Mask Loosens Mid-Course in Head and Neck IMRT With an SIB
A 58-year-old man with T2N2M0 base of tongue squamous cell carcinoma is receiving definitive IMRT, 66 Gy in 30 fractions, with weekly cisplatin. He was simulated in a 5-point thermoplastic mask with a bite block. At fraction 15 he reports that he has lost 6.5 kg, about 8% of his baseline weight, that the mask feels loose, and that he has grade 2 mucositis. The CBCT shifts for the last three fractions were 6, 7, and 8 mm, all in the anterior and inferior direction, and the therapist reviewing the CBCT notes that the mandible appears rotated relative to the planning scan. The plan is a simultaneous integrated boost: 66 Gy to the GTV, 60 Gy to the high-risk PTV, and 54 Gy to the elective neck, with a parotid mean dose constraint of 26 Gy.
1. What should the therapist do before delivering fraction 16, and why is simply applying the CBCT shift the wrong response?
Read the worked answer
The therapist should stop the schedule, inform the radiation oncologist and the medical physicist, and initiate a re-simulation with a new mask. Applying the shift and treating is wrong because the problem is not a random setup offset. Three consecutive shifts in the same anterior-inferior direction, growing in size, together with 8% weight loss, a loose mask, and a rotated mandible on CBCT, describe a systematic change in the patient's anatomy and immobilization. A shift corrects the patient's position relative to the couch, not the fact that the mask no longer reproduces the anatomy it was made for.
The clinical consequences are concrete. The GTV and high-risk PTV were contoured on the original scan; with the mandible rotated and the tongue base moved, the 66 Gy region may now sit partly in tissue that no longer needs it and partly miss tissue that does. The parotid may have moved into or out of the high-dose volume. And because the shifts exceed the setup margin used in planning, the delivered dose is already different from the planned dose.
Document the observations, the weight, and the CBCT findings, and treat this as a replanning event, not a setup problem.
2. After a new CT is acquired in a new mask, what dosimetric comparison should be made, and what results would trigger a formal plan revision?
Read the worked answer
Recalculate the original approved plan on the new CT in the new mask, then compare three things against the original plan and against the constraints.
Target coverage. Look at PTV D95 and D99 for the 66 Gy, 60 Gy, and 54 Gy levels. A common working trigger is the high-dose PTV D95 falling below 95% of the prescription, or a visible cold region inside the GTV. The SIB makes this critical, because a shift that barely moves the 54 Gy elective volume can move the 66 Gy boost off the tumor base.
Organs at risk. Parotid mean dose is the headline constraint at 26 Gy. If the recalculated mean dose exceeds 26 Gy, or rises by several gray compared with the approved plan, that alone justifies re-optimization, since xerostomia is a permanent late effect. Also check spinal cord, mandible, and constrictors.
Delivered dose so far. Use the CBCT shifts and, if available, deformable registration to estimate what the first 15 fractions actually delivered. This informs whether the remaining fractions need compensation and what the patient should be told.
If coverage is maintained, parotid dose is within tolerance, and the cumulative dose is acceptable, the original plan may continue in the new mask with new image-guidance baselines. Any failure of those checks means re-plan. The decision is documented with the numbers that drove it.
3. What is the risk of continuing to treat with the loose mask and the original plan, stated in tissue-specific terms?
Read the worked answer
The anterior-inferior drift moves the treated volume relative to the tumor base and the neck. The most likely pattern is a hot region displaced toward the mandible and anterior oral cavity, with the posterior tongue base and the high-risk nodal levels receiving less than planned. In a 30-fraction SIB course the residual 15 fractions carry half the total dose, so a persistent 6-8 mm error is not a small perturbation.
Specific risks: underdosage of the GTV at the tongue base raises the chance of local recurrence at the primary site; underdosage of the 60 Gy high-risk PTV raises the risk at involved nodes; the parotid can be pushed into the high-dose region and lose function; the mandible can receive more dose than planned, which matters for osteoradionecrosis, particularly with concurrent cisplatin and poor dentition. Meanwhile the patient's mucositis is already grade 2 and would likely worsen with a hot spot in the oral cavity.
Continuing also destroys the dosimetric record. Once the setup is known to be invalid, any later reconstruction of what the patient received becomes guesswork, which compromises both the clinical decision and any future comparison.