Original Investigations
The Application of Dynamic Contrast-Enhanced MRI and Diffusion-Weighted MRI in Patients with Maxillofacial Tumors Erina Kitamoto, DDS, Toru Chikui, DDS, PhD, Shintaro Kawano, DDS, PhD, Masahiro Ohga, RT, Kouji Kobayashi, RT, Yoshio Matsuo, MD, Takashi Yoshiura, MD, PhD, Makoto Obara, PhD, Hiroshi Honda, MD, PhD, Kazunori Yoshiura, DDS, PhD Rationale and Objectives: To elucidate the characteristics of four types of tumors, including squamous cell carcinoma (SCC), malignant lymphoma (ML), malignant salivary gland tumors (MSGTs), and pleomorphic adenoma (Pleo), in the maxillofacial region using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and diffusion-weighted MRI (DW-MRI）data. Materials and Methods: A total of 59 tumors were included in this research. DCE-MRI and DW-MRI were performed. We applied the Tofts and Kermode model (TK model) for the DCE-MRI data and obtained three dependent parameters: the influx forward volume transfer constant into the extravascular extracellular space from the plasma (Ktrans), the fractional volume of extravascular extracellular space per unit volume of tissue (ve), and the fractional volume of plasma (vp). Results: Among the Ktrans values, there were no significant differences between the three types of malignant tumors; however, there was a significant difference between the SCC and Pleo (P = .0099). The ve values of the Pleo were highest, with significant differences compared to the other categories (SCC, P = .0012; ML, P = .0017; and MSGT, P = .041). The ML had the lowest ve values, and there were significant differences between ML and the other two types of malignant tumors (SCC, P = .0278 and MSGT, P = .0062). In 14 (24%) cases, apparent diffusion coefficient (ADC) could not be measured because of poor image quality. The ADC values of the ML were lowest, whereas those of Pleo were highest, similar to that observed for ve. Conclusions: The Pleo tumors had lower Ktrans values and higher ve values, which are useful for differentiating them from the malignant tumors. Moreover, the ve was also useful for establishing a diagnosis of ML. Key Words: DCE-MRI; TIC; ADC; pharmacokinetic analyses; TK model; head and neck tumor. ªAUR, 2014
S
quamous cell carcinoma (SCC), malignant lymphoma (ML), and salivary gland tumors (SGTs) are commonly found in the maxillofacial region. The treatment choice for SCC and SGT is complete surgical removal, whereas benign SGTs are sometimes treated with long-term follow-up. Meanwhile, ML is treated with chemo-
Acad Radiol 2014; -:1–7 From the Faculty of Dental Science, Department of Oral and Maxillofacial Radiology, Kyushu University, Maidashi 3-1-1, Higashi-ku, Fukuoka 8128582, Japan (E.K., T.C., K.Y.); Faculty of Dental Science, Section of Oral and Maxillofacial Oncology, Division of Maxillofacial Diagnostic and Surgical Sciences, Kyushu University, Fukuoka, Japan (S.K.); Department of Medical Technology, Kyushu University Hospital, Fukuoka, Japan (M.O., K.K.); Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan (Y.M., T.Y., H.H.); and Philips Electronics Japan, Ltd 2-13-37, Tokyo, Japan (M.O.). Received March 2, 2014; accepted August 25, 2014. Address correspondence to: E.K. e-mail: [email protected] ªAUR, 2014 http://dx.doi.org/10.1016/j.acra.2014.08.016
therapy, radiotherapy, or a combination of these modalities. Therefore, it is important to differentiate ML from other tumors as soon as possible in the early stage (1,2). Magnetic resonance imaging (MRI) is effective for diagnosing tumors and has some advantages over traditional techniques, especially in identifying soft tissue lesions (3,4). The validity of dynamic contrast-enhanced MRI (DCE-MRI) has previously been established in SGT (5–10). The most conventional assessment with DCE-MRI involves the use of characteristics of the time-intensity curves (TICs) of regions of interest (ROIs). The simple quantification of TIC, such as the use of parameters including the time to peak enhancement (Tpeak) and washout ratio (WR), is commonly used (5,7–10). The Tpeak is considered to represent the microvessel count, and the threshold between benign and malignant tumors has been described in previous articles to be approximately 120 seconds (5). The WR represents the difference in the concentration of contrast medium (CM) between the intravascular (arterial) and extravascular 1
