proton therpay for acc hed23573

ORIGINAL ARTICLE

Proton therapy for head and neck adenoid cystic carcinoma: Initial clinical outcomes

Okechukwu R. Linton, MD,1 Michael G. Moore, MD,2 Joseph S. Brigance, MD,2 Don-John Summerlin, DMD, MS,3 Mark W. McDonald, MD1,4*

1Department of Radiation Oncology, Indiana University School of Medicine, Indianapolis, Indiana, 2Department of Otolaryngology/Head and Neck Surgery, Indiana UniversitySchool of Medicine, Indianapolis, Indiana, 3Department of Pathology, Indiana University School of Medicine, Indianapolis, Indiana, 4Indiana University Health Proton Therapy Cen-ter, Bloomington, Indiana.

Accepted 10 December 2013

Published online 13 March 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/hed.23573

ABSTRACT: Background. The purpose of this study was to report out-comes of proton therapy in head and neck adenoid cystic carcinoma.Methods. We conducted a retrospective analysis of 26 patients treatedbetween 2004 and 2012. Twenty patients (77%) had base of skullinvolvement; 19 (73%) were treated for initial disease and 7 (27%) forrecurrent disease. Twenty patients were treated postoperatively, 6 afterbiopsy alone and 24 had positive margins or gross residual disease.Median dose delivered was 72 Gy (relative biological effectiveness[RBE]).Results. Median follow-up was 25 months (range, 750 months). The 2-year overall survival was 93% for initial disease course and 57% for

recurrent disease (p5 .19). The 2-year local control was 95% for initialdisease and 86% for recurrent disease (p5 .48). The 2-year distantmetastatic rate was 25%. Late toxicity of grade 0 or 1 was seen in 17patients, grade 2 in 5, grade 3 in 2, grade 4 in 1, and grade 5 in 1.Conclusion. Initial outcomes of proton therapy are encouraging. Longerfollow-up is required. VC 2014 Wiley Periodicals, Inc. Head Neck 37:117124, 2015

KEY WORDS: adenoid cystic carcinoma, proton therapy, head andneck, dose-escalation, radiation therapy

INTRODUCTIONAdenoid cystic carcinoma (ACC) is an uncommon secre-tory gland tumor most often arising in salivary tissue inthe head and neck. Its natural history is marked by a pro-pensity for perineural tumor spread and both local recur-rence and distant metastatic disease. Many patients have ahistory of slow indolent tumor growth with a prolongedclinical course, but those with lymphatic invasion, solidtumor histology, or high-grade transformation seem tohave a more aggressive clinical course and reduced sur-vival.13 The incidence of ACC seems to be decreasingover time, and there is suggestion that survival is superiorin women compared to men.4 Surgical resection and post-operative radiotherapy are the standard of care, whereassystemic therapy has been of little utility.5

Patients with advanced tumor stage disease68 andinvolvement of the skull base have poor outcomes,9,10 asdisease cannot usually be surgically extirpated and thelower radiation tolerance of closely adjacent critical nor-mal structures, such as the optic apparatus, brainstem, and

temporal lobes, can preclude delivery of radical doses ofradiotherapy. In these patients, local disease control is ofspecial importance, given the profound sequelae of dis-ease progression at the skull base.

Given the suboptimal results of X-ray-based radiationtherapy for patients with unresectable or gross residualdisease or those with skull base involvement,6,9,1113 doseintensification has been explored. Prior studies have sug-gested a dose response relationship for ACC,1416 wherehigher doses seem to be associated with improved anddurable local tumor control. Neutron therapy has beenused to provide a biologically more potent dose of radia-tion therapy for patients with unresectable salivary glandtumors: although the 5-year locoregional control rate forpatients with ACC and skull base involvement was a dis-appointing 15% to 23%,11,12 it improved with addition ofa Gamma Knife radiosurgery boost.17

Proton therapy is a modality of radiation therapy with abiologic effect normalized to be identical to megavoltageX-rays, but with superior dose localization facilitatingdose-escalation while respecting the dose constraints ofsurrounding normal structures.18 We report initial clinicaloutcomes of proton therapy at our institution for head andneck ACC.

MATERIALS AND METHODSInstitutional review board approval was obtained for

this retrospective review. Twenty-seven consecutivepatients were treated at the Indiana University Health Pro-ton Therapy Center between June 2004 and April 2012

*Corresponding author: M. W. McDonald, Indiana University Health ProtonTherapy Center, 2425 North Milo B. Sampson Lane, Bloomington, IN 47408-1398. E-mail: [email protected]

This work was presented in part at the 95th annual meeting of the AmericanRadium Society, April 27 to May 1, 2013, Scottsdale, Arizona.

