Original Article | Neuroimaging and Head & Neck

https://doi.org/10.3348/kjr.2018.19.2.328
pISSN 1229-6929 · eISSN 2005-8330
Korean J Radiol 2018;19(2):328-333

Early Changes in Apparent Diffusion Coefficient for

Salivary Glands during Radiotherapy for Nasopharyngeal
Carcinoma Associated with Xerostomia
Qing Zhang, MD1, 2*, Yu-Mei Wei, MD3*, Yuan-Gang Qi, MD4, Bao-Sheng Li, MD5
1
Department of Radiology, Affiliated Hospital of Shandong Academy of Medical Sciences, Jinan 250031, China; 2Postdoctoral Mobile Station,
Tianjin Medical University, Tianjin 300070, China; 3Department of Head and Neck Radiotherapy, Shandong Provincial Hospital Affiliated to
Shandong University, Jinan 250022, China; Departments of 4Radiology and 5Radiation Oncology, Shandong Cancer Hospital Affiliated to Shandong
University, Shandong Academy of Medical Sciences, Jinan 250017, China

Objective: To evaluate the early changes in the apparent diffusion coefficient (ADC) of the salivary glands during
radiotherapy (RT) and their association with the degree of xerostomia at 6 months after RT in patients with nasopharyngeal
carcinoma (NPC).
Materials and Methods: We enrolled 26 patients with NPC who underwent RT. Each patient underwent diffusion-weighted
MRI of the salivary glands at rest and with gustatory stimulation within 1 week before RT and 2 weeks after the beginning
of RT. The ADC at rest (ADCR) and increase and increase rate with stimulation (ADCI, ADCIR) of the submandibular and
parotid glands were calculated. The differences in the variables’ values between 2 weeks after the beginning of RT and
baseline (ΔADCR, ΔADCI, and ΔADCIR) were compared to the degree of xerostomia at 6 months after RT.
Results: The ADCR of the submandibular and parotid glands were both significantly higher at 2 weeks after the beginning
of RT than found at baseline (both p < 0.01). The ADCI and ADCIR for the parotid glands were both significantly lower at 2
weeks after the beginning of RT than found at baseline (both p < 0.01). ΔADCI and ΔADCIR of the parotid glands were
associated with the degree of xerostomia at 6 months after RT (r = -0.61 and -0.72, both p < 0.01).
Conclusion: The ADCs of the salivary glands change early during RT. The differences in the ADC increase and increase rate
of the parotid glands between 2 weeks after the beginning of RT and baseline were associated with the degree of xerostomia
at 6 months after RT.
Keywords: Diffusion-weighted imaging; MR imaging; Parotid gland; Dry mouth; Nasopharyngeal carcinoma; Head and neck;
Radiation-induced complication; Quality of life

INTRODUCTION xerostomia, due to damage to the salivary glands, causes

difficulties in mastication and swallowing and enhances the
Xerostomia is a common radiation-induced complication risk of dental problems, subsequently degrading the quality
in patients with nasopharyngeal carcinoma (NPC) after of life in long-term survivors; however, we have no effective
a radical course of radiotherapy (RT) (1, 2). Persistent control methods (3, 4). The early detection and prevention
Received November 2, 2016; accepted after revision August 2, 2017.
This study was supported by the Projects of Medical and Health Technology Development Program in Shandong Province, People’s Republic of
China (Grant No 2015WS0158).
*These authours contributed equally to this work.
Corresponding author: Bao-Sheng Li, MD, Department of Radiation Oncology, Shandong Cancer Hospital Affiliated to Shandong
University, Shandong Academy of Medical Sciences, 440# Jiyan Road, Jinan 250017, China.
• Tel: 86053167626162 • Fax: 86053167626162 • E-mail: [email protected]
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://
creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium,
provided the original work is properly cited.