Academic Radiology, Vol -, No -, - 2014
KITAMOTO ET AL
(equilibrium) phases; therefore, tumors with high cellularity exhibit higher WR values. However, the WR depends on the scan parameters (5,8). In SGT, the presence of the rapid increase–rapid decrease pattern suggests a high possibility of Warthin’s tumor, which is associated with a high blood flow and a small amount of extravascular extracellular space (EES). On the contrary, the pattern of a monotonic or slow increase with a plateau with both longer Tpeak and lower WR suggests a high possibility of pleomorphic adenoma (Pleo), which results from a poor blood flow with abundant stromal tissue. A short Tpeak with a low WR signifies a malignant tumor (5,7–10). Pharmacokinetic analyses of DCE-MRI can be used to obtain quantitative information regarding tissue permeability and are applied in many tissues, such as the brain (11), prostate (12), and breast (13). However, reports of pharmacokinetic analyses of the maxillofacial region are scarce (14–20). Diffusion-weighted MRI (DW-MRI) is an imaging technique that can be used to visualize the Brownian motion of molecules, causing incoherent phase shifts that result in signal attention. DW-MRI can also be successfully applied in the differential diagnosis of SGT (9,21–23). In the present study, we attempted to elucidate the characteristics of four types of tumors, including SCC, ML, malignant salivary gland tumors (MSGTs), and Pleo, in the maxillofacial region using DCE-MRI and DW-MRI data. The first aim was to elucidate the validity of the quantitative analysis for differentiating benign neoplasms (Pleo) from malignant tumors. The second was to elucidate its validity for differentiating among three types of malignant tumors. MATERIALS AND METHODS Study Population
We investigated patients with untreated primary head and neck tumors who were referred to the Department of Oral Surgery and underwent DCE-MRI at our institute between April 2009 and October 2013. The institutional review board approved this study, and the requirement for informed consent was waived. Of the patients, one was omitted from this study because of poor images; therefore, 55 patients were enrolled in this study. The characteristics of the patients (n = 55; 26 male and 29 female; average age, 64 years; age range, 22–91) and diseases (n = 59) are described in Table 1. A total of 32 lesions were histopathologically diagnosed as SCC (23 tongue, five oral floor, and four buccal mucosa region), 14 were diagnosed as ML (two tongue, three buccal mucosa, two maxilla, five submandibular, one parotid space, and one neck), two were diagnosed as adenoid cystic carcinoma (one maxilla and one mandible), four were diagnosed as mucoepidermoid carcinoma (two oral floor and two maxilla), and seven were diagnosed as Pleo (three submandibular gland, one parotid gland, one maxilla, one parapharyngeal space, and one lip). 2
TABLE 1. Characteristics of the Patients (n = 55) and Lesions (n = 59) Sex (n = 55) Primary site (n = 59)
Diagnosis (n = 59)
SCC T stage (n = 32)
Male Female Tongue Maxilla Mandible Buccal mucosa Oral floor Submandibular gland Parotid gland area Parapharyngeal space Neck SCC ML Adenoid cystic carcinoma Mucoepidermoid carcinoma Pleomorphic adenoma T4 T3 T2
26 29 25 6 6 7 7 4 2 1 1 32 14 2 4 7 15 10 7
ML, malignant lymphoma; SCC, squamous cell carcinoma.