Contract grant sponsor: This work was supported in part by the Jesse N.Jones, III, Memorial Fund for Head and Neck Cancer Research at the IndianaUniversity Melvin and Bren Simon Cancer Center.

HEAD & NECKDOI 10.1002/HED JANUARY 2015 117

for head and neck ACC. One patient who was reirradiatedwith palliative intent (30 Gy in 10 fractions) wasexcluded from analysis. The remaining 26 patients weretreated with curative intent: 19 (73%) were treated fortheir initial disease course and 7 (27%) were treated forrecurrent disease. Twenty patients (77%) were treatedafter surgery, and 18 of these (90%) had positive marginsor gross residual disease. The remaining 6 patients (23%)were treated after biopsy alone.

Skull base involvement was defined as tumor involvingthe infratemporal fossa, clivus, foramen ovale, foramenrotundum, cavernous sinus, sphenoid sinus, orbit, or intra-cranial tumor extension. In the subset of 19 patientstreated in their initial disease presentation, 14 had skullbase involvement and none had metastatic disease at pre-sentation. In the subset of 7 patients treated for recurrentdisease, 6 had received prior head and neck radiation, 6had skull base involvement, and 3 had known distant met-astatic pulmonary disease, but received aggressive localtherapy because of the clinically indolent nature of theirdistant metastases and the importance of obtaining localcontrol. Table 1 lists the patients characteristics. Half thepatients were treated by the senior author.

Included in the analysis are 2 patients who receivedsome component of photon therapy. One patient wasreferred for a proton therapy boost after initial delivery of61.2 Gy of intensity-modulated radiation therapy (IMRT)at an outside institution. One patient received 14.4 Gy ofIMRT before beginning proton therapy because of herweight initially being in excess of our table tolerance.The remaining patients were treated exclusively with pro-ton therapy.

Proton treatment planning involved creation of customimmobilization devices, including a thermoplastic facemask, alpha cradle body mold, and acquisition of a non-contrast treatment planning CT scan with 1-mm slicethickness. Preoperative and postoperative imaging wascoregistered to delineate the initial extent of tumor, areasof residual disease, and postoperative anatomy. The targetvolume covered the primary tumor bed and course of theinitial tumor and included coverage of the skull base,tracking any involved cranial nerve through the skull baseforamina to the surface of the brainstem. Ipsilateral cervi-cal lymphatics were irradiated in the 1 node-positivepatient. Limited regional nodal irradiation was deliveredelectively in 7 other patients per the preference of one ofthe treating physicians.

Proton treatment planning was 3-dimensional (3D) con-formal radiotherapy using uniform scanning beam deliv-ery, which delivers a uniform spread of Bragg peakacross each field.19 The choice of beam angles and num-ber of beams varied by the extent and complexity of eachindividual patients tumor. Brass apertures were fabricatedto define the shape of each field and conform to the lat-eral extent of the target, with Lucite range compensatorsmilled to shape the distal end of the proton beam, achiev-ing distal conformality to the target. Treatment optimiza-tion involved multiple iterative adjustments in individualfield shape and design to achieve the desired target cover-age and normal tissue sparing. Our institutional practiceis to use a 2-mm expansion on clinical target volumes tocreate a planning target volume, which accounts for

patient set-up uncertainties. Individual beam parametersincluding compensator smearing were selected with con-sideration of the surrounding tissue heterogeneities andpotential uncertainties inherent in each unique beam path.Range uncertainty is dependent on depth or energy butgenerally 2 to 4 mm in the head and neck. Details of ourproton beam delivery system have been previously pub-lished.20 Daily kilovoltage orthogonal X-ray images weretaken to align the patient before treatment of each field,with adjustments made on a robotic patient positionerwith 6 degrees of freedom to achieve near stereotactictreatment alignment, verified by a physician before treat-ment delivery.21

Patients were treated with once daily fractionation of1.8 to 2 Gy (relative biological effectiveness [RBE]). Pro-ton dose is expressed in Gy (RBE) with an RBE of 1.1compared to megavoltage X-ray therapy. A strategy ofhigh-dose radiotherapy was pursued per institutionalapproach, based on the prior published experience withhigh-dose proton radiotherapy in this disease.22 The totaldose administered was 75.6 Gy (RBE) for gross diseaseand 6670.2 Gy (RBE) for negative margins. For thosewith positive margins, the dose was 70.2 to 72 Gy (RBE),attenuated to 66.6 Gy (RBE) in 1 patient because of prox-imity to the optic nerve with a single sighted eye, and to68.4 Gy (RBE) in 1 patient because of physician prefer-ence. No patient received concurrent systemic therapy.

