Proto9n Beam Radiation Oncology Update

Proton Therapy Achieves Good Long-Term Local Control of Sinonasal Cancer
Elsevier Global Medical News. 2010 Mar 3, S London

CHANDLER, Ariz. (EGMN) - Proton beam therapy yields a high 87% rate of local control at 5 years in patients with locally advanced cancer of the sinuses or nasal cavity, according to a retrospective analysis of 99 patients who were treated with this emerging technology.

The results highlight the need for multicenter prospective studies to better define how this therapy fits into the management of sinonasal cancer, presenting author Dr. Annie W. Chan said at a head and neck cancer symposium sponsored by the American Society for Radiation Oncology.

"Even with 3-D-conformal radiation therapy [3DCRT] or intensity-modulated radiation therapy [IMRT], the local control rate [of sinonasal cancer] has not been very impressive," said Dr. Chan, a radiation oncologist at Massachusetts General Hospital in Boston. "There is a lot of room for improvement in just local control."

Protons are as effective as the photons of conventional radiation therapy in tissue, she observed. But protons have some distinct physical properties - such as a lower dose where the beam first enters tissue and a rapid dose falloff as it exits the target - that offer advantages.

Dr. Chan and her coinvestigators retrospectively analyzed outcomes among 99 patients who underwent proton beam therapy at Massachusetts General Hospital between 1991 and 2003 for newly diagnosed sinonasal cancer.

The patients had a median age of 51 years, and the sex distribution was nearly equal. Two-thirds had stage T4b disease. The most common tumor types were sinonasal undifferentiated carcinoma (SNUC) or squamous cell carcinoma (34%) and neuroendocrine tumors or esthesioneuroblastomas (28%).

A third of the patients had a biopsy alone, whereas the rest had a partial or gross total resection. In all, 27% of patients received chemotherapy.

The total radiation dose was 70 GyE, of which protons made up 57%. Patients received combined proton and photon therapy because the hospital did not have a dedicated proton machine at the time, Dr. Chan explained. "For the sinus area and upper neck, we used protons alone. For the lower neck, we used photons only," she said.

With a median duration of follow-up of 5.3 years among all patients and 8.5 years among living patients, the estimated 5-year rates of local and regional control were 87% and 89%, respectively, according to Dr. Chan, who reported receiving study funding from a government grant.

The rate of local control did not differ significantly by tumor type, extent of surgery (resection vs. biopsy only), type of surgical approach (transfacial or craniofacial vs. endoscopic surgery or craniotomy alone), or T stage (T3b vs. T4a vs. T4b).

In a multivariate analysis, patients had a higher risk of death if they had a Karnofksy performance status score of 90% or lower at the time of radiation vs. a higher score (hazard ratio 3.1, P = .003); SNUC or squamous cell carcinoma vs. other tumor types (HR 2.4, P = .002); or a T4b stage vs. a T3 or T4a stage (HR 3.6, P less than .001). In contrast, the extent of surgery and the surgical approach did not significantly influence this outcome.

The first site of failure was most commonly a distant site (seen in 26% of patients), followed by a local site (11%) and a regional site (8%).

Patients had an increased risk of distant metastases if they had a primary tumor located in the sphenoid or ethmoid sinuses vs. maxillary sinuses or nasal cavity (P = .007) or a T4b stage (P = .002). "Maybe T4b stage is a group of patients in which we should consider induction chemo rather than giving induction chemo to all patients with sinonasal cancer," Dr. Chan commented.

The leading grade 3 or higher late toxicity was soft tissue toxicity (mainly sinocutaneous fistulas occurring along transfacial scars), seen in 9% of patients at 5 years, Dr. Chan said. The other toxicities of this severity were visual or ocular (7%), bone (6%), and neurologic (3%).

Patients were more likely to develop a grade 4 sinocutaneous fistula if they had a surgery by the transfacial or craniofacial surgical approach (P = .03). No one who was treated with an endoscopic approach or craniotomy alone experienced this complication.

In addition, patients were more likely to develop grade 2 or higher neurologic toxicity if their duration of radiation therapy was 40 days or less (P = .002) or if they had adenoid cystic carcinoma vs. other tumor types (P = .002). "For the adenoid cystic carcinoma, we usually cover the perineural spread in our clinical target volume, which could increase the risk of neurologic toxicity," she commented.

A comparison of the 5-year local control rates achieved with proton therapy in the study and with 3DCRT or IMRT in historical series favored the former (87% vs. 21%-63%). "However, all of these studies are single-institution retrospective studies," Dr. Chan cautioned. "Definitive conclusions therefore cannot be made."

Rates of grade 3 or higher late complications with proton therapy were comparable to those with the other two types of radiation therapy.

"Proton beam results in very encouraging outcomes in patients with locally advanced sinonasal cancer," Dr. Chan concluded. "Multi-institutional prospective studies, particularly in comparison with IMRT, are necessary to define the role of protons in the treatment of this rare and aggressive malignancy."

Commenting on the study in a related press briefing, Dr. Louis B. Harrison, chair of radiation oncology at the Beth Israel Medical Center in New York, said that the cost of proton therapy "is something that we all grapple with."

The incremental benefit of this therapy is something that the field is striving to learn, according to Dr. Harrison, who reported having no conflicts of interest. But "you can't learn unless there are enough centers doing it, so that you can accumulate the data. So this is really something that requires a lot of attention in our current-day health care debate."

Dr. Chan disclosed no conflicts of interest.


--------------------------------------------------------------------------------


click for full