ORIGINAL ARTICLE
Characterization of head and neck lesions with
diffusion-weighted MR imaging and the apparent
diffusion coefficient values
Hazem M. El Shahat, Hadeer S. Fahmy, Ghada K. Gouhar
*
Department of Radiology, Faculty of Medicine, Zagazig University, Egypt Received 13 July 2013; accepted 25 August 2013
Available online 18 September 2013
KEYWORDS DWI; Head and neck
Abstract Purpose: The aim of this study was to determine the role of diffusion-weighted MR imaging (DWI) and the apparent diffusion coefficient (ADC) in characterization of head and neck lesions.
Patients and methods: MR imaging including diffusion-weighted sequences was performed on 43 patients presented with head and neck lesions. Images were obtained with a diffusion-weighted fac-tor (b facfac-tor) of 100, 500, and 1000 s/mm2. ADC maps were reconstructed, and the ADC value of the lesions was calculated.
Results: The mean ADC value of malignant tumors was (1.02 ± 0.22)· 10 3
mm2/s (n = 31). The mean ADC value of benign tumors was (1.62 ± 0.27)· 10 3
mm2/s (n = 12). The mean ADC of lymphomas was significantly lower than that of carcinomas. The difference in the ADC value between the malignant tumors and benign lesions was statistically significant (p < 0.001). Selection of (1.2)· 10 3mm2/s as a threshold value of ADC for differentiating benign from malignant
tumors yielded the best result, with an accuracy of 94%, sensitivity of 95%, specificity of 92%, posi-tive predicposi-tive value of 92% and negaposi-tive predicposi-tive value of 94%.
Conclusion: DWI and the ADC measurement are promising, non-invasive imaging approach that can be used for characterization of head and neck lesions. It can help differentiate malignant tumors from benign lesions.
2013 Production and hosting by Elsevier B.V. on behalf of Egyptian Society of Radiology and Nuclear Medicine.
1. Introduction
In spite of early enthusiasm, MR imaging did not have the last word in histological specificity and did not eliminate the need for biopsies or aspirations of the lesions. Spin echo imaging is still the mainstay of MR imaging, but now various new tech-niques hold promise for the future of head and neck imaging. Diffusion-weighted imaging (DWI) with magnetic resonance
* Corresponding author. Tel.: +20 01010109902.
E-mail addresses:[email protected],ghadagouhar@yahoo. com(G.K. Gouhar).
Peer review under responsibility of Egyptian Society of Radiology and Nuclear Medicine.
Production and hosting by Elsevier
Egyptian Society of Radiology and Nuclear Medicine
The Egyptian Journal of Radiology and Nuclear Medicine
www.elsevier.com/locate/ejrnm
www.sciencedirect.com
0378-603X 2013 Production and hosting by Elsevier B.V. on behalf of Egyptian Society of Radiology and Nuclear Medicine.
http://dx.doi.org/10.1016/j.ejrnm.2013.08.002
Open access under CC BY-NC-ND license.
relies upon the relative diffusivity of water protons within the tissue. This technique is based on the amount of random (Brownian) motion that water protons undergo. This tech-nique has proven quite useful for brain imaging in differentiat-ing between infarcted tissue and other pathological processes
(1). The potential utility of DWI for evaluating extracranial neoplastic disease is only recently recognized(2).
Diffusion-weighted images are obtained by applying pairs of opposing magnetic field gradients around the refocusing pulse of a T2*-weighted sequence. Water molecules will be de-phased by the first gradient and rede-phased by the second gradi-ent. If the water molecules are stationary, no net dephasing is expected. Movement of the tissue water molecules between the two opposing gradients will result in dephasing, depicted as signal loss on the diffusion-weighted images. Some intrinsic T2-weighted information, however, is contained in such images. Thus arithmetic processing is performed between the sets to generate apparent diffusion coefficient (ADC) maps, eliminating contributions from T2 signal changes(3).
The signal loss in diffusion-weighted images will be propor-tional to the amount of water molecule movement and the strength of the gradients (b-value). By repeating the sequence with different b-values, the observed signal loss can be quanti-fied using the ADC(4).