328 Copyright © 2018 The Korean Society of Radiology

ADC for Salivary Glands during Radiotherapy

of xerostomia would be of great value in practice. Radiotherapy

The development of xerostomia likely depends on All patients were fixed with a thermoplastic mask. The
various prognostic factors: radiation dose distribution, simulation CT images (slice thickness 3 mm) ranged from
demographics, and tumor and treatment-related factors the vertex cranii to 5 cm below the clavicular heads. All
(5). Many studies aimed at improving the predictive models patients received IMRT with a simultaneous integrated
for baseline subjective scores and patient-dependent risks, boost. The gross tumor volume (GTV) included the primary
but the results were discrepant (6, 7). The mechanism NPC and lymph node-involved tumors exhibited by contrast-
of radiation-induced xerostomia might be complicated. enhanced imaging. Clinical target volume 1 (CTV 1), as the
Building a predictive model that summarizes all factors high-risk region, was outlined by adding a 5-mm margin to
would be difficult. GTV, which contained the inferior sphenoid sinus, clivus,
A diffusion-weighted MRI (DWI) is commonly performed basis cranii, nasopharynx, ipsilateral parapharyngeal space,
in patients with head and neck diseases (8). The parameters posterior third of the nasal cavity, maxillary sinuses, and
of the apparent diffusion coefficient (ADC) can reflect the level II, III, and Va lymph nodes. CTV 2, as the low-risk
microstructures or pathophysiologic conditions of tissues, region, included the lower neck under the cricothyroid
even during the early stages of the disease (9). Some membrane. A 3 mm margin was also suitable in regions
studies found that the ADC of the parotid glands changed close to the brainstem, optic nerves, and optic chiasm.
(i.e., increased or decreased in different studies) after All CTVs were outlined according to the Radiation Therapy
gustatory stimulation compared to the baseline ADC (10, Oncology Group’s (RTOG) recommendations (13). The
11). Several studies investigated the diffusional alternations planning target volume (PTV) was bounded by appending a
in the salivary glands after RT (11, 12). Although these 5 mm margin to the CTV in all directions.
findings differed from each other, the ADC values certainly The prescribed RT dose was 66 Gy/30 fractions, 60 Gy/30
could reflect the function of the salivary glands and the fractions, and 54 Gy/30 fractions to the planning gross
grade of xerostomia after RT in some ways. To the best of tumor volume, PTV 1 and PTV 2, respectively, at 5 fractions
our knowledge, the early changes in the ADC of the salivary per week. After a course of RT, the total dose was 36.45
glands during RT and their association with xerostomia ± 5.62 Gy (mean ± standard deviation; range 25.21−48.83
after RT has not been documented. In this study, we sought Gy) to the parotid glands and 55.71 ± 5.13 Gy (range
to evaluate the early changes in the ADC (2 weeks after the 49.35−69.22 Gy) to the submandibular glands.
beginning of RT) of the salivary glands and their association
with the xerostomia degree at 6 months after RT. Imaging Protocol
Each patient underwent MRI within one week prior to RT
MATERIALS AND METHODS and two weeks after the beginning of RT. All patients were
asked to fast for at least 4 hours before the examination.
Patients All MRI examinations involved a 3T MR system (Achieva
From June 2014 to December 2015, 26 consecutive 3.0T; Philips Healthcare, Eindhoven, The Netherlands). The
patients with NPC who were only undergoing intensity- patients were imaged supine with a 16-channel head-neck
modulated RT (IMRT) were enrolled in another project where array coil. An Axial and sagittal T1-weighted spin-echo
the changes over time in the ADC of the salivary glands sequence (repetition time [TR] 400 ms, echo time [TE] 10.5
during RT was observed. None of the patients had a history ms, slice thickness 4 mm, slice interval 1 mm, field of view
of salivary gland disease prior to RT. No patient received [FOV] 22 x 22 cm, matrix 136 x 109) and fast spin-echo T2-
any treatment for related clinical symptoms of xerostomia weighted images were obtained first as routine procedure.
during or after RT. Here, we retrospectively analyzed these The T2-weighted sequence (TR 5563 ms, TE 46 ms, slice
data to investigate any relationships to the degree of thickness 4 mm, slice interval 1 mm, FOV 22 x 22 cm,
xerostomia at 6 months after RT. The Institutional Ethics matrix 136 x 109) was used as a reference for the following
Committee approved the study. An informed consent form DWI examinations to improve the outlining of the salivary
was signed by each patient before enrollment. glands. The images ranged from the basis cranii to the
undersurface of the submandibular glands and contained
the whole volume of the salivary glands.

kjronline.org Korean J Radiol 19(2), Mar/Apr 2018 329

 Zhang et al.

