Dynamic susceptibility contrast perfusion MR imaging in distinguishing ...

15 downloads 78 Views 684KB Size Report
lesions. Mean and SD of DSC%. Range of. DSC%. Malignant neoplasm. 40. -Adenoid cystic carcinoma. 6. 49.1 ± 7.5. 37.5–56.7. -Mucoepidermoid carcinoma. 5.
European Journal of Radiology 77 (2011) 73–79

Contents lists available at ScienceDirect

European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad

Dynamic susceptibility contrast perfusion MR imaging in distinguishing malignant from benign head and neck tumors: A pilot study Ahmed Abdel Khalek Abdel Razek a,∗ , Lamiaa Galal Elsorogy a , Nermin Yehia Soliman a , Nadia Nada b a b

Diagnostic Radiology Department, Mansoura Faculty of Medicine, 62 Elgomheryia Street, Mansoura 35512, Egypt Pathology Department, Mansoura Faculty of Medicine, 62 Elgomheryia Street, Mansoura 35512, Egypt

a r t i c l e

i n f o

Article history: Received 16 May 2009 Received in revised form 16 July 2009 Accepted 16 July 2009 Keywords: Tumor Perfusion MR imaging Malignant

a b s t r a c t Purpose: To preliminarily investigate the utility of dynamic susceptibility contrast perfusion MR imaging in distinguishing malignant from benign head and neck tumors. Material and methods: Seventy eight patients with head and neck masses underwent single shot dynamic susceptibility contrast T2*-weighted perfusion weighted MR imaging after bolus infusion of gadoliniumDTPA was administrated. The signal intensity time curve of the lesion was created. Dynamic susceptibility contrast percentage (DSC%) was calculated and correlated with pathological findings. Results: The mean DSC% of malignant tumor (n = 40) was 39.3 ± 9.6% and of benign lesions (n = 38) was 24.3 ± 10.3%. There was a statistically significant difference of the DSC% between benign and malignant tumors (P = 0.001) and within benign tumors (P = 0.001). When DSC% of 30.7% was used as a threshold for differentiating malignant from benign tumors, the best results were obtained: accuracy of 84.6%, sensitivity of 80% and specificity of 89.2%. Conclusion: Dynamic susceptibility contrast perfusion weighted MR imaging is a non-invasive imaging technique that can play a role in differentiation between malignant and benign head and neck tumors. © 2009 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Differentiation between benign and malignant head and neck tumor and prediction of pathological nature and grading of these tumors are essential for treatment planning and for prognosis of malignant tumors. Differentiation between benign and malignant head and neck tumors may be problematic with CT and MR imaging in some cases [1,2]. Parameters of CT perfusion seem to overlap between malignant tumors and benign lesions with small area of covering [3] and PET CT shows variable physiologic FDG uptake with false negative results in necrotic tumors [4,5]. The evaluation of the relaxation times, enhancement characteristics, magnetization transfer ratios or apparent diffusion coefficients has proved unreliable [6–9]. Fine needle aspiration biopsy is sensitive, but it may be inconclusive, have discordant cytology findings, or suffer from inadequate samples [10,11]. Dynamic susceptibility contrast (DSC) perfusion MR imaging involves repetitive serial imaging through the tumor during the passage of blood that has been labeled with either contrast material or with an endogenous magnetic tracer label [12–15]. It can be obtained with bolus tracking technique that monitors the pas-

∗ Corresponding author. Tel.: +20 161948567; fax: +20 502259146. E-mail address: [email protected] (A.A.K.A. Razek). 0720-048X/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2009.07.022

sage of contrast medium through a capillary bed. It is based upon inhomogeneity of the magnetic field during the passage of a short bolus of contrast medium through the capillary bed [12–17]. It has been used for evaluation of tumor angiogenesis in different regions of the body notably for brain, hepatic and breast malignancy [12–20]. Our purpose was to preliminarily investigate the utility of dynamic susceptibility contrast perfusion MRI in distinguishing malignant from benign head and neck tumors. 2. Material and methods A prospective study was conducted on 81 consecutive patients (52 male and 29 female) with head and neck masses. Their age ranged from 18 years to 71 years (mean age 43 years). The patients enrolled in this study were patients referred to the MR unit with clinical head and neck masses that were candidates for surgery or biopsy. Three patients were excluded from the study: one patient has a cystic mass, another patient had a motion artifact and last patient showed a small lesion (8 mm) with poor image quality due to a susceptibility artifact. The finally studied group included 78 patients. All patients underwent routine T1 and T2 weighted MR imaging, dynamic susceptibility perfusion weighted MR imaging and finally routine post-contrast MR imaging of head and neck. We obtained institutional review board approval and informed consent from the patients before the study.