Local control (LC), distant metastasis (DM), and over-all survival (OS) were the events analyzed. Local tumorrecurrence was defined as enlargement of residual tumoror regrowth of tumor at the primary tumor site or withinor adjacent to the volume of the preoperative tumorextent. Patients with metastatic disease at the time oftreatment (n5 3) were excluded from analysis in determi-nation of the estimate of the risk of developing DM. Thetime of follow-up was measured from the end date ofradiation treatment. Two-year event estimates were calcu-lated using the KaplanMeier method and 95% confi-dence intervals (CIs) are presented. The log-rank test wasused to perform univariate analysis of patient and diseasecharacteristics. Variables with a p value < .2 on univari-ate analysis were selected for inclusion in a Cox multivar-iate regression analysis using the block entry method toexplore association with LC and OS.

Information on acute and late toxicity was retrospec-tively gathered from weekly treatment status notes, thetreatment completion summary, follow-up notes, and cor-respondence from follow-up with other physicians andgraded according to the Common Terminology Criteriafor Adverse Events (version 4.0) of the National CancerInstitute.

RESULTSThe median follow-up on all patients was 25 months

(range, 754 months) and 27 months in patients that arealive. The 2-year estimate of OS was 82% (95% CI, 66%to 98%). The 2-year estimate of LC was 92% (95% CI,82% to 100%). The 2-year estimate for the developmentof DM (23 of 26 patients in analysis) was 25% (95% CI,5% to 45%). DM developed in 5 patients at 1 through 25months after treatment. Four patients had DM as the ini-tial site of relapse. One had synchronous local and distant

LINTON ET AL.

118 HEAD & NECKDOI 10.1002/HED JANUARY 2015

metastases. No patients developed cervical nodalrecurrence.

Univariate analysis did not demonstrate a statisticallysignificant relationship between any patient or treatmentvariable and LC or OS. The variables analyzed were initialversus recurrent disease, presence or absence of skull baseinvolvement, sex, age (continuous variable), primary tumorsite, surgical resection versus biopsy alone, solid versusnon-solid histology, positive versus negative margin status,perineural invasion, lymphovascular space invasion, pres-ence or absence of gross residual disease, T classification,N classification, group stage, total dose (continuous vari-able), and prior radiation versus radiation nave.

Multivariate Cox regression analysis for OS in a modelincluding the variables of age (p5 .16), prior radiation(p5 .06), lymphovascular space invasion (p5 .06), pres-ence of gross residual disease (p5 .17), perineural

TABLE 1. Patient characteristics.

Characteristic No. of patients (%)

SexMale 12 (46)Female 14 (54)

Age, yMedian 46Range 2173

HistologyCribriform/tubular 18 (69)Solid 8 (31)Primary tumor site

Ethmoid sinus 2 (8)Maxillary sinus 5 (19)Nasal cavity 2 (8)Nasopharynx 2 (8)Orbit 4 (15)Parotid gland 7 (26)Submandibular gland 4 (15)Major salivary 11 (42)Minor, accessory, or ectopic salivary 15 (58)

Skull base involvementCavernous sinus 9 (35)Foramen rotundum 6 (23)Foramen ovale 6 (23)Infratemporal fossa 6 (23)Orbit 4 (15)Clivus 3 (12)Sphenoid sinus 2 (8)

Treated at time ofInitial presentation 19 (73)Recurrent disease 7 (27)

Base of skullInvolved 20 (77)Not involved 6 (23)

Margin statusPositive 24 (92)Negative 2 (8)

Gross residual disease presentYes 12 (46)No 14 (54)

Perineural invasionYes 19 (73)No 7 (27)

Lymphovascular space invasionYes 8 (31)No 18 (69)

Bone invasionYes 14 (54)No 12 (46)

T classificationI 2 (8)II 0III 4 (15)IV 20 (77)

N classificationN0 25 (96)N2 1 (4)

M classificationM0 23 (89)M1 3 (12)