An axial spin-echo echo-planar DWI sequence was then SPSS 18.0 software (SPSS Inc., Chicago, IL, USA).
conducted (TR 5563 ms, TE 46 ms, matrix 136 x 109,
b-values 0, 200, 500, 800 s/mm2, number of signal averages RESULTS
2, scan time 87 seconds); the imaging FOV, slice thickness,
and interval were equal to those used in the T2-weighted For the 26 patients (17 men; mean age, 51 years [range
imaging. After a DWI sequence obtained at rest, six 100-mg 37−63]), at rest, the ADCR (x 10-3 mm2/s) of the parotid
tablets of ascorbic acid were placed in the patients’ mouths. and submandibular glands were both significantly higher
The patients were asked not to chew them. Immediately at 2 weeks after the beginning of RT than it was before
after, the DWI was repeated 10 times with 13-second RT (1.35 [1.20−1.52] vs. 0.94 [0.88−0.96], p < 0.01; 1.57
intervals between the contiguous sequences. [1.47−1.69] vs. 1.24 [1.18−1.29], p < 0.01) (Fig. 2). The
ADC of the parotid glands increased and then decreased in
Data Acquisition And Analysis fluctuation, and the peak value appeared within the first 6
The ADC maps for all DWI images were constructed and minutes after the beginning of stimulation both before and
transferred automatically to a workstation (Extended MR after RT. The ADCI (x 10-3 mm2/s) and ADCIR (%) values for
Workspace 2.6.3.4; Philips Healthcare). Three adjacent the parotid glands after stimulation were both significantly
sections containing the biggest volumes of each salivary lower at 2 weeks after the beginning of RT than they were
gland were used for analysis. In the T2-weighted images, the
region of interest (ROI) on each section was outlined by hand
such that it contained as much of the gland parenchyma as
possible. These ROIs were copied to the matching ADC maps
automatically for ADC calculations (Fig. 1).
The ADC values at rest (ADCR) and increase and increase
rate with stimulation (ADCI, ADCIR) for the bilateral
submandibular and parotid glands were calculated. The
ADCI was defined as the ADC difference between the peak
value during stimulation and the baseline value at rest.
The ADCIR was defined as the ratio of ADCI to the baseline
value at rest. The differences in the absolute values of A B
these variables between 2 weeks after the beginning of RT Fig. 1. Measurement of ADC of salivary glands.
T2WI image (A) and corresponding ADC map (B) for 60-year-old male
and baseline (ΔADCR, ΔADCI, and ΔADCIR) were compared
patient with nasopharyngeal carcinoma before radiotherapy. After
with the degree of xerostomia at 6 months after RT. The placing ROI in manner that enclosed as much of gland parenchyma
clinical degree of xerostomia before and at 6 months after as possible in reference to T2WI, ROI was automatically copied to
corresponding ADC map for further analysis. ADC = apparent diffusion
RT was assessed according to the morbidity scoring system coefficient, ROI = region of interest, T2WI = T2-weighted imaging
developed by the RTOG/European Organization for Research
and Treatment of Cancer (EORTC) (14).

Statistical Analysis
All data were presented as a percentage (%) or median
(interquartile range). A Wilcoxon paired test was used to
compare the ADC values of the salivary glands at rest and
during stimulation as well as before and 2 weeks after
the beginning of RT. Spearman’s Rank-Order Correlations
were used to test the associations between gender, age,
A B
ΔADCR, ΔADCI, ΔADCIR, and RT dose of the parotid and
Fig. 2. ADC maps of parotid glands of 53-year-old male patient
submandibular glands and the degree of xerostomia at 6
with nasopharyngeal carcinoma before (A) and during (B) RT.
months after RT. The level of statistical significance was Mean ADC of parotid glands at rest was 0.89 x 10-3 mm2/s before RT
set at p < 0.05. Statistical analyses involved the use of the and increased to 1.24 x 10-3 mm2/s during RT. RT = radiotherapy