In normal tissue or in areas exhibiting vasogenic edema, the motion of water molecules is not limited and no restricted dif-fusion should be noted. In tissues with cytotoxic edema or in highly cellular regions, however, there is a diffusion restriction, which can be measured both qualitatively and quantitatively
(5).
Hypercellular tissue, as occurring within malignant tumors, will show low ADC values. Non-tumoral tissue changes such as edema, inflammation, fibrosis and necrosis are expected to show low cellularity, in strong contrast with viable tumor. This results in high ADC values. An inverse correlation between the ADC value and tumor cellularity in experimental models has been shown, and this was clinically validated(6,7).
In the head and neck region, DWI may have several possi-ble applications. DWI has demonstrated usefulness to discrim-inate specific histological tumor types, especially to differentiate benign solid lesions from malignant masses. DWI may be of particular value in characterization of neck lymph nodes. Differentiation of post-treatment tissue changes and persistent or recurrent cancer is another topic in which DWI may be very helpful. The inflammatory or post-treatment tissue does not demonstrate restricted diffusion, likely due to a relatively low cellularity. Recurrent or residual disease, how-ever, contains regions of increased cellularity and thus should demonstrate restricted diffusion. Furthermore, DWI can also be used as a whole body imaging at high b-values with ADC mapping to exclude or confirm metastatic disease or a second primary tumor. Moreover, DWI seems to be a safer and more affordable method considering the absence of radiation and to the higher cost of FDG-PET/CT(8–11).
Technical standarization, however, is still not achieved; re-sults obtained depend on selection of b-values. Magnetic field inhomogeneity and suboptimal placement of receiver coils can have a negative impact on image quality. Interpretation of DWI in the head and neck requires training and experience(4). The purpose of our study was to evaluate the role of DWI and ADC measurement in initial characterization of head and neck lesions.
2. Patients and methods
The study involved 43 patients (25 male, 18 female; age range 21–72 years). The inclusion criteria were patients with un-treated head and neck lesions. Informed consent was taken from all patients. MR imaging was performed on a 1.5-T MR unit (GE, Signa Exite). All patients underwent non-en-hanced T1-weighted (T1-W), fat-suppressed T2-weighted (T2-W), and gadolinium-enhanced fat-suppressed T1-weighted images. Diffusion-weighted MR images were obtained using a multislice single-shot spin-echo echo-planar sequence. Diffu-sion-Weighted MR images were acquired with a b-factor of 100, 500 and 1000 s/mm2, and ADC maps were generated.
Quantitative analysis of the ADC map was made. A region of interest (ROI) was drawn using an electronic cursor around the margin of the solid portion of the mass, taking care to avoid the cystic parts. The final diagnosis was made by endos-copy-guided biopsy in 27 patients or by histopathological examination after surgery in 16 patients. Statistical analysis was done using Excel and SPSS (Statistical Package for Social Science version 10). Data were described as mean + standard deviation (SD). Data were analyzed to test statistically signifi-cant differences. To compare between two groups, Student’s t test was applied. Receiver operating characteristic (ROC) curves were generated to determine the cutoff value, to evalu-ate the diagnostic capability of the ADC value for use in the differentiation between malignant and benign lesions and in the differentiation between lymphoma and carcinoma. 3. Results
The final diagnosis of the lesions is listed inTable 1. The mean ADC value of malignant lymphomas was (0.64 ± 0.16)· 10 3
Table 1 Diagnosis of head and neck lesions.
Diagnosis No. of lesions Malignant (n = 31)
Lymphoma 12
Squamous cell carcinoma (nasopharynx) 4 Squamous cell carcinoma (oropharynx) 3 Squamous cell carcinoma (palatine tonsil) 2 Squamous cell carcinoma (larynx) 6 Adenoid cystic carcinoma (parotid) 2
Chordoma 2
Benign (n = 12)
Pleomorphic adenoma (parotid) 8 Inflammatory lesion (occipital) 1 Schwannoma (jugular foramen) 1 Meningioma (cavernous sinus) 2
Table 2 Mean ADC values of malignant tumors and benign lesions.