74

A.A.K.A. Razek et al. / European Journal of Radiology 77 (2011) 73–79

All MR images were acquired on a 1.5 T scanner (Symphony; Siemens Medical Systems, Erlangen, Germany) equipped with a self-shielding gradient set (30 mTm maximum gradient strength, 120 T/m/s slew rate). The lower part of the circularly polarized (CP) head coil and a standard two-element CP neck array coil were used. All patients underwent T1-weighted images (TR/TE of 800/15 ms) and T2-weighted fast spin-echo images (TR/TE of 6000/80 ms) with a section thickness of 5 mm, an inter-slice gap of 1.5 mm, a fieldof-view (FOV) of 25–30 cm and an acquisition matrix of 256 × 224. Multi-slice echo-planar imaging gradient echo sequence was used. The scanning parameters applied were: TR = 2280 ms, TE = 47 ms, number of excitation = 1, flip angle = 80◦ , section thickness = 5 mm, inter-slice gap = 1.5 mm, FOV = 25–30 cm, matrix = 128 × 128 and signal bandwidth = 1345 Hz/pixel. These parameters were selected to obtain images with adequate resolution with decreased image distortion. Dynamic susceptibility contrast T2*-weighted MR images were obtained following administration of gadopentate dimegulumine in a dose of 0.2 mmol/kg body weight. The injection was performed by automatic injector in the right arm at a rate of 5 ml/s followed by 20 ml saline. The rapid injection of high dose of gadopentate dimegulumine was chosen to achieve high concentration of contrast medium that was well tolerated with all patients. The gadolinium was administrated after 8 s of data acquisition, so that there was base line data. The data acquisition time was 110 s, and the time between the data points was 2 s. The number of slices was 20 slices with 55 acquisitions for each slice, and the total number of images obtained was 1100 images. After the dynamic study, a routine post-contrast study was obtained. One radiologist (SN) determined the region-of-interest (ROI) around the suspicious lesion using an electronic cursor. The ROI was placed around the margin of the homogenously enhanced mass (Fig. 1). When heterogeneity in the signal intensity was observed, an ROI was placed around the enhanced, solid part of the tumor and avoided the cystic or necrotic part of the tumor, based upon contrast-enhanced MR imaging results. In patients with multiple cervical lymphadenopathies, the ROI was placed around the largest one. Also, for patients with malignant tumors and metastatic lymph nodes, the ROI was placed around the primary tumor only. The size of ROI varies from 2.2 to 19.6 cm2 (mean 7.4 cm2 ). The time signal intensity curve for the ROI was plotted. The horizontal axis of the plot represented the time and the vertical axis represented the signal intensity (Fig. 2). Then, the ROI was copied into the pre- and post-contrast T1 weighted images.

Fig. 2. Calculation of DSC% from time–signal intensity curve: The upper arrow denotes the unenhanced lesion signal intensity (S0) and the lower arrow denotes the maximum contrast-enhanced signal intensity (SI).