330 Korean J Radiol 19(2), Mar/Apr 2018 kjronline.org

ADC for Salivary Glands during Radiotherapy

before RT (0.27 [0.21−0.33] vs. 0.36 [0.32−0.42], p < 0.01; rate of the submandibular glands did not alter significantly.
22.17 [16.30−25.20] vs. 38.36 [32.92−46.79], p < 0.01). We hypothesized that the sensitivity to radiation might be
The ADC for the submandibular glands also increased slightly different between the parotid and submandibular glands.
after stimulation both before and after RT, but the ADCI (x The differences in the ADC increase and increase rate of
10-3 mm2/s) and ADCIR (%) did not differ significantly (0.13 the parotid glands were associated with the degree of
[0.07−0.16] vs. 0.15 [0.10−0.19], p = 0.19; 12.31 [8.48− xerostomia at 6 months after RT.
17.23] vs. 16.24 [11.18−18.29], p = 0.43). In previous studies, the ADC values of the salivary glands
Before RT, xerostomia assessed by the RTOG/EORTC at rest varied greatly. The ADC of the parotid glands varied
morbidity score was grade 0 for all patients. At 6 months from 0.28 to 2.42 x 10-3 mm2/s even in healthy volunteers
after RT, xerostomia varied across patients: grade 0 for 1 (15, 16). In the present study, the ADC for the parotid
patient (3.85%), grade 1 for 12 patients (46.15%), grade 2 and submandibular glands at rest was 0.94 (0.88−0.96)
for 11 patients (42.30%) and grade 3 for 2 patients (7.70%). and 1.24 (1.18−1.29) x 10-3 mm2/s. Some authors believed
The spearman correlation coefficient indicated that ΔADCI that this discrepancy was likely attributed to the b values
and ΔADCIR of the parotid glands were associated with used because previous research showed that the ADC of
the degree of xerostomia (r = -0.61, p < 0.01; r = -0.72, the parotid gland is negatively affected by b values (10).
p < 0.01). The ΔADCR of the parotid glands was also The ADC is affected also by pulse sequences, accelerating
significantly related to xerostomia (p = 0.04) but the factors, histological components, fat saturations, and bulk
correlation coefficient was as small as 0.41. At the same motions (17, 18). Moreover, some authors insisted that high
time, the ΔADCR of the submandibular glands, gender, age b values ought to be used when assessing the diffusion of
and RT dose to the parotid and submandibular glands was the salivary glands (11) because the ADC values of obtained
not significantly related to the degree of xerostomia at biological tissues at low b values were thought to reflect
6 months after RT (r = 0.35, 0.21, 0.49, 0.51, and 0.57, the fast water motion of the blood and saliva flows known
respectively, all p > 0.05). as perfusion (10); whereas, the high b values were thought
to reflect the slow water motion known as diffusion.
DISCUSSION The contribution of perfusion to the ADC decreases with
increasing b values (19). However, Cho et al. (20) found
The DWI of the salivary glands at rest and with gustatory that salivary flow rates were severely reduced in patients
stimulation has been valuable for visualizing the diffusional who had undergone RT. We used low and high b values
alterations of the salivary glands before and after RT in in this study to reflect both the perfusion and diffusion
patients with NPC (11). We wondered whether the early changes of the salivary glands after RT. We found increased
changes in the ADC of the salivary glands of such patients ADC values for salivary glands during stimulation before
during RT were associated with xerostomia. Here, we RT, and the reaction to stimulation still occurred 2 weeks
evaluated the association of early changes (2 weeks after after the beginning of RT, which is similar to that found in
the beginning of RT) in the ADC of the salivary glands previous studies (11, 21).
during RT with the degree of xerostomia at 6 months The ADC changes of the salivary glands induced by
after RT. Because the parotid gland mainly secretes saliva radiation have been examined by some authors, but the
in stimulated conditions while the submandibular gland results differed greatly. Zhang et al. (16) first reported
mainly works at rest, some studies only investigated the reduced parotid ADC after irradiation. However, some
ADC changes of the parotid gland with stimulation (11). later research found significantly increased parotid ADC
In this study, we investigated the ADC changes of the after RT (11, 12, 22). The b-values were different in
parotid and submandibular glands. Our preliminary results these studies. In our study, we found that the ADC of the
indicated that the mean ADC values for the parotid and salivary glands increased after RT. Our results aligned with
submandibular glands at rest were both significantly higher previous studies. Zhang et al. (11) also found the increase
at 2 weeks after the beginning of RT than they were before between the maximum and the baseline was greater after
RT. The ADC increase and increase rate of the parotid glands RT than that found before RT. They attributed this finding
with stimulation during RT were both significantly lower to increased water contents in the damaged parotid glands
than they were before RT. The ADC increase and increase and delayed emptying of saliva. However, we found that the

