Prostate IMRT Plan With Unexpected Rectal Dose and a Hip Prosthesis
A 69-year-old man with cT2cN0M0 prostate adenocarcinoma (Gleason 4+3, PSA 11 ng/mL) is planned for 78 Gy in 39 fractions to the prostate with a 7-field IMRT technique. He has a left total hip arthroplasty placed 6 years ago. The planner reports that the PTV coverage is acceptable (D95 = 77.4 Gy), but the rectal V70 is 18% (planning constraint V70 < 10%), and the left femoral head mean dose is 51 Gy. The patient is on aspirin and reports occasional nocturia. The attending asks whether to accept, re-plan, or change technique. The CT dataset was acquired with a 3 mm slice thickness and the prosthesis causes streak artifact across the posterior pelvis on several slices.
1. How should the streak artifact and the prosthesis influence your evaluation of the reported dose metrics, and what practical steps would you take before accepting or rejecting the plan?
Read the worked answer
Streak artifact from a metallic hip prosthesis is a CT-number error, not a geometric error. Because the planning system computes dose from electron density derived from those CT numbers, the dose in and immediately downstream of the artifact can be wrong by several percent, and the reported rectal V70 and femoral head mean dose may not be trustworthy if the artifact overlaps those structures.
Practical steps:
- Review the artifact on the planning CT slice by slice and contour the artifact region; check whether the rectal contour or PTV overlaps artifact.
- Where possible, use a metal-artifact-reduction reconstruction or acquire an additional scan and fuse; if the dose calculation depends on corrupted densities, the plan should be recalculated with corrected densities (assign water-equivalent density to artifact regions) and the metrics compared.
- Verify that the prosthesis was contoured as a high-density structure so the TPS does not treat it as soft tissue; this affects both attenuation and the dose distribution.
- Confirm the reported metrics were computed on the same calculation grid (e.g., 2-3 mm) used clinically and that the dose was not reported on a coarse grid.
- Only then decide whether the rectal V70 of 18% is real. If it is real, the plan fails the constraint and should be re-planned regardless of artifact.
2. The rectal V70 is confirmed at 18% after artifact correction. What re-planning strategies would you consider, and how would you weigh them against the patient's comorbidities and the femoral head dose?
Read the worked answer
The rectal V70 of 18% exceeds the institutional constraint of <10% and is associated with late rectal toxicity, so the plan should not be accepted as-is. Options:
- Re-optimize the IMRT fluence with stricter rectal objectives (V70, V60, Dmax) and allow modest PTV heterogeneity; this is usually the first step and often recovers several percent of V70 without changing technique.
- Add a rectal spacer (hydrogel) if the patient is a candidate; this physically displaces the rectum and reliably reduces rectal dose, but requires a second procedure and is not appropriate in every patient.
- Consider a different technique (VMAT with rectal avoidance sectors, or a brachytherapy boost) if IMRT re-optimization cannot meet constraints.
Weighing the femoral head: a mean dose of 51 Gy to the left femoral head is above typical tolerance for avascular necrosis risk (keep mean < 45 Gy and V40 low where possible). The prosthesis itself is not a contraindication to pelvic RT, but the femoral head dose should be reduced if achievable without compromising PTV coverage. Because the patient is on aspirin, the bleeding risk is not a RT planning issue, but his nocturia suggests baseline urinary morbidity that should be documented.
Decision: re-plan with tighter rectal and femoral head objectives; if V70 remains >10%, add a spacer or change technique. Do not accept the current plan.