Clinical Oncology xxx (2018) 1e7

Contents lists available at ScienceDirect

Clinical Oncology
journal homepage: www.clinicaloncologyonline.net

Original Article
Guidelines for Clinical Target Volume Definition for Perineural Spread
of Major Salivary Gland Cancersq
K. Armstrong *, J. Ward *, N.M. Hughes y, A. Mihai *, A. Blayney z, C. Mascott x, R. Kileen y,
J. Armstrong *
* Department of Radiation Oncology, Beacon Hospital, Dublin, Ireland
y
Department of Radiology, St. Vincent’s University Hospital, Dublin, Ireland
z
Department of Anatomy, Royal College of Surgeons, Ireland
x
Department of Surgery, Beacon Hospital, Dublin, Ireland

Received 3 December 2017; received in revised form 23 July 2018; accepted 23 July 2018

Abstract
Aims: Postoperative radiotherapy is the standard of care for resected major salivary cancers that are at risk of locoregional recurrence. Of the various histological
subtypes, perineural invasion is most common in adenoidcystic carcinomas of the three major salivary glands e parotid, submandibular and sublingual. The
clinical target volume (CTV) for these cases must include the relevant cranial nerve pathways at risk. A contouring atlas was devised for delineation of the CTV of
the nerves supplying the major salivary glands.
Materials and methods: Using standard anatomy texts and e-anatomy sources the nerves supplying the major salivary glands were identified. Subsequently the
pathways of the nerves were drawn on an archived patient’s planning computed tomography scan.
Results: The innervation of the major salivary glands has been identified and studied.
Both bone and soft tissue CTVs have been delineated. A full set of images and CTVs of all the relevant transverse computed tomography slices has been archived,
a number of which are printed in this article.
Conclusions: Variation in CTV delineation is a recognised problem in a variety of anatomic sites. Guidelines and atlases can standardise practice and may
improve the safety and efficacy of therapy. An atlas has been generated to guide clinicians in delineating the CTVs for perineural spread in major salivary gland
cancers.
Ó 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Keywords: Adenoidcystic; intensity-modulated radiation therapy; perineural; postoperative; salivary; target volumes

Introduction pair analyses and population-based studies have shown a

At present, postoperative radiotherapy (PORT) is the
standard of care for resected major salivary cancers that are
at significant risk of locoregional recurrence. Due to the
relative rarity of these cancers and the variety of histological
types occurring in three different glands, there are no pro-
spective trials attesting to its value. Retrospective analyses
consisting of comparative studies, case control matched
q Presented at the Third Annual Beacon Hospital Stereotactic Radiosurgery and
Stereotactic Body Radiotherapy Symposium, 17e18 October 2014 in Dublin, Ireland.
Author for correspondence: K. Armstrong, Department of Radiotherapy,
Beacon Hospital, Beacon Court, Sandyford, Dublin 18, Ireland. Tel: þ35-312-
936691.
E-mail address: kevinarms@gmail.com (K. Armstrong).
reduction in locoregional relapse and prolongation of
overall survival when PORT is used for selected cases [1,2].
Well-recognised indications for PORT include T3 or T4 pri-
maries, involved nodes and close or positive resection
margins. Radiotherapy may also be indicated for deep lobe
involvement, intraoperative tumour spillage and high-
grade tumours. Adenoidcystic cancers have an increased
risk of locoregional recurrence and PORT is generally indi-
cated, even in the absence of one of the above indications.
Small retrospective series have shown improved outcomes
when PORT is used for adenoidcystic salivary gland cancers
[3e5]. A large modern retrospective multi-institutional
analysis used the National Cancer Data Base of the USA to
analyse practice patterns and outcomes of PORT for