Diagnosis Mean ADC values Malignant (n = 31) Lymphomas (n = 12) (0.64 ± 0.16)· 10 3 mm2/s Carcinomas (n = 17) (1.02 ± 0.22)· 10 3mm2/s Chordomas (n = 2) (0.96 ± 0.14)· 10 3mm2/s Benign (n = 12) (1.62 ± 0.27)· 10 3mm2/s
mm2/s. The mean ADC value of carcinomas was (1.02 ±0.22)· 10 3mm2/s. The mean ADC value of benign
tu-mors was (1.62 ± 0.27)· 10 3
mm2/s. There was a statistically significant difference in ADC values between malignant and be-nign tumors (p < 0.001).Table 2shows mean ADC values of malignant tumors and benign lesions. Cases are presented in
Figs. 1–5.
ROC curve analysis revealed that the threshold ADC value (1.2)· 10 3mm2/s used for differentiating between benign and
malignant head and neck lesions had an accuracy of 94%, sen-sitivity of 95%, specificity of 92%, positive predictive value of 92% and negative predictive value of 94%.
4. Discussion
Head and neck cancers account for the sixth most common type of cancer worldwide, with betel, tobacco and alcohol con-sumption being important risk factors causing significant mor-bidity and mortality. Squamous cell carcinoma (SCC) is the most common malignant histology in the head and neck region and originates from the epithelial lining of the upper aerodiges-tive tract. Approximately two-thirds of patients with head and neck cancers present with advanced stage disease, commonly involving regional lymph nodes, which require an amplified
Fig. 1 Malignant lymphoma. (A) Axial T2-weighted fat-suppression spin-echo MR image shows multiple bilateral jugular and spinal accessory chains enlarged cervical lymph nodes. (B) Axial diffusion-weighted echo-planar MR image at b factor of 1000 s/mm2, shows
high signal intensity of the enlarged lymph nodes. (C) Apparent diffusion coefficient (ADC) map shows low signal intensity of the lymph nodes with low ADC value (0.62· 10 3mm2/s). (D) Axial T2-weighted fat-suppression spin-echo MR image, in another patient shows enlarged left-sided submandibular lymph node. (E) Axial diffusion-weighted echo-planar MR image at b factor of 1000 s/mm2, shows high
signal intensity of the enlarged lymph node. (F) Apparent diffusion coefficient (ADC) map shows low signal intensity of the lymph node with low ADC value (0.61· 10 3mm2/s).
and aggressive treatment regimen consisting of neoadjuvant therapy and extensive surgery(12–14).
Differentiation of malignant head and neck tumors from be-nign lesions and accurate definite diagnosis are essential for treatment planning as well as for prognosis of malignant tu-mors. A variety of imaging techniques can help in characteriza-tion of head and neck masses. Ultrasound has a role in cystic lesions but it cannot determine the nature of solid masses. Con-ventional CT and MRI continue to be the primary imaging modalities for evaluating head and neck cancers. However, both of these modalities rely on volumetric and morphological crite-ria and consequently suffer from low sensitivity and accuracy when making the diagnosis (15). Moreover, post-treatment changes can be difficult to separate from tumor recurrence, as both entities may present with similar imaging features. Advanced MR imaging techniques provide information regard-ing the metabolic, molecular and pathophysiological aspects of a tumor. These techniques include the use of proton and
phosphorous MR spectroscopy, dynamic contrast-enhanced MR imaging and diffusion-weighted imaging in initial diagnos-tic characterization of head and neck cancers(14). The potential of FDG-PET/CT in evaluating SCC of the head and neck have demonstrated in numerous studies. Since FDG ([18F]-fluoro-2-deoxy-glucose) uptake is not specific to cancer, false positive findings, owing to inflammatory processes and metabolically ac-tive regions are common(16). Biopsy is commonly used, but it is invasive and may give false results(4,5,17).
Diffusion-weighted echo-planar MR imaging is a technique for evaluation of the rate of microscopic random water diffu-sion in tissues. The extent of translational diffudiffu-sion of mole-cules measured in the human body is referred to as the ADC. The ADC is expected to vary according to the cellular density of the lesion(18).
The aim of this study was to evaluate the role of DWI and ADC measurement in the characterization of head and neck lesions.