MRI Protocol
All experiments were performed using the Achieva 1.5T Nova Dual MR device (Philips Healthcare, Eindhoven, The Netherlands) with a maximum gradient strength of 66 mT/m. We used the Sense Neurovascular Coil (SENSE-NV-16) for all experiments. A reference scan was acquired immediately before the DCE-MRI scan to confirm that the scan range covered the tumor. The scan parameters were identical to those of the DCE scan, except for the flip angle (FA), which was 5 . The axial DCE scan consisted of a 10 slice, 3D–T1 fastfield echo sequence (time of repetition (TR)/time to echo (TE), 6.1/4.6 milliseconds; 15 FA; field of view, 220 mm; rectangular field of view (RFOV), 87.5%; slice thickness, 5 mm; SENSE factor, 1.8; acquisition voxel (millimeters), 1.72 1.82 5.00; and reconstruction voxel (millimeters), 0.86 0.86 5.00. Thirty seconds after the start of data acquisition, the operator initiated the injection of CM. We obtained 80 phases with a temporal resolution of 3.5 seconds; therefore, the total duration of imaging was 4 minutes and 40 seconds. A single injection of gadopentetate dimeglumine (Magnevist; Bayer Schering Pharma, Osaka, Japan) was administered at a dose of 0.1 mmol/kg through the antecubital vein at a rate of 2 mL/s via a mechanical injector, immediately followed by a 20-mL saline flush. Before performing DCE-MRI, T2-weighted and T1-weighted (T1w) axial images, coronal T2-weighted images, sagittal T1w images, and axial DW images were acquired. For the DW images, single-shot spin-echo echo-planar imaging sequences were used with a TR of 4716 milliseconds, a TE of 79 milliseconds, spectral presaturation with inversion
Academic Radiology, Vol -, No -, - 2014
DCE-MRI AND DW-MRI IN MAXILLOFACIAL TUMORS
TABLE 2. Results of the TK Model Analysis
Diagnosis Pleo Malignant tumor P value
Ktrans
ve
vp
Kep
AUGC
Mean SD
Mean SD
Mean SD
Mean SD
Mean SD
0.064 0.014 0.095 0.032 .0089
0.354 0.060 0.194 0.067
The Application of Dynamic Contrast-Enhanced MRI and Diffusion-Weighted MRI in Patients with Maxillofacial Tumors Erina Kitamoto, DDS, Toru Chikui, DDS, PhD, Shintaro Kawano, DDS, PhD, Masahiro Ohga, RT, Kouji Kobayashi, RT, Yoshio Matsuo, MD, Takashi Yoshiura, MD, PhD, Makoto Obara, PhD, Hiroshi Honda, MD, PhD, Kazunori Yoshiura, DDS, PhD Rationale and Objectives: To elucidate the characteristics of four types of tumors, including squamous cell carcinoma (SCC), malignant lymphoma (ML), malignant salivary gland tumors (MSGTs), and pleomorphic adenoma (Pleo), in the maxillofacial region using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and diffusion-weighted MRI (DW-MRI）data. Materials and Methods: A total of 59 tumors were included in this research. DCE-MRI and DW-MRI were performed. We applied the Tofts and Kermode model (TK model) for the DCE-MRI data and obtained three dependent parameters: the influx forward volume transfer constant into the extravascular extracellular space from the plasma (Ktrans), the fractional volume of extravascular extracellular space per unit volume of tissue (ve), and the fractional volume of plasma (vp). Results: Among the Ktrans values, there were no significant differences between the three types of malignant tumors; however, there was a significant difference between the SCC and Pleo (P = .0099). The ve values of the Pleo were highest, with significant differences compared to the other categories (SCC, P = .0012; ML, P = .0017; and MSGT, P = .041). The ML had the lowest ve values, and there were significant differences between ML and the other two types of malignant tumors (SCC, P = .0278 and MSGT, P = .0062). In 14 (24%) cases, apparent diffusion coefficient (ADC) could not be measured because of poor image quality. The ADC values of the ML were lowest, whereas those of Pleo were highest, similar to that observed for ve. Conclusions: The Pleo tumors had lower Ktrans values and higher ve values, which are useful for differentiating them from the malignant tumors. Moreover, the ve was also useful for establishing a diagnosis of ML. Key Words: DCE-MRI; TIC; ADC; pharmacokinetic analyses; TK model; head and neck tumor. ªAUR, 2014