The MR images and time–signal intensity curves were independently reviewed by three radiologists (A.A., E.L., S.N., with 20, 12, 8 years of experience in head and neck imaging, respectively) who were blinded to the other clinical, imaging, and final pathological examinations. The quality of the MR images was evaluated based on if they were acceptable or not. The time–signal intensity curve was evaluated qualitatively for the return of signal intensity to the base line and 2nd recirculation time, and then quantitative analysis of the curve was done. Disagreement between radiologists was resolved in consensus. The DSC% was calculated as: S0 − SI/S0 × 100% where S0 represents the signal intensity of the lesion just before descent of signal intensity, and SI represents the signal intensity at peak descent [17]. The contrast-enhanced percentage (CE%) for contrast-enhanced images was calculated as: SI − S0/S0 where SI and SO are signal intensity of the ROI on the contrast-enhanced and pre-contrast images, respectively. The final diagnosis was done with surgical biopsy (n = 47), fine needle aspiration biopsy (n = 19) and core biopsy (n = 12). The biopsy was done after MR imaging with a time delay 5–15 days. The specimens were interpreted by a pathologist experienced in head and neck. The data were determined to be parametric with a normal distribution. Data analysis was done to test for statistical significance. To compare between benign and malignant tumors, a two-tailed Student t-test was used. To compare within pathological subtypes of benign or malignant tumors and grading of malignant tumors, oneway ANOVA test was used. A value of P < 0.05 was defined as being significant. The data were presented in the bar graph. We used the receiver operating characteristic (ROC) curve to determine the cut off value of DSC% and CE% for differentiating malignant from benign tumors. The cut off point was chosen at the highest sensitivity and specificity. The sensitivity, specificity, accuracy and area under the curve were computed with exactly 95% confidence interval based on F distribution. The statistical analysis of data was done by using SPSS program (Statistical package for social science version 10). 3. Results

Fig. 1. Region-of-interest (ROI) localization: Axial susceptibility perfusion weighted MR image shows the ROI is drawn by hand around the margin of the parapharyngeal mass.

The pathological results reported 40 malignant tumors and 38 benign lesions. Malignant tumors included squamous cell

A.A.K.A. Razek et al. / European Journal of Radiology 77 (2011) 73–79

75

Table 1 DSC% of head and neck neoplasm at dynamic susceptibility perfusion weighted images. Pathology

No. of lesions

Mean and SD of DSC%

Range of DSC%

Malignant neoplasm -Adenoid cystic carcinoma -Mucoepidermoid carcinoma -Rhabdomyosarcoma -Non-Hodgkin lymphoma -Squamous cell carcinoma -Metastasis

40 6 5 3 5 19 2

49.1 ± 7.5 45.6 ± 6.9 45.4 ± 7.6 41.8 ± 5.1 34.8 ± 8.6 29.8 ± 3.6

37.5–56.7 36.3–55.2 38.3–52.7 35.3–47.3 19.6–51.4 27.2–32.4

Benign neoplasm -Angiofibroma -Glomus tumor -Pleomorphic adenoma -Schwanoma -Inverted papilloma -Warthin tumors -Parathyroid adenoma -Fibroma

38 2 2 17 6 3 3 3 2

52.2 ± 6.3 51.9 ± 3.1 23.4 ± 4 21.5 ± 3 20.2 ± 3.4 19.6 ± 2 17.4 ± 1.1 15.1 ± 0.6

47.7–56.7 49.7–54.1 18.6–29.1 17.9–27.7 17.3–24.1 17.6–21.6 16.4–18.6 14.7–15.6

carcinoma (n = 19), adenoid cystic carcinoma (n = 6), mucoepidermoid carcinoma (n = 5), non-Hodgkin lymphoma (n = 5), rhabdomyosarcoma (n = 3) and hematogenous skull base metastasis from cancer breast (n = 2). The benign lesions were pleomorphic adenoma (n = 17), schwanoma (n = 6), Warthin tumors (n = 3), inverted papilloma (n = 3), parathyroid adenoma (n = 3), fibroma (n = 2), angiofibroma (n = 2) and glomus tumors (n = 2). They were located in the pharynx (n = 26), parotid space (n = 24), visceral space (n = 16), oral cavity (n = 6), masticator space (n = 5), parapharyngeal space (n = 4), submandiblar region (n = 4) and skull base (n = 4). The diameter of the tumors ranged from 1.2 cm to 15 cm (mean 7.6 cm) Tables 1 and 2 shows the mean, standard deviation and range of DSC% and CE% of head and neck neoplasm respectively. The curve shows a base line area before contrast medium injection that is followed by rapid decrease in signal intensity. The signal intensity either returns (n = 30) or fails to return (n = 48) to the base line. Another, smaller 2nd peak of signal intensity loss was detected in 14 malignant tumors and 8 benign lesions. There was insignificant difference in DSC% of the 2nd recirculation peak between malignant tumors and benign lesions (P = 0.08). The DSC% and CE% of the CSF were 3.2 ± 0.3% and 0.08 ± 0.2%, respectively. The DSC% of malignant tumor was 39.3 ± 9.6% and of benign lesions was 24.3 ± 10.3%. The difference in DSC% between malignant tumors and benign tumors was statistically significant (P = 0.001). The CE% of malignant and benign tumors was 0.55 ± 0.3% and 0.36 ± 0.2%, respectively. There was insignificant Table 2 CE% of head and neck neoplasm at pre- and post-contrast T1 weighted MR imaging. Pathology