kjronline.org Korean J Radiol 19(2), Mar/Apr 2018 331

 Zhang et al.

ADC increase and increase rate of the parotid glands with induced xerostomia in patients with head and neck cancer:
stimulation were both significantly lower during RT than pathogenesis, impact on quality of life, and management.
Head Neck 2004;26:796-807
that found before RT. We performed MRI during instead
4. Jellema AP, Slotman BJ, Doornaert P, Leemans CR, Langendijk
of after RT. Perhaps the secretion is damaged while the
JA. Impact of radiation-induced xerostomia on quality of life
salivary duct system remains normal in the early stages after primary radiotherapy among patients with head and
while undergoing RT. Kato et al. (21) found that the mean neck cancer. Int J Radiat Oncol Biol Phys 2007;69:751-760
increase rate was significantly smaller in patients with 5. van de Water TA, Bijl HP, Westerlaan HE, Langendijk JA.
primary xerostomia than it was for those without it. We Delineation guidelines for organs at risk involved in radiation-
induced salivary dysfunction and xerostomia. Radiother Oncol
supposed that the decrease in ADC increase rate of the
2009;93:545-552
parotid glands during RT may have the same mechanism
6. Grégoire V, De Neve W, Eisbruch A, Lee N, Van den Weyngaert D,
as primary xerostomia. This indicated the response of the Van Gestel D. Intensity-modulated radiation therapy for head
salivary glands to stimulation was damaged during RT. and neck carcinoma. Oncologist 2007;12:555-564
Moreover, we found that the differences in the ADC increase 7. Leek H, Albertsson M. Pilocarpine treatment of xerostomia in
and increase rate for the parotid glands between during and head and neck patients. Micron 2002;33:153-155
8. Xu XQ, Choi YJ, Sung YS, Yoon RG, Jang SW, Park JE, et al.
before RT were associated with the degree of xerostomia at
Intravoxel incoherent motion MR imaging in the head and
6 months after RT.
neck: correlation with dynamic contrast-enhanced MR imaging
The present study had some limitations. First, the and diffusion-weighted imaging. Korean Journal of Radiology
follow-up period after RT was insufficient. The degree of 2016;17:641-649
xerostomia varied over time after RT (23), and we only 9. Wang J, Takashima S, Takayama F, Kawakami S, Saito A,
investigated ADC variables with xerostomia at 6 months Matsushita T, et al. Head and neck lesions: characterization
with diffusion-weighted echo-planar MR imaging. Radiology
after RT. So the follow-up period should be increased to
2001;220:621-630
assess late side effects. Second, we used a different b-value
10. Thoeny HC, De Keyzer F, Boesch C, Hermans R. Diffusion-
than other studies. Although some authors thought that weighted imaging of the parotid gland: influence of the
high b values should be used when evaluating salivary choice of b-values on the apparent diffusion coefficient value.
glands after RT (11), we reasoned that perfusion alterations J Magn Reson Imaging 2004;20:786-790
of the salivary glands after RT should not be neglected. 11. Zhang Y, Ou D, Gu Y, He X, Peng W, Mao J, et al. Diffusion-
weighted MR imaging of salivary glands with gustatory
The standard DWI protocol has not been established, so
stimulation: comparison before and after radiotherapy. Acta
comparing these results is difficult. Finally, our sample size
Radiol 2013;54:928-933
was small. 12. Dirix P, De Keyzer F, Vandecaveye V, Stroobants S, Hermans
In this study, we demonstrated the ADC changes of R, Nuyts S. Diffusion-weighted magnetic resonance imaging
the salivary glands with gustatory stimulation during the to evaluate major salivary gland function before and after
early stages while receiving RT. The difference in the ADC radiotherapy. Int J Radiat Oncol Biol Phys 2008;71:1365-1371
13. Lee N, Harris J, Garden AS, Straube W, Glisson B, Xia P, et
increase and increase rate of the parotid glands between 2
al. Intensity-modulated radiation therapy with or without
weeks after the beginning of RT and baseline was associated
chemotherapy for nasopharyngeal carcinoma: radiation
with the degree of xerostomia at 6 months after RT. The therapy oncology group phase II trial 0225. J Clin Oncol
relationship between the early changes in the ADC of the 2009;27:3684-3690
salivary glands during RT and the degree of xerostomia 14. Cox JD, Stetz J, Pajak TF. Toxicity criteria of the Radiation
should be verified by further investigations. Therapy Oncology Group (RTOG) and the European
Organization for Research and Treatment of Cancer (EORTC).
Int J Radiat Oncol Biol Phys 1995;31:1341-1346
REFERENCES 15. Sumi M, Takagi Y, Uetani M, Morikawa M, Hayashi K, Kabasawa
H, et al. Diffusion-weighted echoplanar MR imaging of the
1. Agulnik M, Epstein JB. Nasopharyngeal carcinoma: current salivary glands. AJR Am J Roentgenol 2002;178:959-965
management, future directions and dental implications. Oral 16. Zhang L, Murata Y, Ishida R, Ohashi I, Yoshimura R, Shibuya
Oncol 2008;44:617-627 H. Functional evaluation with intravoxel incoherent motion
2. Talmi YP, Horowitz Z, Bedrin L, Wolf M, Chaushu G, Kronenberg echo-planar MRI in irradiated salivary glands: a correlative
J, et al. Quality of life of nasopharyngeal carcinoma patients. study with salivary gland scintigraphy. J Magn Reson Imaging
Cancer 2002;94:1012-1017 2001;14:223-229
3. Chambers MS, Garden AS, Kies MS, Martin JW. Radiation- 17. Chang HC, Juan CJ, Chiu HC, Cheng CC, Chiu SC, Liu YJ, et al.

