Volumetric modulated arc therapy planning to reduce ipsilateral submandibular gland and

1B nodal radiation for oropharyngeal head and neck cancer

Luke VanLanen, BS; Nick Nguyen, BS, R.T.(T); Travis Kilmer, BS; Nishele Lenards PhD,
CMD, R.T.(R)(T), FAAMD; Ashley Hunzeker, MS, CMD; Karen Lang, MS, R.T.(T), CMD,
David Reineke PhD
Medical Dosimetry Program at the University of Wisconsin, La Crosse, WI
Introduction
The incidence of oropharynx cancer is on the rise with a predicted 53,000 new diagnoses
in 2019, according to the American Cancer Society.1 The submandibular gland (SMG) is a vital
gland in the human body that secretes saliva. It has been shown to produce 90% of unstimulated
saliva and, when spared, improve the patient’s quality of life.2 The improved quality of life can
be especially observed at night when submandibular secretion is strongly correlated to nighttime
xerostomia.3 Patients with xerostomia have a hypofunction of salivary output resulting in sore
throat, altered taste, dental decay, changes in voice quality, and impaired chewing and
swallowing function.4 Side effects can result in reduced nutritional intake and weight loss;
ultimately affecting general health and quality of life.4
Typical head and neck cancer treatments result in the SMG and level 1B nodes receiving
high levels of radiation. Historically, level 1B nodes have been included in the target volumes.
However, various studies have demonstrated that the percentage of 1B nodal involvement is rare.
Lee et al5 showed that for 102 patients with surgically treated oropharyngeal squamous cell
cancer observed from 2010 to 2016, only 4.3% had 1B nodal involvement. Ho et al6 reported
similar findings that only 2.7% of level 1B nodes were affected amongst oropharyngeal cancer
patients whereas Sanguineti et al7 stated that there is < 5% of the level 1B nodes being
impacted. Lee et al5 demonstrated that the incidence of 1B nodal involvement is
rare, while oropharynx cancer rarely travels into the level 1B nodes. The seldom occurrence
of 1B nodal involvement means that the SMG and 1B nodes are considered low risk for
oropharynx cancer. However, the SMG is also included in the planning target
volume (PTV) because of its location to the 1B nodes but does not have any nodal drainage
function. The treatment of PTV volumes typically includes irradiating the ipsilateral SMG and
level 1B nodes. Irradiating the gland and nodes leads to an increase in xerostomia
and a potential decrease in the quality of life for a patient.8 In addition, by adhering to the dose
constraints for the ipsilateral level 1B nodes and SMG, the treatments proved beneficial to the
patients by decreasing the risk of xerostomia and improving quality of life.8
Using volumetric modulated arc therapy (VMAT) to treat the patients has proven to be an
advantage to maintain lower doses to normal tissues. Dai et al9 concluded that there
was an improvement in delivering lower doses to the mandible and oral cavity while also having
improved target dose homogeneity when compared to static-field intensity-modulated radiation
therapy (IMRT). Other advantages to using VMAT over static-field IMRT included having a
shorter delivery time and a smaller number of monitor units.9
The goal of this study was to examine the dose that the SMG/1B receives by excluding it
from limited research showing the planning results of attempting to avoid the ipsilateral SMG
and level 1B nodes. Researchers hypothesized that by excluding the SMG and level 1B nodes
from the low risk PTV, dose to the SMG and level 1B nodes can significantly decrease. To
research the possibly of lowering the dose to the SMG/1B nodes, retrospective planning was
done with the SMG/1B nodes subtracted from the low-risk CTV. The overall goal of the
retrospective plan is to evaluate the dose to the SMG/1B nodes by planning to get coverage for
the re-structured low-risk PTV Retrospective planning was done on cases using a reconstructed
PTV to spare the SMG and level 1B nodes that were created by a radiation oncologist.
Methods and Materials
Selection and Setup
The selection criteria for this retrospective study included patients with a history
of oropharyngeal squamous cell carcinoma, base of tongue (BOT), or tonsil cancer and had
received prior radiation therapy to the oropharyngeal head and neck area. A total of 5 patients
were collected from a single clinical institution. Out of all 5 patients, 4 had unilateral disease
and 1 of the patients was treated bilaterally. For this study, each side of the neck was evaluated
exclusively for the bilateral case, totaling 6 sites. The exclusion criteria for this study included
patients not requiring surgical removal of SMGs, considered high risk due to metastatic disease
to the lymph nodes, and/or gross tumor extending into the SMG.