No. of lesions

Mean and SD of CE%

Range of CE%

Malignant neoplasm -Adenoid cystic carcinoma -Mucoepidermoid carcinoma -Rhabdomyosarcoma -Non-Hodgkin lymphoma -Squamous cell carcinoma -Metastasis

40 6 5 3 5 19 2

0.67 ± 0.8 0.54 ± 0.1 0.53 ± 0.2 0.52 ± 0.1 0.51 ± 0.1 0.50 ± 0.5

0.57–0.7 0.32–0.8 0.32–0.9 0.32–0.7 0.42–0.6 0.44–0.5

Benign neoplasm -Angiofibroma -Glomus tumor -Pleomorphic adenoma -Schwanoma -Inverted papilloma -Warthin tumors -Parathyroid adenoma -Fibroma

38 2 2 17 6 3 3 3 2

0.51 ± 0.1 0.46 ± 0.1 0.43 ± 0.1 0.40 ± 0.6 0.46 ± 0.1 0.43 ± 0.8 0.52 ± 0.5 0.29 ± 0.3

0.51–0.5 0.46–0.6 0.25–0.6 0.31–0.6 0.37–0.6 0.37–0.5 0.47–0.6 0.27–0.3

Fig. 3. Bar graph of malignant and benign tumors: (A) Bar graph of DSC% shows the mean DSC% of malignant tumor (39.3%) is significantly different (P = 0.001) than benign tumors (24.3%). (B) Bar graph of CE% shows there is overlap in CE% of malignant (0.55) and benign (0.0.36) head and neck tumor with insignificant difference in between (P = 0.07).

difference (P = 0.07) in the CE% between the malignant and benign tumors (Fig. 3). The DSC% of malignant tumors ranged from 19.6% to 56.7% (mean = 39.3%). There was insignificant difference within malignant tumors (P = 0.07). The adenoid cystic (Fig. 4) and mucoepidermoid carcinoma showed very similar high ranges of DSC%. Squamous cell carcinoma showed low DSC% compared to other carcinomas and sarcomas but not reach to a significant level. The DSC% of benign tumors ranged from 14.7% to 56.7% (mean = 24.3 ± 10.3%) (Fig. 5). There was significant difference in the DSC% within benign tumors (P = 0.001). The lowest DSC% was noted in patient with fibroma (14.7%). Pleomorphic adenomas and schwanomas revealed a wide range of signal intensity loss. There was significant difference in the DSC% between vascular benign tumors and other benign tumors. Benign vascular lesions (2 angiofibroma and 2 glomus tumors) showed marked DSC% that was mistaken as malignant lesions. The ROC curve stated that when the DSC% of 30.7% was used as a threshold for differentiating malignant from benign tumors, it showed accuracy of 84.6%, sensitivity of 80%, specificity of 89.2% and area under the curve of 0.86 (0.77–0.99 95% confidence interval). The threshold of DSC% (33%) used for differentiating malignant from benign, non-vascular tumors revealed accuracy of 93%, sensitivity 91%, specificity of 79%% and an area under the curve of 0.93 (0.86–1 95% confidence interval). The threshold of CE% used was 0.44% revealed accuracy of 62.3%, sensitivity 75%, specificity of 49.6% and an area under the curve of 0.73 (0.62–0.84 95% confidence interval) (Fig. 6). 4. Discussion Tumor angiogenesis is a complex process where new vessels grow towards and within the tissue with increased angiogenic

76

A.A.K.A. Razek et al. / European Journal of Radiology 77 (2011) 73–79

Fig. 4. Adenoid cystic carcinoma of the tongue: (A) Post-contrast sagittal T1 weighted MR image shows: an intense enhancing mass is seen involving the intrinsic muscles of the tongue. (B) Time–signal intensity curve shows the calculated DSC% is 52.3%.