https://doi.org/10.1016/j.clon.2018.08.018
0936-6555/Ó 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Armstrong K, et al., Guidelines for Clinical Target Volume Definition for Perineural Spread of Major Salivary
Gland Cancers, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.08.018
2 K. Armstrong et al. / Clinical Oncology xxx (2018) 1e7
adenoidcystic carcinomas. PORT for salivary adenoidcystic
carcinoma was associated with improved survival even for
those with early-stage disease [6].
Of the various histologic subtypes, perineural invasion is
most common in adenoidcystic carcinomas of the three
major salivary glands e parotid, submandibular and sub-
lingual. The reported incidence of histologically docu-
mented perineural invasion varies substantially in the
reported literature. Garden et al. [7] from the MD Anderson
Hospital in Texas reported perineural invasion of a named
nerve in 28% (55/198) of patients with adenoidcystic cancer.
Vrielinck et al. [8] from the University Hospitals KUL,
Belgium reported an incidence of 52% in 37 patients. The
high variance in reported incidences may relate to the use of
different definitions for perineual invasion, i.e. gross versus
microscopic.
Because of the high incidence of perineural invasion it
has been standard practice to include the pathways of the
relevant nerves in the radiotherapy fields when delivering
PORT. However, the evidence justifying this is very limited.
A retrospective analysis of adenoidcystic cancers (major and
minor salivary gland origin) from the University of Califor-
nia San Francisco by Chen et al. [9] reported the rate of skull
base recurrence for patients treated with surgery alone.
There were 50 such patients, 25 of whom had histological
evidence of perineural invasion. There were two cases of
skull base recurrence, both in patients with histological
evidence of perineural invasion. Thus, the crude rate of skull
base recurrence was only 4% (2/50) for all adenoidcystic
patients treated with surgery alone and 8% (2/25) for those
with proven perineural invasion [9].
One retrospective series reported on 140 patients with
pathological evidence of perineural invasion at the time of
initial surgery. Allen et al. [10] from the University of Cali-
fornia, Davis, Cancer Center, Sacramento reported that 16/
140 (11%) had major (named) nerve involvement. Ninety-
four patients (67%) received PORT to the primary site and
the portal films of 65 of these patients were available for
review. PORT reduced the actuarial probability of skull base
recurrence from 15% to 5% (P ¼ 0.03). Crude rates of skull
base recurrence were 6% (2/35) and 10% (3/30), respectively,
for patients whose skull base were and were not encom-
passed in the irradiation field. The 5-year overall survival for
patients who experienced a skull base recurrence was 19%
compared with 91% for those who did not (P < 0.001). The
high mortality associated with skull base recurrence pro-
vides a rationale for inclusion of the relevant nerve path-
ways in the clinical target volume (CTV) of PORT.
The major series that examined the use of PORT for major
salivary gland cancers and for adenoidcystic cancers in
particular are mostly historical and used two-dimensional
radiation therapy. These older technologies did not
mandate a sophisticated three-dimensional approach to
target volume definition. With the advent of three-
dimensional conformal radiotherapy (3DCRT) and
intensity-modulated radiation therapy (IMRT), a more
precise three-dimensional approach to target volume defi-
nition is required. Furthermore, as radiotherapy technology
evolves it may become increasingly possible to spare organs
Please cite this article in press as: Armstrong K, et al., Guidelines for Clinical Target Volume Definition for Perineural Spread of Major Salivary
Gland Cancers, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.08.018
at risk that are in intimate proximity to target volumes.
Including the complex pathways of cranial nerves in a CTV
leads to challenges in treatment planning. The superior
depth dose characteristic of particle therapy may facilitate
adequate coverage of target volumes while sparing critical
normal structures [11,12].
If a decision is made to include pathways of potential
perineural spread, then the CTV for these cases must
include the relevant cranial nerve pathways at risk. Con-
touring atlases standardise CTV definition in other organs
and are now routinely used in clinical practice as a way of
standardising care [13,14]. Ko et al. [15] presented a con-
touring guide for head and neck cancers with perineural
invasion. However, they did not include the ninth nerve for
parotid cancers, nor the pathways for the sublingual gland.
Mourad et al. [16] described a contouring atlas adopted by
the Radiation Therapy Oncology Group for cranial nerves
IXeXII. This article identifies the individual nerves at risk
for each major salivary gland primary site and provides an
atlas that identifies the CTVs for perineural spread.
Materials and Methods
The nerve supply of the three major salivary glands were
identified by reference to standard anatomy textbooks
[17,18]. The pathways and interactions between the nerves
were described and depicted in diagrams in these text-
books. These nerve pathways were then identified on
internet-based ‘e-anatomy’ transverse computed tomogra-
phy images [19]. This work was carried out primarily by
authors KA and JW and checked by CM and AB. The depic-
tion of these pathways is often incomplete and not always
represented on transverse images. Transverse images are
most often used in the process of target volume definition.
Therefore, a complete set of transverse images was required
depicting the pathways. This work was carried out pri-
marily by authors NH and checked by RK. Blind validation
was not carried out, nor was inter-observer variability
checked.
To acquire these images, the nerve pathway CTVs were
delineated on an archived patient planning computed to-
mography scan. On the patient’s scan, bone CTVs were
assumed to be the entire dimension of the bone canal or
foramen. Unfortunately, the soft tissue pathways are
inherently less distinct and the actual nerves are invisible
on computed tomography scans. Consequently, the soft
tissue components of the nerve pathway CTVs are less
distinct and are delineated with less certainty. These soft
tissue CTVs are therefore more generous and less accurate.
Results
Parotid Gland
For the parotid gland, two important nerve pathways
were delineated. The facial nerve, which travels through the
substance of the parotid gland, and the glossopharyngeal
 K. Armstrong et al. / Clinical Oncology xxx (2018) 1e7 3