All patients underwent a computed tomography (CT) simulation for radiation treatment
planning. The patient was positioned head-first supine with a custom 5-point thermoplastic mask
and neck support on an indexed s-frame. The 5 CT scans were performed using contrast, for
improved nodal delineation, with a slice thickness of 3mm.
Contouring
Target volumes included the gross tumor volume (GTV), clinical target volume (CTV),
and PTV and were delineated by the radiation oncologist. The original PTV was created from a 3
mm expansion of the CTV which included the ipsilateral SMG and level 1B nodes. Previously
treated head and neck tumors were re-planned with attempts to spare ipsilateral SMG and level
1B nodes to account for the lack of disease involvement. The new CTV, in an attempt to spare
the ipsilateral SMG and level 1B nodes for lack of disease involvement, excluded these
structures. The ipsilateral SMG and level 1B nodes were contoured by the physician and
excluded from the CTV. The new CTV was now abutting these structures. The new
PTV was then recontoured by a radiation oncologist for retrospective planning by placing a 3
mm expansion on the CTV. The PTV also had 5 mm subtracted from the skin contour. The only
change to the treatment volume in the retrospective plans was the absence of the SMG and
level 1B nodes. Normal structures contoured in each plan included; skin, mandible, parotid
gland, spinal cord, spinal cord protective ring volume (PRV), brain, brain stem, brain stem PRV,
optic nerve, optic chiasm, orbit, and lens. The PRV expansion for the spinal cord was 5 mm,
and the PRV expansion for the brain stem was 3 mm.
Treatment Planning
Patients in this study were treated with doses ranging from 66-70 Gy. Four of the patients
were treated to 70 Gy and only one patient was treated to 66 Gy. The lower dose of 66 Gy is due
to early staged disease which warrants a lower prescription dose. All treatment regimens were
given prescriptions that matched the standard of care and matched
the original prescriptions. Each plan was normalized using a volume-based
normalization technique. The normalization used was 95% of the PTV volume receiving the
prescription dose.
The treatment planning systems (TPS) used for the planning of all five patients was
Monaco (Elekta). Treatment plans were generated using a VMAT technique consisting of 2-3
arcs and 360 degrees of rotation. The plans were re-created all using 6 MV energy. The
retrospective re-plans aimed to maintain the same optimization goals for target coverage
while improved organs at risk (OAR) sparing were optimized, with the goal of receiving the
equal or less dose than the original plan. A primary focus remained on reproducing adequate
OAR sparing so the re-plans were equally safe or safer because each structure receives equal or
less dose than the original plan. They were also equally effective at treating the tumor mass since
the only difference in the plans is the avoidance of the ipsilateral SMG/1B nodes, which were
found to be disease free.
To evaluate the different planning techniques, a DVH for each technique was generated
to collect data for dosimetric analysis. The primary focus of comparison is dose to the SMG and
dose to the level 1B nodes. Planning target coverage and dose to OAR structures are also utilized
to ensure the retrospective plan is within 2% of the original plan. Evaluations of individual plans
were kept consistent for each plan. The PTV coverage was analyzed by the percentage of the
target volume that is covered by the prescription dose or greater. The minimum target
coverage required for each plan was the volume receiving 95% of the prescription dose
(V95%) and each plan needed to, at the least, meet 95% of the volume covered by 100% of the
prescription dose. The researchers also compared the volume receiving 100% of the prescription
dose (V100%). The main goal was to match original prescribed coverage by
the radiation oncologist as well as match the PTV coverage from the original plan. The dose to
the ipsilateral SMG was the next priority of comparison. While maintaining coverage to the
PTV, sparing of the SMG is considered successful if it has a mean dose of less than
39 Gy. Ensuring the SMG is spared while achieving PTV coverage is the overall goal of the
study along with sparing the level 1B nodes. The dose at the level 1B nodes was then compared
to the original plan. The level 1B nodes was considered successfully spared if it receives a mean
dose of less than 39 Gy.15 The OAR structures were then compared via the two treatment plans
for each patient that included the following: mandible, parotid gland, spinal cord, spinal cord
PRV, brain, brain stem, brain stem PRV, optic nerve, optic chiasm, orbit, and lens. Maximum
and mean doses were to be recorded for the contoured structures at risk.
Results
Patient Plan Characteristics
Patient 1 had 3 PTV volumes being treated to 3 dose levels: 70 Gy, 63 Gy, and 56
Gy. The doses to the volumes included 70 Gy, 63 Gy, and 56 Gy. The normalization for the
PTV’s was kept the same such that 95% of the volumes were covered by the prescription dose.