Surgery before proton therapyYes 20 (77)No 6 (33)

Neck dissection

TABLE 1. Continued

Characteristic No. of patients (%)

Yes 13 (50)No 13 (50)

Type of surgeryParotidectomywith facial nerve preservation and infratem-poral fossa resection

4

with facial nerve sacrifice, infratemporalfossa resection

1

with facial nerve sacrifice, infratemporalfossa resection, segmentalmandibulectomy

1

with facial nerve preservation 1Orbital exenteration 2Orbital exenteration, anterior skull baseresection, maxillectomy

1

Orbital debulking 1Maxillectomy with orbital floor resection 1Maxillectomy with orbital floor resection,infratemporal fossa resection

1

Transnasal, transethmoid anterior skull baseresection

1

Transoral, transpalatal partialnasopharyngectomy

1

Nd:YAG* laser transmaxillary debulking ofmaxillary and nasopharyngeal tumor

1

Total rhinectomy with transnasal, transeth-moid anterior skull base resection

1

Wide local excision of buccal mucosal tumor 1Submandibulectomy 1Submandibulectomy with partialmandibulectomy

1

Extent of surgeryGross total resection 14 (54)Subtotal resection 8 (31)Biopsy only 4 (15)

Prior radiation therapyYes 6 (23)No 20 (77)

Proton dose in Gy (RBE)Median 72Range 6675.6

Abbreviations: Nd:YAG, neodymium-doped yttrium aluminum garnet; RBE, relative biologicaleffectiveness.

PROTON THERAPY FOR HEAD AND NECK ADENOID CYSTIC CARCINOMA


invasion (p5 .17), and skull base involvement (p5 .17)yielded no variables with a statistically significant associ-ation with OS. Multivariate Cox regression analysis forLC in a model, including the variables of sex (p5 .064)and age (p5 .11) yielded no variables with a statisticallysignificant association with LC.

The 2-year LC for patients treated for their initial dis-ease course was 95% (95% CI, 85% to 100%) and forrecurrent disease it was 86% (95% CI, 60% to 100%;Figure 1; p5 .48). The 2-year OS for those treated forinitial disease was 93% (95% CI, 81% to 100%) versus57% for those with recurrent disease (95% CI, 20% to

94%; Figure 2; p5 .19). The 2-year LC for patients with-out skull base involvement was 100% and 90% (95% CI,77% to 100%) for those with skull base involvement (Fig-ure 3; p5 .21). The 2-year OS for those without skullbase involvement was 100% compared to 77% (95% CI,58% to 97%) for those with skull base disease (Figure 4;p5 .17).

The 2-year LC for patients irradiated after biopsy only(n5 6) was 100% compared to 90% (95% CI, 77% to100%) for patients treated with surgery and postoperativeradiation (p5 .97). The 2-year LC for patients treatedwith macroscopic residual disease with or without

FIGURE 1. Local control for patients treated for initial versusrecurrent disease.

FIGURE 2. Overall survival for patients treated for initial versusrecurrent disease.

FIGURE 3. Local control for patients treated for skull base versusnon-skull base disease.

FIGURE 4. Overall survival for patients treated for skull base ver-sus non-skull base disease.

LINTON ET AL.


surgical resection (n5 12) was 93% (95% CI, 79% to100%) compared to 92% (95% CI, 76% to 100%) forpatients treated without gross residual disease (p5 .64).

Fourteen patients with skull base involvement weretreated with proton therapy in their initial disease presen-tation. Among these patients, the median follow-up timewas 27 months (range, 1150 months). The 2-year esti-mates of LC, OS, and DM were 93%, 91%, and 32%,respectively (Figure 5).

Among the 7 patients with recurrent disease, 6 hadreceived one or more courses of prior radiation therapy.The median follow-up for these 7 patients was 24 months(range, 743 months). The 2-year estimates of LC, OS,and DM were 86%, 57%, and 25%, respectively.

Pattern of failure analysis

Isolated local recurrences occurred in 2 patients at 3months and 45 months after treatment, respectively. Bothtumors originated in the ethmoid sinus, involved the baseof skull, and both patients had gross residual disease atthe time of radiation. One patient had been treated forrecurrent, previously irradiated disease and progressedwithin the 75.6 Gy (RBE) dose region. The secondpatient was treated for their initial disease presentation,receiving radiation only to the postoperative volume ofresidual tumor, and developed recurrent tumor within theprimary tumor bed, which, for unclear reasons, wasexcluded from the radiation treatment volume. It wasnoted that in this patient, the preoperative MRI had notbeen coregistered to aid in delineation of the initial dis-ease extent. One additional patient with a nasal cavity pri-mary developed an orbital apex recurrence accompaniedby bone metastases 2 months after radiation therapy, andwas noted to have suboptimal coverage of the foramenrotundum.