Fig. 5. Nasal inverted papilloma: (A) Axial post-contrast T1 weighted images shows a well-defined enhancing mass is seen in the right nasal cavity. (B) Time–signal intensity curve shows the DSC% of benign tumor is 17.3%.

activity that is mediated by factors released from malignant tumor cells therefore, perfusion scans can potentially image tumor activity before the tumor develops gross anatomic distortion. The role of imaging has begun to shift from purely anatomical to provide information on tumor physiology (21–30). Dynamic contrast-enhanced perfusion weighted MR imaging yield different quantitative parameters that reflect the angiogenesis. The MR imaging methods used for imaging of angiogenesis are dynamic susceptibility contrast perfusion weighted MRI (DSC–MRI) and dynamic contrast-enhanced MRI (DCE–MRI). While contrast enhancement in DCE–MR depends on several factors such as pre-contrast T1 value of the tumor, tumor perfusion, tumor interstitial matrix, capillary density and capillary permeability, DSC–MR T2*-perfusion weighted imaging only responds to more specific changes in microvascular perfusion [21–30]. DCE–MRI permits a full depiction of the wash-in and wash-out contrast kinetics within tumors, and thus provides insight into the nature of the bulk tissue properties [29,30]. DCE–MRI has been used to differentiate benign from malignant neoplasm in different regions of the body based upon different rates of contrast uptake. In head and neck, it has been used for characterization of cervical lymph nodes, parotid tumors and for differentiation of

post-treatment changes from recurrent tumors. The success of this method has been questioned due to overlap in quantitative parameters of benign and malignant tumors [31–34]. Recently, Bisdas et al. [12] have been used gadolinium enhanced spin echo (SE) of echo-planar T2 weighted MR of extracranial head and neck tumors with calculation of perfusion parameters, however, this study lack of validation of perfusion estimates derived from imaging. Also, Michaely et al. [21] characterize and quantify the vascularization and hemodynamic characteristics of head and neck tumors with a dynamic 3D time-resolved echo-shared angiographic technique (TREAT) using the regular contrast agent bolus. But, the study was done on a small number of patients without calculation of perfusion parameters. Rumboldt et al. [3] reported that CT perfusion (CTP) shows promise in distinguishing benign and malignant processes, primarily by means of mean transit time maps. But, perfusion parameters seem to overlap for malignant lesions, the salivary glands, and the thyroid glands. Dynamic susceptibility contrast MR imaging is well suited for evaluation of tumor angiogenesis since the degree of signal loss depends on the volume of the intravascular space within the tumor and on the concentration of injected contrast agent within the blood [14–18]. It reflects the physiology of the microcircula-

A.A.K.A. Razek et al. / European Journal of Radiology 77 (2011) 73–79

77

Fig. 6. Receiver operating characteristic (ROC) curve: (A) ROC curve for DSC% shows the DSC% of 30.7% was used as a threshold for differentiating malignant from benign tumors with an accuracy of 84.6%, sensitivity of 80%, specificity of 89.2% and an area under the curve of 0.86. (B) ROC curve for DSC% shows the DSC% used for differentiation malignant tumors from benign non-vascular lesions is 33% with accuracy of 93%, sensitivity of 91%, specificity 79% and an area under the curve of 0.93. (C) ROC curve for CE% shows the threshold of CE% used is 0.445 with an accuracy of 62.3%, sensitivity 75%, specificity of 49.6% and area under the curve of 0.73.

tion, especially the microvasculature and the extracellular space [19–20]. To our knowledge, no previous studies in the literature explored the role of DSC–MR imaging in head and neck. We evaluated the potential use of dynamic susceptibility perfusion weighted MR imaging in differentiating benign from malignant head and neck masses as well as the different sub-types of malignant and benign tumors. In this study, the difference in the DSC% between malignant and benign head and neck tumors at dynamic susceptibility contrast perfusion MR imaging was statistically significant (P = 0.001), but tissue sampling is done for specific tissue diagnosis. This was explained by highly vascularity with increased capillary perfusion of malignant tumors compared to benign tumors. The tumor blood vessels are typically dilated and tortuous with abnormal branching pattern, blind ends and no organization into arterioles, capillaries and venules [24–26]. Not only the number, but also, the size of the vessels has been found to increase with increased tumor angiogenesis. Several unique properties of new vessels including increased tumor blood volume, arterio-venous shunt formation, altered capillary transit time and increased capillary permeability [24–27].