The mean dose to the ipsilateral SMG and level 1B nodes was 65.87 Gy and 66.48 Gy for the
original plan, respectively. The retrospective plan produced a mean dose of 61.62 Gy and 58.83
Gy to the ipsilateral SMG and level 1B nodes, respectively. This resulted in a –6.45% difference
for the ipsilateral SMG and a –11.51% difference for the ipsilateral level 1B nodes.
Patient 2 had 3 PTV volumes being treated to three different doses which included 70 Gy,
63 Gy, and 59.4 Gy. This patient had bilateral disease, therefore, the SMG and level 1B nodes on
both sides of the patient were attempted to be spared. The normalization for the PTV’s was kept
the same such that 95% of the volumes were covered by the prescription dose. For the left side of
the patient, the mean dose to the SMG and level 1B nodes was 63.30 Gy and 62.28 Gy for the
original plan, respectively. On the same side, the retrospective plan produced a mean dose of
62.39 Gy and 58.31 Gy to the ipsilateral SMG and level 1B nodes, respectively. This resulted in
a –1.44% difference for the SMG and a –6.37% difference for the level 1B nodes. For the right
side of the patient, the mean dose to the SMG and level 1B nodes was 63.61 Gy and 63.15 Gy
for the original plan, respectively. On the same side, the retrospective plan produced a mean dose
of 58.31 Gy and 53.66 Gy to the ipsilateral SMG and level 1B nodes, respectively. This resulted
in a –8.33% difference for the SMG and a –15.03% difference for the level 1B nodes.
Patient 3 had 2 PTV volumes being treated to 2 different doses. The doses to the volumes
included 63 Gy and 54 Gy. The normalization for the PTV’s was kept the same such that 95% of
the volumes were covered by the prescription dose. The mean dose to the ipsilateral SMG and
level 1B nodes was 54.78 Gy and 43.51 Gy for the original plan, respectively. The retrospective
plan produced a mean dose of 37.50 Gy and 27.71 Gy to the ipsilateral SMG and level 1B nodes,
respectively. This resulted in a –31.54% difference for the ipsilateral SMG and a –36.31%
difference for the ipsilateral level 1B nodes.
Patient 4 had three PTV volumes being treated to three different doses. The doses to the
volumes included 70 Gy, 63 Gy, and 56 Gy. The normalization for the PTV’s was kept the same
such that 95% of the volumes were covered by the prescription dose. The mean dose to the
ipsilateral SMG and level 1B nodes was 67.87 Gy and 66.64 Gy for the original plan,
respectively. The retrospective plan produced a mean dose of 41.88 Gy and 42.03 Gy to the
ipsilateral SMG and level 1B nodes, respectively. This resulted in a –38.30% difference for the
ipsilateral SMG and a –36.93% difference for the ipsilateral level 1B nodes.
Patient 5 had three PTV volumes being treated to 3 different dose levels. The doses to the
volumes included 70 Gy, 63 Gy, and 56 Gy. The normalization for the PTVs was kept the same
such that 95% of the volumes were covered by the prescription dose. The mean dose to the
ipsilateral SMG and level 1B nodes was 71.73 Gy and 68.18 Gy for the original plan,
respectively. The retrospective plan produced a mean dose of 70.15 Gy and 61.80 Gy to the
ipsilateral SMG and level 1B nodes, respectively. This resulted in a –2.20% difference for the
ipsilateral SMG and a –9.36% difference for the ipsilateral level 1B nodes.
Statistical Methods
Wilcoxon signed rank tests were used rather than paired t-tests or a doubly multivariate
repeated measures analysis due to very small sample sizes, outliers, and missing values observed
in some of the samples. The sample size contains two measurements for patient 2. Statistical
analysis was conducted using R (R Core Team, 2019).16
A level of significance of 5% was used for each individual hypothesis tests and no correction
was applied to control the Type 1 error rate for multiple testing because the Type 2 error rate
already elevated due to the small sample sizes and lowering the Type 1 error rate would only
make the Type 2 error rate larger.
Statistical Results
The results from the new plans created in this study demonstrated no significant
difference in population distributions between the dose received by the ipsilateral submandibular
gland for the new target size and the original target size (P = 0.132) at the 5% level of
significance. However, the responses for level 1B nodes (P = 0.026) were significantly different,
with the distribution of the new target size being shifted to the left of the distribution for the old
target size. That is, generally lower values were seen with the new target size. Side-by-side
boxplots for the mean dose to the ipsilateral SMG and level 1B nodes for each target size are
given by Figure 3 and Figure 4.