nerve, which provides the gland with its parasympathetic
innervation.
Facial Nerve
The facial nerve (Figure 1) enters the internal acoustic
meatus accompanied by the eighth cranial nerve. It passes
anteriorly towards the geniculate ganglion. At the ganglion
it turns abruptly posteriorly. It passes laterally and poste-
riorly to the utricle of the vestibule. At this level it is su-
perior and medial to the middle ear. It then moves
inferiorly, where it enters the facial canal within the mas-
toid. Within the canal it travels inferiorly, medially and
anteriorly, until it exits the stylomastoid foramen. Finally, it
travels laterally, through the gap between the mastoid
process and the mandible, to the parotid gland.
Glossopharyngeal Nerve
The glossopharyngeal nerve (ninth cranial nerve) pro-
vides parasympathetic innervation to the parotid gland,
where it has a secretomotor effect. The nerve leaves the
cranium via the jugular foramen. The tympanic nerve
branches off the glossopharyngeal nerve within the jugular
foramen. The tympanic nerve has mixed sensory and
parasympathetic fibres. It penetrates the temporal bone and
enters the cavity of the middle ear. Here, it forms the
tympanic plexus; a network of nerves that provide sensory
innervation to the middle ear, internal surface of the tym-
panic membrane and Eustachian tube (Figure 2).
Including the tympanic nerve pathway between the ju-
gular foramen and foramen ovale in the CTV may involve
delivering almost a full dose to the middle ear. This struc-
ture should therefore be delineated as an organ at risk and
have a doseevolume histogram generated as part of the
planning process. The lesser petrosal nerve carries the
parasympathetic fibres from the tympanic plexus to the otic
ganglion, where it synapses. The lesser petrosal nerve
passes through the foramen ovale (Figure 3).
The pathway was continued inferiorly for a very short
distance, to the otic ganglion, which is inferior to foramen
ovale. The parasympathetic fibres were then drawn passing
anterolaterally between the mandible and the mastoid,
where it runs within the auriculotemporal nerve, before
giving off parotid branches. There was no need to contour
the nerve beyond the parotid itself as it is already included
in the parotid CTV.
Submandibular and Sublingual Glands
Both the submandibular and the sublingual gland receive
parasympathetic innervation from the same source e the
facial nerve, via its chorda tympani nerve. The chorda
tympani branches off the facial nerve within the facial canal
and runs through the middle ear to exit the skull at the
petrotympanic fissure, which is close to the external portion
of the foramen ovale. Then it joins the lingual nerve before