Differentiating pathologic types of head and neck tumors, is essential for treatment planning and establishing the prognostic factors of these lesions. The degree of angiogenesis is important in the evaluation of different tumor types and the determination of the biologic aggressiveness of malignant tumors. Dynamic susceptibly contrast imaging has been used for evaluation of brain and hepatic tumors [12–17]. In this study, the DSC% of malignant tumors varies according to pathological type. Mucoepidermoid and adenoid cystic carcinoma show the highest DSC% with overlap within malignant tumors. Squamous cell carcinoma showed low DSC% compared to other carcinomas and sarcomas but not reach to a significant level. In this study, pleomorphic adenomas and schwanomas show a wide range of DSC% that depends on the degree of tumor vascularity. Angiofibroma and glomus tumors are benign, vascular tumors that show marked DSC% and were mistaken as malignancy. Glomus tumors and angiofibromas usually have a specific MR imaging appearance, but pleomorphic adenomas and schwanomas can be quite variable. This was explained by Schnall [35] who stated that during the first pass, the concentration of gadolinium determined by the density of blood vessels and the permeability of the capil-

78

A.A.K.A. Razek et al. / European Journal of Radiology 77 (2011) 73–79

laries that perfuse the tissue. Any lesion with high vascularity and permeability, such as vascular tumors, will demonstrate a rapid increase in gadolinium concentration. On the other hand, glomus tumor and angiofibromas show characteristic signal void structures with intense contrast enhancement on routine MR imaging which is sufficient for an accurate and definitive diagnosis [36–38]. Echo-planar MR imaging system is the method of choice for dynamic susceptibility perfusion imaging with rapid image acquisition allows greater anatomical coverage. The single shot EPI produces a temporal resolution of 2 s. During a 2 s acquisition window it is usually possible to acquire approximately 15–20 slices. The signal-to-noise ratio can be improved by using a high dose of contrast medium (i.e. 2 or more mmol/kg). A series of at least five pre-contrast images should be collected prior to the passage of bolus to improve the estimation of signal intensity baseline during analysis [29,30,35]. In this study, we applied only 2 mmol/kg which is sufficient for adequate signal-to-noise ratio of the images. Schnall [35] reported that the ROI used should be placed only over the enhanced component of the lesion and should not encompass the entire lesion. The ROI analysis is prone to errors since it is operator dependent. Fischbein et al. added that multiple ROIs with a mean value should be tried if the lesion shows significant signal heterogeneity. However, volume averaging and areas of micronecrosis is the source of sampling error. In this study, to avoid bias from the necrotic part of tumors, the ROI was drawn by freehand and excluded areas of gross necrosis. In this study, there is failure of the signal intensity to return to the base line due to leakage of contrast medium. Padhani showed that when a bolus of low molecular weight paramagnetic contrast agent passes through a capillary bed, it is transiently confined to within the vascular space. In dynamic susceptibility perfusion weighted brain studies with an intact blood brain barrier, the contrast agent behave as true intravascular agents, while in extracranial region such as liver and breast, the contrast agent rapidly passes into the extravascular extracellular space (leaky space). In tumors, 12–45% of the contrast medium leaks into the extracellular space [15,24–30]. Further studies are recommended using macro-aggregate contrast medium without leakage of contrast medium to increase the diagnostic performance of this technique [36]. In this study, the recirculation of contrast medium with a 2nd smaller and broader peak was detected. Jackson [27] reported that the contrast recirculates through the body, and a second recirculation peak is seen following the first. As the contrast continues to circulate, the bolus disperses and widens so that the second peak is lower and broader than the first peak. The recirculation peak could not help in narrowing the differential diagnosis between malignant tumors and benign lesions. Echo-planar MR imaging in head and neck may be associated with susceptibility artifacts at the air/tissue interface [26,27]. In this study, a susceptibility artifact was noted in one patient with a small lesion (8 mm) at the skull base. The image quality of single shot echo-planar sequence is still too poor to detect millimetersized lesions with great confidence. Software improvements may improve image quality with subsequent decreased susceptibility artifacts. There are few limitations to this study. The patient population studied is a diverse set of malignant and benign tumors with a mixture of pathological types in different regions of head and neck at variable ages. Further studies are needed to detect histopathological subtypes of tumors since this is important in treatment planning of head and neck tumors and to evaluate the tumors and outliers of each group explicitly based on age and location. There is no intraobserver kappa statistics, and the statistical sample size in each category is too small. We assessed only the primary tumor and did not assess the associated cervical lymph nodes. Lastly, we use low