332 Korean J Radiol 19(2), Mar/Apr 2018 kjronline.org

ADC for Salivary Glands during Radiotherapy

Effects of gender, age, and body mass index on fat contents 21. Kato H, Kanematsu M, Toida M, Kawaguchi T, Shibata T, Kajita
and apparent diffusion coefficients in healthy parotid glands: K, et al. Salivary gland function evaluated by diffusion-
an MRI evaluation. Eur Radiol 2014;24:2069-2076 weighted MR imaging with gustatory stimulation: preliminary
18. Liu YJ, Lee YH, Chang HC, Huang TY, Chiu HC, Wang CW, et results. J Magn Reson Imaging 2011;34:904-909
al. A potential risk of overestimating apparent diffusion 22. Studer G, Kirilova A, Jaffray D, Dawson L, Lockwood G, Bayley
coefficient in parotid glands. PLoS One 2015;10:e0124118 A, et al. Major salivary gland function: diffusion-weighted
19. Le Bihan D, Breton E, Lallemand D, Aubin ML, Vignaud J, MRI (DWI) assessment before, during and after radiation
Laval-Jeantet M. Separation of diffusion and perfusion therapy. International Journal of Radiation Oncology, Biology,
in intravoxel incoherent motion MR imaging. Radiology Physics 2005;63:S361
1988;168:497-505 23. Lin A, Kim HM, Terrell JE, Dawson LA, Ship JA, Eisbruch A.
20. Cho MA, Ko JY, Kim YK, Kho HS. Salivary flow rate and clinical Quality of life after parotid-sparing IMRT for head-and-neck
characteristics of patients with xerostomia according to its cancer: a prospective longitudinal study. Int J Radiat Oncol
aetiology. J Oral Rehabil 2010;37:185-193 Biol Phys 2003;57:61-70

kjronline.org Korean J Radiol 19(2), Mar/Apr 2018 333