Fig 1. (A) Internal auditory meatus where the seventh cranial nerve exits the brain (black arrows). This meatus is large as it contains both the
seventh nerve and the eighth nerve. (B) High resolution magnetic resonance imaging using constructive interference in steady state sequence
shows the origin of the facial nerve (white arrow) from the lateral margin of the inferior pons with the vestibulocochlear nerve (white
arrowhead) and traverses the cerebellopontine angle cistern to enter to the internal auditory meatus. (C) The facial nerve exits the internal
auditory canal and travels anteriorly towards the geniculate ganglion (black arrowhead). At the ganglion it turns abruptly posteriorly and travels
posterior and lateral to the utricle of the vestibule and enters the facial canal within the mastoid. (D) Path of cranial nerve VII within the
temporal bone where it passes through the ‘Z’-shaped facial canal (two white arrows).

Please cite this article in press as: Armstrong K, et al., Guidelines for Clinical Target Volume Definition for Perineural Spread of Major Salivary
Gland Cancers, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.08.018
4 K. Armstrong et al. / Clinical Oncology xxx (2018) 1e7

Fig 2. (A) The glossopharyngeal nerve arises from the lateral medulla and travels in the lateral cerebellomedullary cistern to exit the skull via the
pars nervosa in the jugular foramen (white arrows). (B) The pars nervosa (delineated by the asterisk) is the smaller anteromedial part of the
jugular foramen. Also depicted in this image is the foramen ovale (black arrows) through which the lesser petrosal nerve, a branch of the
glossopharyngeal nerve, travels to the otic ganglion. (C) The glossopharyngeal nerve then travels in between the internal carotid artery and the
internal jugular vein, descends anteromedial to the internal carotid artery initially and then traverses the artery becoming posteromedial to the
styloid process (white arrow) and the stylopharyngeus muscle. The asterisk in (C) denotes the expected location of the glossopharyngeal nerve
at the level of the styloid process. (D) This image uses soft tissue windowing at the level of the styloid process to show the relationship of the
internal carotid artery (black arrows), styloid process (white arrows) and the expected location of the glossopharyngeal nerve (asterisk).

synapsing in the submandibular ganglion and subsequently
innervating both glands.
Because the chorda tympani passes through the middle
ear, the contouring of the pathway for these two glands was
begun at the foramen ovale, which the mandibular nerve
passes through, giving off the lingual nerve (Figure 3).
Starting here allowed for variation as to the level at which
the lingual nerve combines with the chorda tympani, a
landmark that differs in each individual. It also allowed for
variation as to the level at which the mandibular nerve
gives off the lingual nerve. Furthermore, the submandibular
ganglion was assumed to be superior to the submandibular
gland on the planning system.
Submandibular Gland
Following its commencement at the foramen ovale
(Figure 3), the submandibular gland’s pathway was
continued beneath the lateral pterygoid muscle, then infe-
riorly between the medial pterygoid muscle and the ramus
of the mandible (Figure 4). Directly inferior to this point the
submandibular ganglion was marked. This was sufficient for
the submandibular gland’s pathway, as the gland itself lies
directly inferior to this point.
Sublingual Gland
The pathway for the sublingual gland includes the
pathway described above for the submandibular gland. It
also continues anteriorly beyond the submandibular gan-
glion as far as the site of sublingual salivary gland itself
(Figure 4).
Discussion
It is accepted practice that the relevant cranial nerve
pathways for adenoidcystic cancers of the major salivary
gland cancers should be included in the CTV for PORT. To
date there has been no published guideline or contouring
atlas to assist in this process. This article provides such a
contouring atlas. A limitation of this study is that magnetic
resonance imaging (MRI) images were not used to assist in
nerve delineation. MRI images would not have provided an
advantage in delineation of skull base anatomy. The
enhanced soft tissue definition of MRI images can identify
nerves themselves and the soft tissue structures that define
the pathways. The decision not to utilise MRI images was
deliberate as many practitioners and centres do not have
access to image fusion facilities and must rely on transverse
computed tomography images to identify anatomy.
Parotid Gland
The seventh cranial nerve has a complex anatomic rela-
tionship with the middle and inner ear. Pacholke et al. [20]

