Proton therapy for head and neck cancer: Rationale, potential indications, practical considerations, and current clinical evidence

Figures & data

Table I. Outcomes and toxicity rates for various head and neck cancers treated with radiation therapy.*

Clinical site	Radiation dose-limiting structure	Typical radiation dose	Grade 3+ toxicity rate	Local-regional tumor control rate	Radiation therapeutic ratio
Early-stage true vocal cord (2)	Thyroid cartilage and arytenoids	63 Gy/28 fx	<1%	∼95%	Very high
T3 Base of tongue (3)	Mandible	74.4 Gy/62 fx, BID	8%	82%	High
Nasopharynx (5)	CNS; visual apparatus	74.4 Gy/62 fx, BID	10%	76%	Fair
Paranasal sinus tumor (7)	CNS; visual aparatus	64.8-74.4 Gy at 1.2 Gy/fx, BID	20%	63%	Fair
Skin cancer with clinical perineural invasion (unpublished data)	CNS; visual apparatus	74.4 Gy/62 fx, BID	35%	55%	Fair to poor

*X-ray-based radiation therapy.

Abbreviations: CNS, central nervous system; fx, fractions; BID, twice a day.

Figure 1. This shows a comparison of an intensity-modulated radiotherapy (IMRT) plan and a proton therapy (PT) plan for a 46-year-old female with a spindle cell carcinoma involving the right maxillary sinus (pT4N0) with involvement of the infraorbital nerve, the orbit, the pterygomaxillary space, and the infratemporal fossa, status post endoscopic maxillectomy, rhinotomy, and orbital decompression. For both plans, the prescribed dose to the primary target was 74.4 Gy/CGE in 1.2 Gy/CGE twice-daily fractions. The low neck was also prescribed 50 Gy in 2 Gy fractions with 6 MV photons with concurrent weekly cisplatin during radiation therapy.

Figure 1A shows color-wash depictions of the dose distributions achieved with the PT plan (left) and the IMRT plan (right). As apparent from the axial, coronal, and sagittal images, there is greater integral dose with the IMRT plan. Figure 1B shows the dose-volume histogram (DVH) comparison between the two plans; as apparent from the curves on the far right, the target coverage was the same for the two plans, but the PT plan delivered a significantly lower dose to the optic structures, including the chiasm, the lacrimal glands, the retinae, and the optic nerves as well as the brainstem and parotids. The differences in dose to these structure are further detailed in Table II.

Table II. Comparison of normal-tissue exposure between IMRT and PT plans* for sinonasal carcinoma.

Radiation dose-limiting structure	Radiation plan (Dose)	Minimum dose	Mean dose	Maximum dose
Chiasm	IMRT (Gy)	21.8 Gy	41.4 Gy	63.1 Gy
	PT (CGE)	0.8 CGE	8.6 CGE	35.2 CGE
Right optic nerve	IMRT (Gy)	57.6 Gy	65.2 Gy	70.3 Gy
	PT (CGE)	12.7 CGE	48.7 CGE	72.5 CGE
Right retina	IMRT (Gy)	37.4 Gy	59.4 Gy	72.4 Gy
	PT (CGE)	0.3 CGE	31.4 CGE	75.9 CGE
Right lacrimal gland	IMRT (Gy)	23.4 Gy	37.7 Gy	51.9 Gy
	PT (CGE)	0 CGE	0.9 CGE	5.3 CGE
Left optic nerve	IMRT (Gy)	41.7 Gy	57.2 Gy	65 Gy
	PT (CGE)	0 CGE	6.8 CGE	29 CGE
Left retina	IMRT (Gy)	28.7 Gy	54.6 Gy	59.1 Gy
	PT (CGE)	0 CGE	0.8 CGE	14.1 CGE
Left lacrimal gland	IMRT (Gy)	11.6 Gy	30 Gy	47.1 Gy
	PT (CGE)	0 CGE	0.9 CGE	5.3 CGE

*Summary of critical-structure dose data from Figure 1. Doses for IMRT are given in Gy and doses for PT are given in CGE (cobalt gray equivalent); differences in critical-structure dose that are likely to result in differences in preservation of function are shown in bold.

Abbreviations: IMRT, intensity-modulated radiotherapy; PT, proton therapy.