Discussion
The results of this study showed that a CTV drawn to avoid the ipsilateral SMG and
level 1B nodes can replicate a plan that includes those structures in the target volume. The results
demonstrated adequate coverage to a PTV drawn to spare the SMG/level 1B nodes can be
achieved. However, comparing PTV coverage from the original plan to the new plan is not
equivalent because they are not the same volume. We were still able to demonstrate that PTV
coverage was either equal or better in the new plan. This means the area being treated to
prescription dose is the same as the original target volume with the only difference in dose being
the area including the ipsilateral SMG and level 1B node. The SMG and level 1B nodes have not
been shown to include disease5. This means the volume of disease treated in both the original and
retrospective plan should be the same. Utilizing mean dose to measure the dose to the target
volume allows for a great comparison. In this study only one outlier occurred. In patient 3 the
mean dose for PTV 54 changed by 7%. Other than this one volume, the PTV mean doses never
varied by greater than 5% demonstrating that the retrospective plans were created very similar to
the original plan.
Although PTV coverage was met, the mean doses to the ipsilateral SMG and
level 1B nodes were not met. The main goal of this study was to conclude if the retrospective
plan would result in a reduction of dose to the SMG and level 1B nodes. Researchers suggest that
a mean dose of less than 39 Gy spares some salivary function in the SMG.3 Doses to the
ipsilateral SMG exceeded 60 Gy for four plans, with a total of five sites, and only met the mean
dose of 39 Gy for one plan. This means that only 20% of the plans were able to spare the SMG,
and 60% of the plans exceeded a mean of 60 Gy. A reason for this occurring is that for patient 3,
the prescription dose was lower than the rest of the patients. This could have led to the result of
the mean dose to the ipsilateral SMG remaining under 39 Gy while the rest of the patients with
higher prescription doses failed. In many instances the ipsilateral SMG and
level 1B nodes cannot be spared for a functionable plan to be created. Although, sparing the
SMG and level 1B nodes may be an option a radiation oncologist would explore, it may not be
possible to achieve that plan. An aspect that may be noted is that this study only investigated five
patients. It may be possible that expanding the number of patients involved in this study would
yield more favorable results. Sparing the SMG and level 1B would help increase the patient's
quality of life and multiple studies have shown the potential benefit of sparing the salivary
function of a patient.1-3,5, 9 The researchers in this study demonstrated that a plan with
adequate PTV coverage that spares the ipsilateral SMG/level 1B nodes is unlikely to be
achieved.
Conclusion
The purpose of this study was to conclude if the dose to the SMG and level 1B nodes
could be reduced while maintaining coverage to the PTV for patients with oropharyngeal
cancer. The goal was to have a treatment that irradiated the tumor volume but spared the
SMG and level 1B nodal function. When considering mean dose to the ipsilateral SMG and level
1B nodes, the findings of this study did not. This study does not support the idea that a plan can
maintain coverage to the PTV and spare the SMG/1B nodes for patients with oropharyngeal
cancer. The plan does support the fact that a CTV drawn in this fashion still allows dose to the
remaining OAR to meet planning constraints and varied very slightly from the original plans. An
argument could be made that an attempt to spare the ipsilateral SMG and level 1B nodes should
be made even if a plan sparing those structures is likely impossible. If sparing is not possible
then the treatment is identical to the typical dose distribution currently being used to treat this
tumor volume. However, if dose reduction and sparing does occur then the outcome for the
patient would be the same while quality of life significantly better. Finally, we want to conclude
this study by stating that this study should be expanded to more retrospective planning. One of
the limitations of this study was the number of patients evaluated for re-planning. To further
research in this area of treatment our study should be replicated but with a larger number of
patients. A larger sample size would give a better idea of the types of patients this type
of planning technique could benefit.
 References

1. Key Statistics for Oral Cavity and Oropharyngeal Cancers. American Cancer Society.
https://www.cancer.org/cancer/oral-cavity-and-oropharyngeal-cancer/about/key-
statistics.html. Accessed June 24, 2019. 
2. Gensheimer M, Liao J, Laramore G, Parvathaneni U. Safety of submandibular gland-sparing
intensity modulated radiation therapy for head-and-neck cancer. Int J Radiat Oncol Biol
Phys. 2013;87(2):S59. http://dx.doi.org/10.1016/j.ijrobp.2013.06.153 
3. Terhaard C, Dijkema T, Braam P, Raaijmakers C, Roesink J. Importance of sparing
submandibular gland function to improve patient-reported xerostomia. Int J Radiat Oncol
Biol Phys. 2010;78(3):S80. http://dx.doi.org/10.1016/j.ijrobp.2010.07.218 
4. Pinna R, Campus G, Cumbo E, Mura I, Milia E. Xerostomia induced by radiotherapy: An
overview of the physiopathology, clinical evidence, and management of the oral
damage. Ther Clin Risk Manag. 2015(1):171-188. http://dx.doi.org/10.2147/tcrm.s70652 
5. Lee NC, Kelly JR, Park HS, Yarbrough WG, Burtness BA, Husain, ZA. (2017). The risk of
level IB nodal involvement in oropharynx cancer: Guidance for submandibular gland sparing
irradiation. Pract Radiat Oncol. 2017;7(5): e317-e321.