Please cite this article in press as: Armstrong K, et al., Guidelines for Clinical Target Volume Definition for Perineural Spread of Major Salivary
Gland Cancers, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.08.018
 K. Armstrong et al. / Clinical Oncology xxx (2018) 1e7 5

Fig 3. (A) The chorda tympani branches from the facial nerve and traverses the middle ear, which contains three ossicles (thick white arrow). It
enters the petrotympanic fissure (not shown in this image) on the anterior wall of the middle ear cavity and exits the fissure to join the lingual
nerve. Also shown in this image is the foramen ovale (big black triangle) through which the lesser petrosal nerve travels. The thick black arrows
show the apical (upper arrow) and basal (lower arrow) turns of the cochlea. The asterisk is in the vestibule and the lateral semicircular canal is
shown with the thin white arrow. (B, C) Axial and reformatted coronal sequences of the temporal bones. The chorda tympani branches from the
facial nerve and traverses the middle ear. It enters the petrotympanic fissure (black arrows) on the anterior wall of the middle ear cavity and
exits the fissure to join the lingual nerve. The mandibular condyle (*) and temporal bone (white arrow) are marked as reference points. (D)
Transverse image of facial nerve within the facial canal (white arrows). (E) Coronal image of facial nerve within the facial canal (white arrows),
the nerve travels inferiorly before exiting the stylomastoid foramen (black arrow). (F) Stylomastoid foramen (black arrow) where the seventh
nerve exits bone and runs anteriorly to enter the parotid gland.

described how to contour the middle and inner ear as an
organ at risk in radiotherapy planning. The nerve passes
above the middle ear, but in intimate relationship with its
superior component. Thus, even with IMRT and image-
guided radiotherapy it is likely that including the nerve
pathway in the CTV would deliver significant dose to the
middle ear. Emami et al. [21] reported that the TD5/5 for
chronic serous otitis was 55 Gy with conventional frac-
tionation. The distance between the pathway of the seventh
nerve and the cochlea may allow IMRT to deliver adequate
dose to the nerve CTV and spare the cochlea.
Marks et al. [22] reported that a mean dose of whole
organ radiation to the cochlea of 45 Gy in conventional
fractionation resulted in a 30% rate of sensory neural hear-
ing loss. Lee et al. [23] analysed the outcome in 211 patients
with head and neck cancers treated with radiotherapy. They
recommended that the mean dose to the cochlea should be
less than 32 Gy to maintain the risk of grade 2 or higher
tinnitus at less than 20%.
Garden et al. [24] from Texas noted that hearing loss
occurred in 7% (12/166) of patients treated with PORT for
parotid cancers. A variety of doses and non-conformal
techniques were used. The authors did not provide data
about the mechanisms of hearing loss or the dose
relationships.
Lamers-Kuijper et al. [25] from the Netherlands Cancer
Institute analysed the dose to middle and inner ear struc-
tures comparing 3DCRT to IMRT for parotid cancers. The
mean volume of the middle ear receiving 50 Gy or higher
was reduced from 66% with 3DCRT to 33% with IMRT. The
mean volume of the inner ear receiving 50 Gy or higher was
reduced from 14.7% with 3DCRT to 1.4% with IMRT. How-
ever, these 20 cases were adenocarcinomas and did not
have target volumes specifically designed to include po-
tential perineural spread. The authors showed that prox-
imity of the planning target volume to the ear structures
restricted the benefit of IMRT.
The COSTAR trial randomised 110 patients receiving
PORT for parotid cancer to receive either standard 3DCRT or
cochlea-sparing IMRT [26]. Radiotherapy dose to the target
volume was in the range of 60e65 Gy. The mean dose to the
ipsilateral cochlea was 56.2 Gy for 3DCRT and 35.7 Gy for
IMRT (P < 0.001). The primary end point was a measure of
ipsilateral cochlea hearing loss at 12 months, which
occurred in 14/35 (40%) 3DCRT patients and 11/31 (36%)
IMRT patients (P ¼ 0.80). IMRT reduced the radiation dose
below the accepted tolerance of the cochlea, but this did not
lead to a statistically significant reduction in the proportion
of patients with hearing loss in the ipsilateral ear at 12
months after radiotherapy. Pacholke et al. [20] described