Fig. 2 Nasopharyngeal carcinoma. (A) Axial T2-weighted fat-suppression spin-echo MR image shows mass lesion obliterating the fossa of Rosenmuller at the left side. (B) Axial gadolinium-enhanced T1-weighted spin-echo MR image shows moderate enhancement of the mass. (C) Axial diffusion-weighted echo-planar MR image at b factor of 1000 s/mm2, the lesion appears hyperintense, indicating diffusion restriction. (D) Apparent diffusion coefficient (ADC) map shows low signal intensity of the mass with low ADC value (0.74· 10 3mm2/s).
A study proposed by Wang et al.(19), aiming at character-ization of head and neck masses by diffusion-weighted echo-planar MR imaging, found that the mean ADC value of benign solid tumors was higher than malignant tumors.
Eida et al. (20) studied ADC mapping of salivary gland tumors and concluded that ADC can provide preoperative tissue characterization of salivary gland tumors. Abdel Razek et al. (21) found that in malignant thyroid, nodules show a significantly lower ADC value than that of benign solitary thyroid nodules. White et al.(22) found a statistically signif-icant difference in ADC values between benign and malig-nant lesions involving the paranasal sinuses and skull base. Abdel Razek et al.(23) added that the mean ADC value of nasal and paranasal sinus malignant lesions is significantly different from that of benign lesions. Sumi et al.(24) stated that the ADC value is a helpful parameter in discriminating metastatic nodes in the neck. Abdel Razek et al.(25) found that the ADC map is a new technique that can help in differ-entiation of non-necrotic malignant from benign lymph nodes.
In our study, the mean ADC value of malignant head and neck tumors was significantly lower than that of benign le-sions, in agreement with the previously mentioned studies. This is explained by the difference in histopathologic features of the benign and malignant tumors. Malignant tumors have
enlarged nuclei, hyperchromatism and angulation of nuclear contour, and they show hypercellularity. These histological characteristics reduce the extracellular matrix and the diffu-sion space of water protons in the extracellular and intracel-lular dimensions with a resultant decrease in ADC values
(14,26).
Sumi et al.(27)and Maeda et al.(28)reported that squa-mous cell carcinoma has a higher ADC value than lymphoma because carcinoma may contain small foci of necrosis on histo-pathological examination that are not identifiable on MR images. Sumi et al.(29) reported that the mean ADC value of lymphomas is also lower than metastatic lymph nodes due to the difference in cellularity.
In our study, we have similar results as the mean ADC va-lue of carcinomas was significantly higher than that of malig-nant lymphomas.
Although we found significant differences in ADC values between benign and malignant lesions, there was variability in ADC values within each group. This could be because of the difference in internal architecture between lesions (30). This was in agreement with Hamphries et al.(31)and Abdel Razek et al.(32).
Limitations in our study include unavoidable image distor-tion to some degree and therefore, the ADC values measured in small lesions might be inaccurate. Also, measuring the
Fig. 3 Laryngeal glottic carcinoma. (A) Axial T2-weighted fat-suppression spin-echo MR image shows left-sided tumor mass centered on the true vocal cord, involving the anterior commissure crossing to the other side. (B) Axial gadolinium-enhanced T1-weighted spin-echo MR image shows enhancement of the lesion. (C) Axial diffusion-weighted spin-echo-planar MR image at b factor of 1000 s/mm2, the
lesion appears hyperintense, indicating diffusion restriction. (D) Apparent diffusion coefficient (ADC) map shows low signal intensity of the mass with low ADC value (0.72· 10 3mm2/s).
ADC values of the lesions located adjacent to the air-contain-ing organs is not accurate because of susceptibility artifacts.
Diffusion-weighted MR imaging using echo-planar imag-ing (EPI) has best results owimag-ing to its speed and a lower mo-tion artifact. Diffusion-weighted echo-planar pulse sequence
is the sequence of choice for the quantitative study of diffu-sion, because the diffusion and relaxation effects contribute separately to the MR signal intensity and can be easily separated; furthermore, EPI is a very fast technique that enables data acquisition, with different b values within a
Fig. 4 Chordoma. (A) Sagittal T1-weighted spin-echo MR image shows intermediate signal intensity mass centered on the clivus. (B) Axial T2-weighted spin-echo MR image shows the clival lesion. (C) Axial gadolinium-enhanced T1-weighted spin-echo MR image shows moderate enhancement of the lesion. (D) Axial diffusion-weighted echo-planar MR image at b factor of 1000 s/mm2, shows high signal intensity of the lesion. (E) Apparent diffusion coefficient (ADC) map shows low signal intensity of the mass with low ADC value (0.92· 10 3
reasonably short time. The main trade-off of the EPI pulse sequence is that it is very sensitive to magnetic susceptibility effects, resulting in geometric distortion artifacts that tend to be more severe with increasing b values(19,21,33–35). How-ever, newer techniques such as SPLICE, line scan DWI, fast asymmetric spin-echo and BLADE (PROPELLER), are being introduced in order to acquire better quality DWI data
(14).