molecular contrast medium that associated with leakage of contrast medium and CBV map are not created. Further studies are recommended with usage of macromolecular to avoid leakage with calculation of blood volume of the tumor. Parameters for T1 and T2 are suboptimal with relatively poor spatial resolution. 5. Conclusion We concluded that dynamic susceptibility contrast perfusion weighted MR imaging is a non-invasive MR imaging technique that can play a role in differentiation between benign non-vascular and malignant head and neck tumors as it reflects the difference in vascularization among benign and malignant head and neck tumors. Conflict of interest The authors have declared no conflict of interest. References [1] Yousem DM, Gad K, Tufano RP. Respectability issues with head and neck cancer. AJNR Am J Neuroradiol 2006;27:2024–36. [2] Bhattacharyya N. Prognostic factors for neoplasia and malignancy in a neck mass. Arch Otolaryngol Head Neck Surg 1999;125:303–7. [3] Rumboldt Z, Al-Okaili R, Deveikis J. Perfusion CT for head and neck tumors: pilot study. AJNR Am J Neuroradiol 2005;26:1178–85. [4] Hannah A, Scott AM, Tochon-Danguy H, et al. Evaluation of 18 Ffluorodeoxyglucose positron emission tomography and computed tomography with histopathologic correlation in the initial staging of head and neck cancer. Ann Surg 2002;236:208–17. [5] Fukui M, Blodgett T, Snyderman C, et al. Combined PET-CT in the head and neck Part 2. Diagnostic uses and pitfalls of oncologic imaging. Radiographics 2005;25:913–30. [6] Takashima S, Sone S, Harii A, Okamoto S, Yashida J. Major salivary gland lesions: correlation of MR findings with flow cytometric DNA analysis and prognosis. AJR Am J Roentgenol 1996;167:1297–304. [7] Takashima S, Naguchi Y, Okumura T, Aruga H, Kobyashi T. Dynamic MR imaging in the head and neck. Radiology 1993;189:813–21. [8] Markkola A, Aronen H, Paavonen T, et al. Spin lock and magnetization transfer imaging of head and neck tumors. Radiology 1996;200:369–75. [9] Wang J, Takashima S, Takayama F, et al. Head and neck lesions: characterization with diffusion – weighted echoplanar MR imaging. Radiology 2001;220:621–30. [10] Sack M, Weber R, Weinstein G, Chalian A, Nisenbaum H, Yousem D. Imageguided fine-needle aspiration of the head and neck five years experience. Arch Otolaryngol Head Neck Surg 1998;124:1151–61. [11] Gupta S, Henningsen G, Wallace M, et al. Percutaneous biopsy of head and neck lesions with CT guidance: various approaches and relevant anatomic and technical considerations. Radiographics 2007;27:371–90. [12] Bisdas S, Baghi M, Wagenblast S, Vogl T, Thng C, Koh T. Gadolinium-enhanced echo-planar T2-weighted MRI of tumors in the extracranial head and neck: feasibility study and preliminary results using a distributed-parameter tracer kinetic analysis. J Magn Reson Imaging 2008;27:963–9. [13] Cha S, Knopp E, Johnson G, Wetzel S, Litt A, Zagzag D. Intracranial mass lesions: dynamic contrast-enhanced susceptibility-weighted echo-planar perfusion MR imaging. Radiology 2002;223:11–29. [14] Cha S. Perfusion MRI of brain tumors. Top Magn Reson Imaging 2004;15:279–89. [15] Sugahara T, Korogi Y, Shigematsu Y, et al. Value of dynamic susceptibility contrast magnetic resonance imaging in the evaluation of intracranial tumors. Top Magn Reson Imaging 1999;10:114–24. [16] Ichikawa T, Haradome H, Hachiya J, Nitatori T, Araki T. Characterization of hepatic lesions by perfusion-weighted MR imaging with an echoplanar sequence. AJR Am J Roentgenol 1998;170:1029–34. [17] Ichikawa T, Arab A, Araki T, et al. Perfusion MR imaging with a supermagenetic iron oxide using T2-weighted and susceptibility–sensitive echoplanar sequences: evaluation of tumor vascularity in hepatocellular carcinoma. AJR Am J Roentgenol 1999;173:207–13. [18] Kvistad K, Lundgren S, Fjosne H, Smenes E, Smethurst H, Haraldseth O. Differentiation benign and malignant breast lesions with T2*-weighted first pass perfusion imaging. Acta Radiol 1999;40:45–51. [19] Kvistad K, Rydland J, Vainio J, et al. Breast lesions: Evaluation with dynamic contrast – enhanced T1-weighted MR imaging and with T2*-weighted first pass perfusion MR imaging. Radiology 2000;216:545–53. [20] Kuhl CK, Bieling H, Gieseke J, et al. Breast neoplasms: T2* susceptibilitycontrast, first-pass perfusion MR imaging. Radiology 1997;202:87–95. [21] Michaely H, Herrmann K, Dietrich O, Reiser M, Schoenberg S. Quantitative and qualitative characterization of vascularization and hemodynamics in head and neck tumors with a 3D magnetic resonance time-resolved echo-shared angiographic technique (TREAT)—initial results. Eur Radiol 2007;17:1101–10.