S
quamous cell carcinoma (SCC), malignant lymphoma (ML), and salivary gland tumors (SGTs) are commonly found in the maxillofacial region. The treatment choice for SCC and SGT is complete surgical removal, whereas benign SGTs are sometimes treated with long-term follow-up. Meanwhile, ML is treated with chemo-
Acad Radiol 2014; -:1–7 From the Faculty of Dental Science, Department of Oral and Maxillofacial Radiology, Kyushu University, Maidashi 3-1-1, Higashi-ku, Fukuoka 8128582, Japan (E.K., T.C., K.Y.); Faculty of Dental Science, Section of Oral and Maxillofacial Oncology, Division of Maxillofacial Diagnostic and Surgical Sciences, Kyushu University, Fukuoka, Japan (S.K.); Department of Medical Technology, Kyushu University Hospital, Fukuoka, Japan (M.O., K.K.); Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan (Y.M., T.Y., H.H.); and Philips Electronics Japan, Ltd 2-13-37, Tokyo, Japan (M.O.). Received March 2, 2014; accepted August 25, 2014. Address correspondence to: E.K. e-mail: [email protected] ªAUR, 2014 http://dx.doi.org/10.1016/j.acra.2014.08.016
therapy, radiotherapy, or a combination of these modalities. Therefore, it is important to differentiate ML from other tumors as soon as possible in the early stage (1,2). Magnetic resonance imaging (MRI) is effective for diagnosing tumors and has some advantages over traditional techniques, especially in identifying soft tissue lesions (3,4). The validity of dynamic contrast-enhanced MRI (DCE-MRI) has previously been established in SGT (5–10). The most conventional assessment with DCE-MRI involves the use of characteristics of the time-intensity curves (TICs) of regions of interest (ROIs). The simple quantification of TIC, such as the use of parameters including the time to peak enhancement (Tpeak) and washout ratio (WR), is commonly used (5,7–10). The Tpeak is considered to represent the microvessel count, and the threshold between benign and malignant tumors has been described in previous articles to be approximately 120 seconds (5). The WR represents the difference in the concentration of contrast medium (CM) between the intravascular (arterial) and extravascular 1
Academic Radiology, Vol -, No -, - 2014
KITAMOTO ET AL
(equilibrium) phases; therefore, tumors with high cellularity exhibit higher WR values. However, the WR depends on the scan parameters (5,8). In SGT, the presence of the rapid increase–rapid decrease pattern suggests a high possibility of Warthin’s tumor, which is associated with a high blood flow and a small amount of extravascular extracellular space (EES). On the contrary, the pattern of a monotonic or slow increase with a plateau with both longer Tpeak and lower WR suggests a high possibility of pleomorphic adenoma (Pleo), which results from a poor blood flow with abundant stromal tissue. A short Tpeak with a low WR signifies a malignant tumor (5,7–10). Pharmacokinetic analyses of DCE-MRI can be used to obtain quantitative information regarding tissue permeability and are applied in many tissues, such as the brain (11), prostate (12), and breast (13). However, reports of pharmacokinetic analyses of the maxillofacial region are scarce (14–20). Diffusion-weighted MRI (DW-MRI) is an imaging technique that can be used to visualize the Brownian motion of molecules, causing incoherent phase shifts that result in signal attention. DW-MRI can also be successfully applied in the differential diagnosis of SGT (9,21–23). In the present study, we attempted to elucidate the characteristics of four types of tumors, including SCC, ML, malignant salivary gland tumors (MSGTs), and Pleo, in the maxillofacial region using DCE-MRI and DW-MRI data. The first aim was to elucidate the validity of the quantitative analysis for differentiating benign neoplasms (Pleo) from malignant tumors. The second was to elucidate its validity for differentiating among three types of malignant tumors. MATERIALS AND METHODS Study Population
