution |
|---|---|---|
| SIMULATION | ||
| Patient positioning inside the MR bore | Prone: breast deformation on tableramya and fitting of receiver coil (Figure 2) | Development of a thinner coil or a dedicated MR-linac breast coil |
| Supine: difficulties fitting arms inside bore in standard RT position | Use a minimal or no inclined wedge support, move arms closer together above the head | |
| Deformation of body contour by receiver coil | Disturbed body contour | Use coil bridges to support the coil (Figure 1) |
| Body contour visibility in prone position | With dedicated prone breast coil, body contour and OARs not visible further away from coil | Use an additional coil placed on top of the patient |
| Electron stream effect | Irradiation dose outside the treatment field in an inferior-to-superior direction (Figure 4) | Include chin, arm, and abdominal region in the simulation plan |
| Breathing and cardiac motion during scanning | Motion artefacts | Use a 3D sequence, signal averaging, and left–right phase encoding in protocol design, or use triggering or breath-hold for acquisition |
| CONTOURING | ||
| Surgical clip and/or marker visualization on MRI | Magnetic field distortion and artefacts impeding contouring of target volume (Figure 3) | 1. Use or develop markers or clips with smaller artefacts 2. No marker insertion (only possible in the neoadjuvant setting if no further surgery is required) |
| SIMULATION AND PLANNING | ||
| Geometric accuracy (gradient nonlinearities) in combination with lateral target volumes | Reduced geometric accuracy, increasing with distance from isocenter | 1. Use distortion correction software on scanner 2. Position target as close to scanner isocenter as possible (e.g., shift patient on the table) 3. Include remaining inaccuracy in PTV margin |
| Geometric accuracy (magnetic field inhomogeneities and patient-induced distortions) | Reduced geometric accuracy, especially near tissue–air interfaces | 1. Use high bandwidth acquisition 2. Acquisition of B0 map to assess patient-induced distortion. |
| PLANNING | ||
| Electron return effect | Possible skin dose, chest wall, or lung dose increase (dose increase at tissue–air interfaces) | Pay attention to skin, chest wall, and lung dose constraints in planning, carefully choose beam setup (e.g., use enough beams) |
| Electron stream effect | Irradiation dose outside the treatment field in an inferior-to-superior direction (Figure 4) | Use of bolus material to shield irradiation outside of field |
| Missing electron density information in MR-only workflow | Inaccurate dose calculation without correct electron density assignments | Development of methods for synthetic CT generation from MRI |
| High-density treatment couch material | Unpredictable dose effects by daily replanning | Avoid beam angles passing through the treatment couch edges |
| TREATMENT | ||
| Irradiation through coil | No irradiation through MR receiver coils, only through dedicated hybrid machine coils. Dedicated prone breast coil cannot be used | 1. Try to fit the dedicated MR-linac coil on top of prone patient (only for smaller patients) 2. Design a thinner, more flexible coil for the hybrid system 3. Design a new prone coil for the hybrid system |
| Fixed treatment couch | Interfractional changes in position cannot be corrected by moving the treatment couch | Use online plan adaptation strategies to account for interfractional changes in anatomy |
| Motion during treatment | Geographical miss during treatment or increased PTV margins | Use online gating or tracking when available, e.g., only beam-on when the target volume is within pre-specified boundaries |
MRI, magnetic resonance imaging; MR, magnetic resonance; OAR, organ at risk; PTV, planning target volume; CT, computed tomography; RT, radiotherapy.