A.A.K.A. Razek et al. / European Journal of Radiology 77 (2011) 73–79 [22] Folkman J. Tumor angiogenesis: therapeutic implications. N Engl J Med 1971;285:1182–6. [23] Kuszyk B, Corl F, Franano N, et al. Tumor transport physiology: implications for imaging and imaging-guided therapy. AJR Am J Roentgenol 2001;177:747–53. [24] Fabian Kiessling F, Jugold M, Woenne E, Brix G. Non-invasive assessment of vessel morphology and function in tumors by magnetic resonance imaging. Eur Radiol 2007;17:2136–48. [25] Provenzale J. Imaging of angiogenesis: clinical techniques and novel imaging methods. AJR Am J Roentgenol 2007;188:11–23. [26] Miller J, Pien H, Sahani D, Sorensen A, Thrall J. Imaging angiogenesis: applications and potential for drug development. J Natl Cancer Inst 2005;97:172–86. [27] Jackson A. Analysis of dynamic contrast enhanced MRI. Br J Radiol 2004;77:S154–66. [28] Rijpkema M, Kaanders J, Joosten F, van der Kogel A, Heerschap A. Method for quantitative mapping of dynamic MRI contrast agent uptake in human tumors. J Magn Reson Imaging 2001;14:457–63. [29] Padhani R. MRI for assessing antivascular cancer treatments. Br J Radiol 2003;76:S60–80. [30] Choyke P, Dwyer A, Knopp M. Functional tumor imaging with dynamic contrast-enhanced magnetic resonance imaging. J Magn Reson Imaging 2003;17:509–20.

79

[31] Takashima S, Naguchi Y, Okumura T, Aruga H, Kobyashi T. Dynamic MR imaging in the head and neck. Radiology 1993;189:813. [32] Fischbein N, Noworolski S, Herny R, Kaplan M, Dillon W, Nelson S. Assessment of metastatic cervical adenopathy using dynamic contrast-enhanced MR imaging. AJNR Am J Neuroradiol 2003;24:301–11. [33] Yabuuchi H, Fukuya T, Tajima T, Hachitanda Y, Tomita K, Koga M. Salivary gland tumors: diagnostic value of gadolinium-enhanced dynamic MR imaging with histopathologic correlation. Radiology 2003;226:345–54. [34] Lai P, Yang C, Pan H, et al. Recurrent inverted papilloma: diagnosis with pharmacokinetic dynamic gadolinium-enhanced MR imaging. AJNR Am J Neuroradiol 1999;20:1445–51. [35] Schnall MD. An overview of interpretation strategies for breast MR imaging. MR Clin North Am 2001;9:289–94. [36] Brasch R, Turetschek K. MRI characterization of tumors and grading angiogenesis using macromolecular contrast media: status report. Eur J Radiol 2000;34:148–55. [37] Rao AB, Koeller KK, Adair CF. Paragangliomas of the head and neck: radiologic–pathologic correlation. Radiographics 2002;19:1605–32. [38] Schick B, Kahle G. Radiological findings in angiofibroma. Acta Radiol 2000;41:585–93.