Treatment toxicity

Acute toxicity included transient radiation dermatitis ofmaximum grade II in 16 patients, lateralized mucositis ofmaximum grade II in 7 patients, and 2 patients with gradeII serous otitis. Of the 2 patients treated for orbital tumorswith an intact eye, 1 developed an acute grade III kerati-tis requiring treatment to be held temporarily for resolu-tion. One other patient with a sinonasal primary had abrief treatment interruption for initiation of treatment of anasal cavity methicillin-resistant staphylococcus aureusinfection. All patients completed the prescribed course ofradiation.

Late toxicity of grade 0 or 1 was seen in 17 patients,grade 2 in 5 patients, grade 3 in 2 patients, grade 4 in 1patient, and grade 5 in 1 patient. The grade 3 toxicitiesincluded mandibular osteoradionecrosis in 1 patienttreated for initial disease in which 9 cm3 of the mandible(11.6% relative volume) received a dose of 70 Gy (RBE)or higher, with a maximum mandible dose of 82.9 Gy(RBE). The mandible had not been contoured by thetreating physician, which may explain why alternate beamarrangements were not used to keep the mandible atacceptable dose constraints. Hyperbaric oxygen therapywas used with improvement. A second patient who previ-ously received 50.4 Gy of photon therapy was reirradiatedfor recurrent parotid disease 6 years later to 72 Gy (RBE)and 27 months later developed grade 3 otologic toxicityrequiring tympanic membrane reconstruction. The grade 4toxicity was anticipated unilateral vision loss in a patientwho was irradiated for an orbital tumor that directlyinvolved the optic nerve and consented to the expectationof vision loss. The single case of late grade 5 toxicitydeveloped in a patient with a history of an extensivenasopharyngeal ACC previously irradiated with 54 Gy ofIMRT and who later received 3 Gamma Knife radiosur-gery treatments to different areas of recurrent head andneck disease over a 6-year timeframe. None of thepatients prior radiation treatment plans had been archivedin digital format, so it was impossible to accuratelyreconstruct the lifetime composite dose distribution. Twoyears after the most recent prior radiation, in the absenceof distant metastatic disease, the patient was reirradiatedto 75.6 Gy (RBE) to progressive painful gross recurrentdisease in the infratemporal fossa and frontal sinus. Thepatient had radiographic and clinical regression of diseaseand improvement in pain, but 6 months later developed acerebrospinal fluid leak and meningitis, and subsequentlydied.

DISCUSSIONPommier et al22 have reported results of high-dose pro-

ton therapy for patients with skull base ACC, finding a 5-year local control of 93% (100% at 2 years) in 23 patientswho had received a median dose of 75.9 Gy (RBE),mostly treated with a hyperfractionated accelerated con-comitant boost technique. These results are encouragingin a population of patients with advanced complex dis-ease, and support a strategy of dose escalation, which wasaccomplished with acceptable toxicity using proton ther-apy. All of these patients received some component ofphoton therapy (median proton component was 60% of

FIGURE 5. Local control, overall survival, and distant metastasisin patients treated for initial disease with skull base involvement.



the dose), due in part to the limited clinical availability ofthe former Harvard Cyclotron Laboratory, in contrast toour own experience, which used once daily fractionationand exclusively proton therapy in all but 2 patients.Among our subset of 14 patients with base of skullinvolvement treated for initial disease presentation, the 2-year local control was 93% after a median dose of 72 Gy(RBE) in once daily fractionation.

A review of the literature on ACC, which reportsresults by base of skull involvement or T classification(Table 2) suggests that the results achieved with high-dose proton therapy are among the most favorablereported in the literature for patients with advanced-stagedisease. Such comparisons among different retrospectiveclinical series should be interpreted with caution as thereare marked limitations including the inability to controlfor multiple prognostic factors and selection biases.