http://dx.doi.org/10.1016/j.prro.2017.02.004 
6. Ho FCH, Tham IWK, Earnest A, Lee KM, Lu JJ. Patterns of regional lymph node metastasis
of nasopharyngeal carcinoma: A meta-analysis of clinical evidence. BMC Cancer.
2012;12(1):98. http://dx.doi.org/10.1186/1471-2407-12-98 
7. Sanguineti G, Califano J, Zhou J, Stafford E, Koch W, Tufano R, Gourin C, Sormani M,
Marur S, Forastiere A. Defining the risk of involvement for each neck nodal level in patients
with early T-stage/node-positive oropharyngeal carcinoma. Int J Radiat Oncol Biol Phys.
2008;72(1):1356-1364. http://dx.doi.org/10.1016/j.ijrobp.2008.06.506
8. Jackson WC, Hawkins, PG, Arnould Gs, Yao J, Mayo C, Mierzwa M. Submandibular gland
sparing when irradiating neck level 1B in the treatment or oral squamous cell carcinoma.
Med Dosim. 2019;44(2):144-149. http://dx.doi.org/10.1016/j.meddos.2018.04.003
9. Dai X, Zhao Y, Liang Z, Dassarath M, Wang L, Jin L, Chen L, Dong J, Price RA, Ma CM.
Volumetric-modulated arc therapy for oropharyngeal carcinoma: A dosimetric and delivery
efficiency comparison with static-field IMRT. Physica Medica. 2015;31(1):54-59.
http://dx.doi.org/10.1016/j.ejmp.2014.09.003 
10. Murtaza G, Mehmood S, Rasul S, Murtaza I, Khan EU. Dosimetric effect of limited aperture
multileaf collimator on VMAT plan quality: A study of prostate and headandneck cancers.
Rep Pract Oncol Radiother. 2018;23(3):189-198. https://doi.org/10.1016/j.rpor.2018.02.006.
11. Kim YH, Park D, Park HR, et al. Effect of collimator angles on the dosimetric results of
volumetric modulated arc therapy planning for patients with a locally-advanced
nasopharyngeal carcinoma. J Korean Phys Soc . 2017;70(5):539-544.
http://doi:10.3938/jkps.70.539
12. Ahn BS, Park S-Y, Park JM, Choi CH, Chun M, Kim J-I. Dosimetric effects of sectional
adjustments of collimator angles on volumetric modulated arc therapy for irregularly shaped
targets. Plos One. 2017;12(4):1-14. https://doi.org/10.1371/journal.pone.0174924.
13. Kim YH, Park D, Park HR, et al. Effect of collimator angles on the dosimetric results of
volumetric modulated arc therapy planning for patients with a locally advanced
nasopharyngeal carcinoma. J Korean Phys Soc. 2017;70(5):539-544.
http://doi:10.3938/jkps.70.539
14. Luan S, Heintz P, Sorensen S, et al. The effect of collimator rotation on IMRT treatment
planning. Int J Radiat Oncol Biol Phys. 2005;63(1):S524-S525.
http://doi:10.1016/j.ijrobp.2005.07.886
15. Chen J, Ou D, Hu C. Sparing level IB lymph nodes by intensity-modulated radiotherapy in
the treatment of nasopharyngeal carcinoma. Int J Clin Oncol. 2013;19(6):998-1004.
http://dx.doi.org/10.1007/s10147-013-0650-6
16. The R Project for Statistical Computing. R. https://www.r-project.org/. Accessed September
30, 2019.
 Figures

Figure 1. Sagittal view of the original CTV (green) and PTV (purple) contours in relation to the
ipsilateral SMG (blue) and level 1B nodes (tan) contours.
Figure 2. Sagittal view of the new CTV (green) and PTV (purple) contours in relation to the ipsilateral
SMG (blue) and level 1B nodes (tan) contours.
Figure 3. Boxplots of mean dose (cGy) to Ipsilateral SMG by target size plan.
Figure 4. Boxplots of mean dose (cGy) to level 1B nodes by target size plan.