Concurrent Dose-Escalated Proton Therapy and IMRT for Sinonasal Cancer: Initial Clinical Experience

Reporter: Abigail Berman Milby, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 21, 2012

Presenting Author/Institution: Mark William McDonald, MD Indiana University Health Proton Therapy


  • Sinonasal malignancies are often managed with radiotherapy and chemotherapy alone without surgical resection. In this definitive setting, high doses are needed and therefore the long-term risk to critical structures, particularly the optic structures, is often of concern with the location of these tumors.
  • For patients requiring comprehensive head and neck radiation for sinonasal malignancies, proton therapy may provide dosimetric advantages at the primary tumor site and opportunities for dose escalation near critical normal structures.
  • Proton therapy, however, is not well-suited to treating the long craniocaudal treatment distance required for treatment of the primary disease as well as the neck. In addition, proton therapy is not ideally suited to treating the cervical lymphatics.
  • Intensity-modulated radiation therapy (IMRT) often the better modality for achievement of a homogenous dose to an irregular cervical lymphatic PTV.
    • With IMRT, a simultaneous integrated boost (SIB) is easy to deliver to grossly involved nodes.
    • IMRT is able to spare the skin, parotid glands, submandibular glands, and brachial plexus due to excellent conformality.
    • In addition, IMRT decreases utilization of proton room time.
  • For all of these reasons, using combinations of proton therapy and IMRT for patients requiring sinonasal and neck radiation is attractive. This is a report of the initial clinical experience with the use of concurrent proton therapy and IMRT.


  • Between 2010-2011, there were 10 patients treated with dose-escalated comprehensive head and neck radiation for sinonasal malignancies using a matched proton-IMRT technique.
  • All patients had a T4 primary tumor.
  • There were 5 tumors in the nasopharynx (NP), 3 ethmoid sinuses, 1 maxillary sinus, and 1 nasal cavity.
  • Histologies included sinonasal undifferentiated carcinoma (SNUC), squamous cell carcinoma, esthesioneuroblastoma, and Ewing's sarcoma.
  • Proton therapy was delivered with uniform scanning and was directed at the primary tumor site matched to concurrent IMRT to the bilateral necks. Because a LINAC was not available at the presenters' institution, treatment was coordinated across two facilities.
  • The IMRT plan was half-beam blocked at the matchline, which was feathered by 0.5 cm once during the treatment course. The matchline was generally placed at the bottom of the clivus and every effort was made to avoid matching through high-risk volumes or gross tumor.
  • All patients underwent PET/CT 12 weeks after completion of radiotherapy.


  • The 5 NP patients were treated to a median dose of 76 Gy (RBE). The four patients with SNUC were treated to a median dose of 72 Gy (RBE), two of whom were post-operative (68.4-70.2 (RBE)) and two definitive (75.6 Gy(RBE)).
  • All patients received concurrent chemotherapy except for the 2 treated post-operatively.
  • Median V26 for the maximally spared parotid was 54% (range 17-90%) and the V26 for the contralateral parotid was 68% (51-93%)
  • Median V30 for the oral cavity was 38% (range 10-94%), including one patient who had oral cavity involvement.
  • Toxicity was as follows:
    • 1 grade 3 mucositis and a 1grade 4 hematologic toxicity, and treatment was stopped at 68 Gy. This patient remains without evidence of disease at 23 months.
  • Treatment was stopped early at 68 Gy in one patient with paranasal Ewing's sarcoma who developed leptomeningeal disease and died during the treatment course.
  • The remaining patients are alive with a median follow up time of 10 months (3=21 months).
  • No patients developed matchline fibrosis or a matchline recurrence. No G-tubes were required more than 3 months out from completion of therapy.
  • There have been no visual or CNS toxicities.
  • A total of two patients developed distant metastases.
  • One patient developed a second primary cancer (squamous cell carcinoma).
  • One patient with esthesioneuroblastoma developed a local recurrence that was identified as next to mesh leading to possible dose attenuation in that region. The local recurrence was effectively treated by stereotactic radiosurgery.
  • One year overall survival (OS) and locoregional control was approximately 75% by Kaplan Meier analysis

Author's Conclusions

  • The respective advantages of proton therapy and IMRT can be combined in concurrent treatment to provide dose-escalated comprehensive head and neck radiation for sinonasal malignancies without any apparent toxicities or complications related to the match line.
  • Initial locoregional control is promising with 8/9 patients with controlled local disease.

Scientific/Clinical Implications

  • The authors report good local control and low rates of toxicities for sinonasal tumors treated with a unique approach of IMRT for cervical neck radiation and proton therapy for primary tumor irradiation.
    • These results should be interpreted in the context that this is a small, single center study with less than 1 year median follow up and some patients with as little as 3 months follow up.
  • This is one of the first reports of its kind. A study from Noel et al. at the Centre de Protonthérapie in Orsay, France (Cancer Radiotherapy, 2002) reported similar treatment for nasopharyngeal cancer alone. This report, however, combined proton beam therapy and 3D conformal photons, not IMRT.
  • There are several technical factors that must be considered when using proton beam therapy for sinonasal tumors: the amount of air in the sinuses (versus tumor or secretions) can vastly change the dose distribution. Therefore, frequent verification scans should be used in this population to assure unchanged amounts of air.
  • Carbon ion therapy is another form of particle beam therapy that is often advocated for malignancies approaching the base of skull. Carbon would not be possible for the cervical necks and therefore a similar matched technique with IMRT would need to be used.
  • Intensity modulated proton therapy (IMPT) may be able to improve the dose distribution even further and permit treatment of the nodal chains with IMPT as well.
  • When using two different facilities daily, there is inherently more risk for error. The authors described the measures they take to reduce this error, including checking a light field every day, confirmation to switching matchlines, shared immobilization, one set of contours, IMRT half-beam block. Careful evaluation of daily and composite dose is needed to safely employ this technique across institutions.
  • This technique may be applicable to other sites of radiotherapy as well, including treatment of the mantle field with photons with the mediastinal portion of the treatment with protons. This technique is being piloted at the University of Pennsylvania (Goldsmith et al. ASTRO 2012).


Survivorship Benefits of Proton Therapy
by OncoLink Editorial Team
June 17, 2015

Cancer Center Advertising: Doing More Harm Than Good?
by Rodney Warner, JD
July 21, 2016