Despite these limitations, we believe these data supportdose intensification for high-risk ACC based on the prom-ising initial LC. Our institutional approach to patientswith ACC is surgical resection and postoperative radio-therapy, given that multiple series have suggested thatpatients treated with radiotherapy alone have inferior dis-ease control and survival compared with those treatedwith surgery and postoperative radiotherapy. Skull baseinvolvement and the probability of positive margins orgross residual disease in the cavernous sinus does not, inour opinion, preclude resection of extracranial diseasewith the plan for postoperative high-dose proton therapy.For patients with skull base involvement, gross residualdisease, or positive margins, we utilize high-dose postop-erative proton therapy. Patients who decline surgery orare inoperable may be treated with palliative intent,referred for consideration of neutron therapy, or treatedwith high-dose proton therapy depending on clinicianjudgment of the feasibility of delivery of radical doses ofradiation. Excellent results have been reported with post-operative standard dose photon therapy among patientswith negative margins.15 However, in patients with ear-lier-stage disease in more favorable sites where there isless rationale for dose escalation, proton therapy mayoffer an advantage in mitigating toxicities of therapy. Fig-

ure 6 illustrates a clinical example in which proton ther-apy was used to eliminate oral cavity irradiation andpreserve salivary gland function in a patient with a buccalmucosa ACC.

Patients treated more intensively, either with high-dosephoton6 or proton22 therapy, or neutron therapy withradiosurgical boost,17 seem to have improved disease con-trol compared with older series of nonoperative manage-ment with lower-dose photon therapy, raising thepossibility of using more conformal radiotherapy techni-ques for nonsurgical management in patients for whomradical resection would be excessively morbid. The com-paratively poor results seen in series of patients treatedwith radiation therapy alone for advanced disease mayreflect delivery of insufficient dose of radiotherapy forgross disease, often complicated by proximity of unresect-able disease to critical normal structures that have a lowertolerance dose to radiotherapy. In our series, 6 patientswere treated after biopsy alone, and 14 with macroscopicresidual disease, with results that did not seem inferior tothose treated postoperatively for microscopic residual dis-ease. Similarly, in the series by Pommier et al,22 localdisease control was maintained at both 5 and 10 yearsamong all 11 patients treated for skull base disease withhigh-dose proton therapy after biopsy only.

Our experience with high-dose proton therapy reflects ashort median follow-up time, and more mature follow-upis desired in this disease with a long natural history. Thedata for high-dose radiotherapy is retrospective and hasnot been rigorously compared against standard-dose radio-therapy, which limits the conclusions that can be drawnregarding the role of dose escalation in obtaining LC.

Late treatment toxicity in patients treated for their ini-tial disease course was modest with 1 patient developinggrade 3 toxicity, which is not surprising given the largevolume of mandible treated to a high dose, and 1 patientwith grade 4 vision loss, which was anticipated in treat-ment of an orbital tumor involving the optic nerve in apatient who otherwise would have required enucleation.Additional late toxicity was seen in the smaller subset ofpatients treated for recurrent disease, including 1 deathbecause of cerebrospinal fluid leak and meningitis in a

TABLE 2. Recent series of local control and overall survival for high-risk head and neck adenoid cystic carcinoma.

Author, y Cohort No. of patients Radiation dose Local control Survival

Balamucki, 20126 T4 Sx1 photon 22 69.6 Gy 5-y 43% 5-y 38%Balamucki, 20126 T4 photon alone 32 74.2 Gy 5-y 80% 5-y 51%Bianchi, 20087 T3/4 Sx1 photon 44 NR 5-y 74% 5-y 63%Rhee, 20068 T3/4 photon6 Sx 25 NR 5-y 58% NR for T3/4Munter, 200613 BOS (80%) photon6 Sx 25 66 Gy 2-y 75% 2-y 72%Chen, 200616 T4 Sx1 photon 26 64 Gy 10-y 37% 10-y 49%*Douglas, 199612 BOS neutron6 Sx 84 19.2 nGy 5-y 18% 5-y 31%

Douglas, 200817 BOS neutron1 GK6 Sx 34 19.2 nGy1 12 Gy GK 2-y 82% NRPommier, 200622 BOS proton6 Sx 23 75.9 Gy (RBE) 5-y 93% 5-y 77%Current study, 2014 BOS proton6 Sx 14 72 Gy (RBE) 2-y 93% 2-y 91%

Abbreviations: Sx, surgery; NR, not reported; BOS, base of skull; nGy, neutron Gy; GK, Gamma Knife radiosurgery; RBE, relative biological effectiveness.* Overall survival for T4 tumors treated with surgery with or without radiation. Unspecified number as a subset of 84 total patients. Cause-specific survival. Mixed salivary gland histologies, 85% adenoid cystic carcinoma.Note: Radiation dose is the median prescribed dose. Survival is overall survival unless otherwise indicated.