Please cite this article in press as: Armstrong K, et al., Guidelines for Clinical Target Volume Definition for Perineural Spread of Major Salivary
Gland Cancers, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.08.018
6 K. Armstrong et al. / Clinical Oncology xxx (2018) 1e7

Fig 4. (A) L is the lateral ptyergoid, M the medial pterygoid, the black arrow is the mandibular ramus and the asterisk is the expected location of
the lingual nerve. (B) The path of the parasympathetic supply of the lingual nerve is shown, with the fibres passing anteromedial from the
submandibular ganglion (just cephalad to submandibular gland) to the sublingual gland. H is the hyoglossus e the submandibular ganglion sits
above this muscle. M is the mylohyoid muscle. The lingual nerve passes deep to the mylohyoid and lateral to hyoglossus.

how to delineate the cochlea on radiation planning
computed tomography scans. Nevertheless, accurate
delineation of the cochlea is difficult due to its tiny size. The
organ at risk may be expanded into a planning organ risk
volume (PORV) by expanding the volume by up to 0.3 cm.
Subsequently the IMRT planning optimisation process seeks
to minimise the dose to this PORV. It is possible that in some
patients on the IMRT arm of the COSTAR trial, which has
only been reported as an abstract, the dose recorded for the
cochlea may not be an accurate reflection of the dose
received to the actual organ. When the organ at risk is tiny,
then the ratio of PORV to organ at risk becomes very large. It
is speculated that the actual cochlea may only have been a
very small portion of the PORV. Furthermore, daily variation
in set-up accuracy could also contribute to a variation be-
tween calculated and delivered dose to the cochlea.
Nevertheless, the finding of the COSTAR trial raises con-
cerns about the toxicity that may result when the CTV in-
cludes nerve pathways in proximity to the cochlea.
The standard dose of conventionally fractionated PORT
in parotid cases is 60e66 Gy depending on margin status.
It is accepted that these doses are necessary for the portion
of the CTV related to the primary tumour. Clinicians must
decide whether to include within the CTV the parts of the
cranial nerve pathways (seventh or ninth) which are in
intimate proximity to the middle and inner ear. Each
possible outcome is considered. Inclusion of the full
pathway may be justifiable if there is extensive perineural
invasion in the primary resection specimen or if named
branches of the nerve are microscopically involved. At the
other extreme, if there is no evidence of perineural inva-
sion in the primary resection specimen it may be more
difficult to justify the risk of toxicity. In either scenario, the
patient should be informed of the dilemma. A compromise
is also possible whereby the dose to the primary tumour
bed is maintained at the range of 60e66 Gy with con-
ventional fractionation but the dose where the perineural
CTV and resultant planning target volume overlap with the
middle and inner ear is restricted to approximately 50 Gy
delivered by conventional fractionation. This is a specula-
tive suggestion based on the fact that the relevant portion
of CTV has not been operated on and may not require
60e66 Gy for control of microscopic disease. Ko et al. [15]
made similar recommendations to adapt the dose, taking
account of both the degree of risk in a portion of the CTV
and the tolerance of adjacent organs at risk. Bakst et al.
[27], using a murine model of pancreas cancer, proposed
that radiation may impair perineural invasion by inter-
ruption of paracrine mechanisms underlying perineural
invasion. In their laboratory model they showed such ef-
fects at low doses. However, there are no human labora-
tory or clinical data directly addressing the use of doses in
the range of 50 Gy for control of microscopic perineural
invasion [27].
Submandibular and Sublingual Glands
The chorda tympani (a branch of the facial nerve) carries
taste fibres from the anterior two thirds of the tongue and
parasympathetic fibres to the submandibular ganglion. The
chorda tympani nerve was not included in the CTV for the
submandibular gland as it comes in close contact with the
middle ear, potentially disqualifying it from being safely
irradiated. This seems reasonable in a case where the
operative histology provides no actual evidence of peri-
neural invasion. In the presence of perineural invasion,
particularly if this is other than minimal, the chorda
tympani may be included in the CTV. Ko et al. [15] proposed
including the pathway of the XIIth (hypoglossal) nerve for
submandibular glands on the basis of the nerve’s proximity
to the gland. However, as the XIIth nerve does not innervate
the submandibular gland we have not included its pathway
in the CTV.