Table III. Guidelines and specific considerations in treatment planning of proton therapy.

Treatment planning	Guidelines and considerations
Simulation setup	1. Verify proton stopping-power algorithm for particular treatment planning CT.
	2. Verify proton stopping power of all materials in beam path, including face mask, head holder, and table must be known.
	3. Positioning must facilitate as close a placement of beam nozzle to skin as possible to decrease penumbra.
	4. Evaluate any anatomical sites in which intrafraction motion may result in variation in target location relative to anatomy used in positioning or variations in composition in length of beam entrance path to target (4D CT and or MR).
Contouring	1. Contour all metals and assign them an appropriate Hounsfield unit.
	2. Override any metal artefact with appropriate tissue Hounsfield units.
	3. Contour any structures with potentially variable stopping power values through the course of PT treatment (e.g. a sinus cavity which could be filled with air or tissue-equivalent material); assess the potential impact of variations in sopping power; plan rescanning during treatment and adaptive replanning if necessary.
Beam-angle selection	1. Select angles that optimize the target size and volumetric shape in the beam's eye view—either by minimizing the target size or irregularity or thickness.
	2. Avoid angles that pass through metal.
	3. Avoid angles that stop the beam immediately before a critical structure; in addition to slight uncertainties in the distal range of the beam related to imprecise treatment planning systems, and daily variations in the length and composition of the beam path, there are uncertainties in the RBE at the end of the proton range.
	4. Choose angles that provide the shortest beam path to the target as integral dose increases with the proton-path length and penumbra increases with depth in tissue.
	5. Choose angles with the most stable path composition and fewest tissue-heterogeneity interfaces. The proton range in tissue varies with the different stopping powers of various tissues; intrafraction and interfraction variations in the composition of the path impact the daily proton range in tissue.
	6. Choose beam angles to match the major axes of target motion; beam angles that are perpendicular to major target motion will require much larger PTV expansions.
Uncertainties and expansions	1. Any uncertainties in beam path composition and length must be modeled and accounted for with proximal and distal SOBP margins.
	2. PTV expansion is field-specific with PT (IMRT PTV expansion may always be field specific, as daily image guidance may not be performed between segments).
	3. If entrance-path composition and length are stable, little or no PTV expansion is necessary beyond the proximal and distal SOBP margins. In the beam's eye view, PTV expansion is dominated by interfraction and intrafraction setup errors and organ motion. The use of daily image-based patient localization allows a significant reduction of PTV margin. Generally, in head and neck tumors, the PTV expansion is 3 mm.
	4. Some critical structures may warrant a PTV expansion if they abut the CTV and are at risk for injury with the prescribed PTV dose (e.g. the optic nerve or chiasm with a sinonasal tumor that invades the anterior cranial fossa).
Overall plan evaluation	1. Use the fewest fields necessary to achieve dosimetry goals for target coverage and OARs, and plan robustness. The more fields used, the more tissue exposed to integral dose. The more fields treated each session, the greater the opportunity for intrafraction motion that impacts margins and expansions. However, when complex techniques are required, critical structures are adjacent to the target volume, or there is significant uncertainty in the beam range, the use of more fields may reduce the risk of random and systematic errors and increase the robustness of the daily and overall treatment plan.
Apertures and compensators	1. The distance to the aperture edge must account for penumbra, which will be impacted by distance between nozzle and skin, the field size, and the target depth.
	2. Compensators can be used to increase distal beam edge conformality; a “smearing” value of 5 to 7 mm is used within the target volume and “smoothing” value of 8 mm is used outside the target volume I head and neck tumors to account for daily variations related to intrafraction and interfraction motion and set-error.

*These guidelines and considerations are in use at the University of Florida Proton Therapy Institute and pertinent for double scattered proton therapy in head and neck cancer as well as in other sites. With the development of intensity modulated proton therapy (IMPT) using “scanned” proton beams, there will be less indication for apertures and compensators and some of the considerations with beam angle selection may be moot.

Abbreviations: CT, computed tomography; CTV, clinical target volume; IMRT, intensity-modulated radiotherapy; MR, magnetic resonance; OAR, organs at risk; PT, proton therapy; PTV, planning target volume; RBE, relative biological effectiveness; SOBP, spread-out Bragg peak.