LINTON ET AL.


heavily pretreated patient. Given the small numbersinvolved, there was no statistically significant differencein toxicity among the treatment nave and previously irra-diated cohorts.

Our experience with 7 patients with recurrent disease,6 of whom had previously been irradiated, demonstrateda surprisingly good 2-year LC of 86%. In this subset ofpatients with particularly challenging disease compli-cated by one or more prior courses of radiation therapy,high-dose proton therapy was pursued in an effort toobtain control of skull base disease causing or threaten-ing significant clinical deficits. In 1 patient who hadundergone prior IMRT and 3 radiosurgery procedures,reirradiation may have contributed to subsequent cere-brospinal fluid leak, meningitis, and death, which under-scores the potential for significant toxicity withreirradiation. The safety and efficacy of reirradiation inthis complex cohort of patients cannot be evaluatedfrom this small subset. Comprehensive reviews of therole of reirradiation in head and neck cancer have beenpublished.23

Additional limitations of this study include its retro-spective nature, the relatively small number of patients,and the short median follow-up time, with longer follow-up necessary to assess the durability of local disease con-trol. Our study benefits from a relatively uniform treat-ment approach to these patients over the study period.The importance of careful treatment planning and thecomplexity of care involved in treatment of patients withskull base disease is underscored by the pattern of failureanalysis, which revealed that two thirds of the local fail-ures may have been prevented by changes in treatmentvolume.

Although optimizing LC and reducing toxicity of localtherapy remains an important goal of future research, theinsidious course of ACC and the high propensity for dis-tant metastatic disease24 underscore the importance ofcontinued efforts in evaluating the role of systemic thera-pies and developing novel agents with activity againstACC.

CONCLUSIONSIn our initial clinical experience among patients with

mostly advanced stage head and neck adenoid cystic car-cinoma, high-dose proton therapy provides encouragingpreliminary LC. Longer follow-up is needed to gauge thedurability of disease control and to monitor for late toxic-ities of therapy. Careful treatment planning is essential tooptimize outcomes.

REFERENCES1. Bhayani MK, Yener M, El-Naggar A, et al. Prognosis and risk factors for

early-stage adenoid cystic carcinoma of the major salivary glands. Cancer2012;118:28722878.

2. da Cruz Perez DE, de Abreu Alves F, Nobuko Nishimoto I, de AlmeidaOP, Kowalski LP. Prognostic factors in head and neck adenoid cystic carci-noma. Oral Oncol 2006;42:139146.

3. Seethala RR, Hunt JL, Baloch ZW, Livolsi VA, Leon Barnes E. Adenoidcystic carcinoma with high-grade transformation: a report of 11 cases and areview of the literature. Am J Surg Pathol 2007;31:16831694.

4. Ellington CL, Goodman M, Kono SA, et al. Adenoid cystic carcinoma of thehead and neck: incidence and survival trends based on 19732007 Surveil-lance, Epidemiology, and End Results data. Cancer 2012;118:44444451.

5. Papaspyrou G, Hoch S, Rinaldo A, et al. Chemotherapy and targeted ther-apy in adenoid cystic carcinoma of the head and neck: a review. Head Neck2011;33:905911.

6. Balamucki CJ, Amdur RJ, Werning JW, et al. Adenoid cystic carcinoma ofthe head and neck. Am J Otolaryngol 2012;33:510518.

FIGURE 6. Example of proton therapy for early-stage oral cavity adenoid cystic carcinoma. This patient underwent excision of a 1.2-cm adenoidcystic carcinoma from the right buccal mucosa with negative but close surgical margins, stage I, pT1 N0 M0. Perineural invasion was present onpathology. Proton therapy was used to treat the tumor bed to a dose of 66 Gy (relative biological effectiveness [RBE]) in 33 fractions. A custommouth guard was fabricated to completely shield the gingiva and oral cavity, allowing the distal end of the proton beam to terminate in the mouthguard rather than in normal tissues. The initial treatment volume included elective irradiation along the course of the mandibular nerve to foramenovale to 54 Gy (RBE), with proton beam geometry selected to spare the ipsilateral parotid. The patient developed grade 2 mucositis confined to thetreated portion of the ipsilateral buccal mucosa but developed no taste perversion, has no xerostomia, and has experienced no dental decay 17months after therapy. On the left is an axial slice from the treatment planning CT with color isodose lines. The tumor bed is encircled in red andthe ipsilateral parotid gland in green. On the right is the treatment plan dose volume histogram showing the sparing of the oral cavity and ipsilat-eral parotid gland. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com]



http://wileyonlinelibrary.com

7. Bianchi B, Copelli C, Cocchi R, Ferrari S, Pederneschi N, Sesenna E.Adenoid cystic carcinoma of intraoral minor salivary glands. Oral Oncol2008;44:10261031.