In conclusion, diffusion-weighted MR imaging can help in the characterization of head and neck lesions, and its findings may provide useful information before surgery or biopsy. Diffusion-weighted MR imaging could be included in the routine MR imaging protocol to evaluate head and neck cancers. Diffusion-weighted MR imaging provides comple-mentary information for conventional structural MR imaging techniques.
Conflict of interest
The authors have no conflict of interest to declare. References
(1)Schaefer PW, Copen WA, Lev MH, et al. Diffusion-weighted imaging in acute stroke. Magn Reson Imaging Clin N Am 2006;14:141–68.
(2)Thoeny H, Keyzer F. Extracranial applications of diffusion-weighted magnetic resonance imaging. Eur Radiol 2007;17:1385–93. (3)Koh DM, Padhani AR. Diffusion-weighted MRI: a new func-tional clinical technique for tumor imaging. Br J Radiol 2006; 79:633–5.
(4)Hermans R, Vandecaveye V. Diffusion-weighted MRI in head and neck cancer. JBR-BTR 2007;90:264–7.
Fig. 5 Parotid gland pleomorphic adenoma. (A) Axial T2-weighted spin-echo MR image shows large mass of mixed signal intensity arising in the deep lobe of the left parotid gland, effacing the parapharyngeal space. (B) Axial gadolinium-enhanced T1-weighted spin-echo MR image shows modertae enhancement of the mass. (C) Axial diffusion-weighted echo-planar MR image at b factor of 1000 s/mm2, the mass appears of low to intermediate signal intensity. (D) Apparent diffusion coefficient (ADC) map shows high signal intensity of the mass with high ADC value (1.63· 10 3
(5)Shah GV, Wesolowski JR, Ansari SA, Mukherji SK. New directions in head and neck imaging. J Surg Oncol 2008;97:644–8. (6)Chenevert TL, Meyer CR, Moffat BA, et al. Diffusion MRI: a new strategy for assessment of cancer therapeutic efficacy. Mol Imaging 2002;1:336–43.
(7)Ross BD, Moffat BA, Lawrence TS, et al. Evaluation of cancer therapy using diffusion magnetic resonance imaging. Mol Cancer Ther 2003;2:581–7.
(8)Takahara T, Imai Y, Yamashita T, et al. Diffusion weighted whole body imaging with background body signal suppression (DWIBS): technical improvement using free breathing, STIR and high resolution 3D display. Radiat Med 2004;22:275–82. (9)Ng SH, Yen TC, Chang JT, et al. Prospective study of [18F]
fluorodeoxyglucose positron emission tomography and computed tomography and magnetic resonance imaging in oral cavity squamous cell carcinoma with palpably negative neck. J Clin Oncol 2006;24:4371–6.
(10)Mukherji SK, Bradford CR. Controversies: Is there a role for positron-emission tomographic CT in the initial staging of head and neck squamous cell carcinoma? Am J Neuroradiol 2006;27: 243–5.
(11)Gutzeit A, Antoch G, Kuhl H, et al. Unknown primary tumors: detection with dual-modality PET/CT: Initial experience. Radi-ology 2005;234:227–34.
(12)Pande P, Soni S, Kaur J, et al. Prognostic factors in betel and tobacco related oral cancer. Oral Oncol 2002;38:491–9. (13)McGurk M, Chan C, Jones J, et al. Delay in diagnosis and its
effect on outcome in head and neck cancer. Br J Oral Maxillofac Surg 2005;43:281–4.