Please cite this article in press as: Armstrong K, et al., Guidelines for Clinical Target Volume Definition for Perineural Spread of Major Salivary
Gland Cancers, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.08.018
 K. Armstrong et al. / Clinical Oncology xxx (2018) 1e7
Conclusion
A contouring atlas was devised for inclusion of nerve
pathways in the PORT of adenoidcystic cancers of major
salivary gland cancers.
Acknowledgement
This research was partly funded by a patient donation.
References
[1] Chong L, Armstrong JG. Salivary gland cancer. In: Leibel,
Phillips, editors. A textbook of radiation oncology, 3rd ed.
Philadelphia: Elsevier Saunders; 2010. p. 666e699.
[2] Armstrong JG, Harrison LB, Spiro RH, Fass DE, Strong EW,
Fuks Z, et al. Malignant tumors of major salivary gland origin:
a matched pair analysis of the role of combined surgery and
postoperative radiotherapy. Arch Otolaryngol Head Neck Surg
1990;116(3):290e293.
[3] Fu K, Leibel S, Levine M, Friedlander LM, Boles R, Phillips TL.
Carcinoma of the major and minor salivary glands: analysis of
treatment results and sites and causes of failures. Cancer
1977;40:2882e2890.
[4] Matsuba H, Spector G, Thawley S, Simpson JR, Mauney M,
Pikul FJ. Adenoid cystic salivary gland carcinoma. A histo-
pathologic review of treatment failure patterns. Cancer 1986;
57:519e524.
[5] Miglianico L, Eschwege F, Marandas P, Wibault P. Cervico-
facial adenoid cystic carcinoma: study of 102 cases. Influence
of radiation therapy. Int J Radiat Oncol Biol Phys 1987;13:
673e678.
[6] Lee A, Givi B, Osborn VW, Schwartz D, Schreiber D. Patterns of
care and survival of adjuvant radiation for major salivary
adenoid cystic carcinoma. Laryngoscope 2017;127:2057e2062.
[7] Garden AS, Weber RS, Morrison WH, Ang KK, Peters LJ. The
influence of positive margins and nerve invasion in adenoid
cystic carcinoma of the head and neck treated with surgery
and radiation. Int J Radiat Oncol Biol Phys 1995;32:619e626.
[8] Vrielinck LJ, Ostyn F, Van Damme B, van den Bogaert W,
Fossion E, University Hospital KUL Belgium. The significance
of perineural spread in adenoid cystic carcinoma of the major
and minor salivary glands. Int J Oral Maxillofac Surg 1988;17:
190e193.
[9] Chen AM, Garcia J, Granchi P, Bucci MK, Lee NY. Base of skull
recurrences after treatment of salivary gland cancer with
perineural invasion reduced by postoperative radiotherapy.
Clin Otolaryngol 2009:539e545.
[10] Chen AM, Bucci MK, Weinberg V, Garcia J, Quivey JM,
Schechter NR, et al. Adenoid cystic carcinoma of the head and
neck treated by surgery with or without postoperative radi-
ation therapy: prognostic features of recurrence. Int J Radiat
Oncol Biol Phys 2006;66:152e159.
[11] Grant SR, Grosshans DR, Bilton SD, Garcia JA, Amin M,
Chambers MS, et al. Proton versus conventional radiotherapy
for pediatric salivary gland tumors: acute toxicity and dosi-
metric characteristics. Radiother Oncol 2015;116(2):309e315.
[12] Jensen AD, Poulakis M, Nikoghosyan AV, Welzel T, Uhl T,
Federspil PA, et al. High-LET radiotherapy for adenoid cystic
Please cite this article in press as: Armstrong K, et al., Guidelines for Clinical Target Volume Definition for Perineural Spread of Major Salivary