We investigated patients with untreated primary head and neck tumors who were referred to the Department of Oral Surgery and underwent DCE-MRI at our institute between April 2009 and October 2013. The institutional review board approved this study, and the requirement for informed consent was waived. Of the patients, one was omitted from this study because of poor images; therefore, 55 patients were enrolled in this study. The characteristics of the patients (n = 55; 26 male and 29 female; average age, 64 years; age range, 22–91) and diseases (n = 59) are described in Table 1. A total of 32 lesions were histopathologically diagnosed as SCC (23 tongue, five oral floor, and four buccal mucosa region), 14 were diagnosed as ML (two tongue, three buccal mucosa, two maxilla, five submandibular, one parotid space, and one neck), two were diagnosed as adenoid cystic carcinoma (one maxilla and one mandible), four were diagnosed as mucoepidermoid carcinoma (two oral floor and two maxilla), and seven were diagnosed as Pleo (three submandibular gland, one parotid gland, one maxilla, one parapharyngeal space, and one lip). 2
TABLE 1. Characteristics of the Patients (n = 55) and Lesions (n = 59) Sex (n = 55) Primary site (n = 59)
Diagnosis (n = 59)
SCC T stage (n = 32)
Male Female Tongue Maxilla Mandible Buccal mucosa Oral floor Submandibular gland Parotid gland area Parapharyngeal space Neck SCC ML Adenoid cystic carcinoma Mucoepidermoid carcinoma Pleomorphic adenoma T4 T3 T2
26 29 25 6 6 7 7 4 2 1 1 32 14 2 4 7 15 10 7
ML, malignant lymphoma; SCC, squamous cell carcinoma.
MRI Protocol
All experiments were performed using the Achieva 1.5T Nova Dual MR device (Philips Healthcare, Eindhoven, The Netherlands) with a maximum gradient strength of 66 mT/m. We used the Sense Neurovascular Coil (SENSE-NV-16) for all experiments. A reference scan was acquired immediately before the DCE-MRI scan to confirm that the scan range covered the tumor. The scan parameters were identical to those of the DCE scan, except for the flip angle (FA), which was 5 . The axial DCE scan consisted of a 10 slice, 3D–T1 fastfield echo sequence (time of repetition (TR)/time to echo (TE), 6.1/4.6 milliseconds; 15 FA; field of view, 220 mm; rectangular field of view (RFOV), 87.5%; slice thickness, 5 mm; SENSE factor, 1.8; acquisition voxel (millimeters), 1.72 1.82 5.00; and reconstruction voxel (millimeters), 0.86 0.86 5.00. Thirty seconds after the start of data acquisition, the operator initiated the injection of CM. We obtained 80 phases with a temporal resolution of 3.5 seconds; therefore, the total duration of imaging was 4 minutes and 40 seconds. A single injection of gadopentetate dimeglumine (Magnevist; Bayer Schering Pharma, Osaka, Japan) was administered at a dose of 0.1 mmol/kg through the antecubital vein at a rate of 2 mL/s via a mechanical injector, immediately followed by a 20-mL saline flush. Before performing DCE-MRI, T2-weighted and T1-weighted (T1w) axial images, coronal T2-weighted images, sagittal T1w images, and axial DW images were acquired. For the DW images, single-shot spin-echo echo-planar imaging sequences were used with a TR of 4716 milliseconds, a TE of 79 milliseconds, spectral presaturation with inversion
Academic Radiology, Vol -, No -, - 2014
DCE-MRI AND DW-MRI IN MAXILLOFACIAL TUMORS
TABLE 2. Results of the TK Model Analysis
Diagnosis Pleo Malignant tumor P value
Ktrans
ve
vp
Kep
AUGC
Mean SD
Mean SD
Mean SD
Mean SD
Mean SD
0.064 0.014 0.095 0.032 .0089
0.354 0.060 0.194 0.067