(14)Chawla S, Kim S, Wang S, et al. Diffusion-weighted imaging in head and neck cancers. Future Oncol 2009;5(7):959–75. (15)Ahmad A, Branstetter BFT. CT versus MR: still a tough decision.
Otolaryngol Clin North Am 2008;41:1–22.
(16)Pentenero M, Cistaro A, Brusa M, et al. Accuracy of 18F-FDG-PET/CT for staging of oral squamous cell carcinoma. Head Neck 2008;30:1488–96.
(17)Meuwly J, Lepori D, Theumann N, et al. Multimodality imaging evaluation of the pediatric neck: techniques and spectrum of findings. Radiographics 2005;25:931–48.
(18)Koh D, Collins D. Diffusion-weighted MRI in the body: applications and challenges in oncology. AJR 2007;188(6): 1622–35.
(19)Wang J, Takashima S, Takayama F, et al. Head and neck lesions: characterization with diffusion-weighted echoplanar MR imag-ing. Radiology 2001;220(3):621–30.
(20)Eida S, Sumi M, Sakiha N, et al. Apparent diffusion coefficient of salivary gland tumors: prediction of the benignancy and malignancy. AJNR 2007;28:116–21.
(21)Abdel Razek AK, Sadek A, Ombar O, et al. Role of apparent diffusion coefficient value in differentiation between malignant and benign solitary thyroid nodule. AJNR 2008;29(3):563–8. (22)White ML, Zhang Y, Robinson RA. Evaluating tumors and
tumor-like lesions of the nasal cavity, the paranasal sinuses and the adjacent skull base with diffusion-weighted MRI. J Comput Assist Tomogr 2006;30:490–5.
(23)Abdel Razek A, Samir S, Bilal M. Assessment of nasal and paranasal sinus masses with diffusion-weighted echoplanar MR imaging. J Neuroradiol 2009;36:206–11.
(24)Sumi M, Sakihama N, Suni T, et al. Discrimination of metastatic cervical lymph nodes with diffusion-weighted MR imaging in patients with head and neck cancer. AJNR 2003;24:1627–34. (25)Abdel Razek A, Soliman NY, Elkhamary S, et al. Role of
diffusion MR imaging in cervical lymphadenopathy. Eur Radiol 2006;16(7):1468–77.
(26)Anderson J. Tumors: General features, types and examples. In: Anderson JR, editor. Muir’s textbook of pathology. London: Edward Arnold; 1992. p. 127–56.
(27)Sumi M, Ichikawa Y, Nakamura T. Diagnostic ability of apparent diffusion coefficient for lymphomas and carcinomas of the pharynx. Eur Radiol 2007;17(10):2631–7.
(28)Maeda M, Kato H, Sakuma H, et al. Usefulness of the apparent diffusion coefficient inline scan diffusion weighted imaging for distinguishing between squamous cell carcinomas and malignant lymphomas of head and neck. AJNR 2005;26:1186–92. (29)Sumi M, Van cautern M, Nakamura T, et al. MR microimaging
of benign and malignant nodes in the neck. AJR 2006;186: 749–57. (30)Srinivasan A, Dvorak R, Perni K, et al. Differentiation of benign and malignant pathology in the head and neck using 3T apparent diffusion coefficient values: early experience. AJNR 2008;29:40–4.
(31)Humphries PD, Sebire NJ, Siegel MJ, et al. Tumors in pediatric patients at diffusion-weighted MR imaging: apparent diffusion coefficient and tumor cellularity. Radiology 2007; 245(3):848–54.
(32)Abdel Razek A, Gaballa G, Elhawarey G, et al. Characterization of pediatric head and neck masses with diffusion-weighted MR imaging. Eur Radiol 2009;19:201–8.
(33)Chow KK, Yeung D, Chan Q, et al. DWI of head and neck cancer: the effect of b values on ADC measurements. Proc Int Soc Mag Reson Med 2011;19:2437.
(34)Morelli JN et al. Echo planar diffusion-weighted imaging: possibilities and considerations with 12- and 32-channel head coils. J Clin Imaging Sci 2012;2:31.
(35)Riffel P, Michaely HJ. Improved echo planar diffusion-weighted imaging of the head and neck using syngo ZOOmit. Clin Oncol 2013;2:68–71.