Gland Cancers, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.08.018
7
carcinoma of the head and neck: 15 years’ experience with
raster-scanned carbon ion therapy. Radiother Oncol 2016;118:
272e280.
goire V, Ang K, Budach W, Grau C, Hamoir M, Langendijk,
[13] Gre
et al. Delineation of the neck node levels for head and neck
tumors: a 2013 update. DAHANCA, EORTC, HKNPCSG, NCIC
CTG, NCRI, RTOG, TROG consensus guidelines. Radiother Oncol
2014;110:172e181.
[14] Gregoire V, Evans M, Le QT, Bourhis J, Budach V, Chen A, et al.
Delineation of the primary tumour clinical target volumes
(CTV-P) in laryngeal, hypopharyngeal, oropharyngeal and oral
cavity squamous cell carcinoma: AIRO, CACA, DAHANCA,
EORTC, GEORCC, GORTEC, HKNPCSG, HNCIG, IAG-KHT,
LPRHHT, NCIC CTG, NCRI, NRG Oncology, PHNS, SBRT, SOM-
ERA, SRO, SSHNO, TROG consensus guidelines. pii: S0167-
8140(17)32656-32657. Radiother Oncol 2017. https://doi.org/
10.1016/j.radonc.2017.10.016 [Epub ahead of print].
[15] Ko H, Gupta V, Mourad W, Kenneth S, Hu K, Harrison L, et al.
A contouring guide for head and neck cancers with perineural
invasion. Pract Radiat Oncol 2014;4:e247ee258.
[16] Mourad W, Hu K, Harrison L. Cranial nerves IXeXII contouring
atlas for head and neck cancer. Available at: https://www.
rtog.org/CoreLab/ContouringAtlases/HNAtlases.aspx.
[17] Netter F. Atlas of human anatomy, 2nd ed. East Hanover New
Jersey: Novartis; 1997.
[18] Last RJ. Anatomy. Regional and applied, 6th ed. Edinburgh:
Churchill Livingstone; 1978.
[19] https://radiopaedia.org].
[20] Pacholke HD, Amdur RJ, Schmalfuss IM, Louis D,
Mendenhall WM. Contouring the middle and inner ear on
radiotherapy planning scans. Am J Clin Oncol 2005;28(2):
143e147.
[21] Emami B, Lyman J, Brown A, Coia L, Goitein M,
Munzenrider JE, et al. Tolerance of normal tissue to thera-
peutic radiation. Int J Radiat Oncol Biol Phys 1991;21:109e122.
[22] Marks LB, Yorke ED, Jackson A, Ten Haken RK, Constine LS,
Eisbruch A, et al. Use of normal tissue complication proba-
bility models in the clinic. Int J Radiat Oncol Biol Phys 2010;
3(suppl):s10es19.
[23] Lee TF, Yeh SA, Chao PJ, Chang L, Chiu CL, Ting HM, et al.
Normal tissue complication probability modeling for cochlea
constraints to avoid causing tinnitus after head-and-neck
intensity-modulated radiation therapy. Radiat Oncol 2015;
10:194.
[24] Garden AS, el-Naggar AK, Morrison WH, Callender DL,
Ang KK, Peters LJ. Postoperative radiotherapy for malignant
tumors of the parotid gland. Int J Radiat Oncol Biol Phys 1997;
37:79e85.
[25] Lamers-Kuijper E, Schwarz M, Rasch C, Mijnheer B. Intensity-
modulated vs. conformal radiotherapy of parotid gland tu-
mours: potential impact on hearing loss. Med Dosim 2007;32:
237e245.
[26] Nutting C, Morden J, Beasley M, Bhide S, Emson M, Evans M,
et al. PO-093 COSTAR trial results: 3-D conformal radio-
therapy vs cochlea-sparing IMRT in parotid cancer patients.
Radiother Oncol 2017;122(suppl. 1):44e45.
[27] Bakst R, Lee N, He S, Chernichenko N, Chen C, Carl Le H et al.
Radiation impairs perineural invasion by modulating the
nerve microenvironment. Available at: http://journals.plos.
org/plosone/article?id¼10.1371/journal.pone.0039925.