8. Rhee C, Won TB, Lee CH, et al. Adenoid cystic carcinoma of the sinonasaltract: treatment results. Laryngoscope 2006;116:982986.

9. Prokopakis EP, Snyderman CH, Hanna EY, Carrau RL, Johnson JT, DAmicoF. Risk factors for local recurrence of adenoid cystic carcinoma: the role ofpostoperative radiation therapy. Am J Otolaryngol 1999;20:281286.

10. Lupinetti AD, Roberts DB, Williams MD, et al. Sinonasal adenoid cysticcarcinoma: the M. D. Anderson Cancer Center experience. Cancer 2007;110:27262731.

11. Douglas JG, Laramore GE, AustinSeymour M, Koh W, Stelzer K, Griffin TW.Treatment of locally advanced adenoid cystic carcinoma of the head and neckwith neutron radiotherapy. Int J Radiat Oncol Biol Phys 2000;46:551557.

12. Douglas JG, Laramore GE, AustinSeymour M, et al. Neutron radiotherapyfor adenoid cystic carcinoma of minor salivary glands. Int J Radiat OncolBiol Phys 1996;36:8793.

13. Munter MW, SchulzErtner D, Hof H, et al. Inverse planned stereotacticintensity modulated radiotherapy (IMRT) in the treatment of incompletelyand completely resected adenoid cystic carcinomas of the head and neck:initial clinical results and toxicity of treatment. Radiat Oncol 2006;1:17.

14. Simpson JR, Thawley SE, Matsuba HM. Adenoid cystic salivary gland carci-noma: treatment with irradiation and surgery. Radiology 1984;151:509512.

15. Garden AS, Weber RS, Morrison WH, Ang KK, Peters LJ. The influenceof positive margins and nerve invasion in adenoid cystic carcinoma of thehead and neck treated with surgery and radiation. Int J Radiat Oncol BiolPhys 1995;32:619626.

16. Chen AM, Bucci MK, Weinberg V, et al. Adenoid cystic carcinoma of the headand neck treated by surgery with or without postoperative radiation therapy: prog-nostic features of recurrence. Int J Radiat Oncol Biol Phys 2006;66:152159.

17. Douglas JG, Goodkin R, Laramore GE. Gamma knife stereotactic radiosur-gery for salivary gland neoplasms with base of skull invasion followingneutron radiotherapy. Head Neck 2008;30:492496.

18. McDonald MW, Fitzek MM. Proton therapy. Curr Probl Cancer 2010;34:257296.

19. Anferov VA. Scan pattern optimization for uniform proton beam scanning.Med Phys 2009;36:35603567.

20. Farr JB, Mascia AE, Hsi WC, et al. Clinical characterization of a protonbeam continuous uniform scanning system with dose layer stacking. MedPhys 2008;35:49454954.

21. Allgower CE, Schreuder AN, Farr JB, Mascia AE. Experiences with anapplication of industrial robotics for accurate patient positioning in protonradiotherapy. Int J Med Robot 2007;3:7281.

22. Pommier P, Liebsch NJ, Deschler DG, et al. Proton beam radiation therapyfor skull base adenoid cystic carcinoma. Arch Otolaryngol Head Neck Surg2006;132:12421249.

23. McDonald MW, Lawson J, Garg MK, et al. ACR appropriateness criteriaretreatment of recurrent head and neck cancer after prior definitive radia-tion expert panel on radiation oncology-head and neck cancer. Int J RadiatOncol Biol Phys 2011;80:12921298.

24. Al-Mamgani A, van Rooij P, Sewnaik A, Tans L, Hardillo JA. Adenoidcystic carcinoma of parotid gland treated with surgery and radiotherapy:long-term outcomes, QoL assessment and review of the literature. OralOncol 2011;48:278283.

LINTON ET AL.


proton therpay for acc hed23573

Documents