- Email: [email protected]
Role of dynamic MRI in the evaluation of head and neck cancers treated with radiation therapy
Int. J. Radiation Oncology
PII SO360-3016(96)00595-O
ELSEVIER
l
Biol. Phys., Vol. 37, No. 4, pp. 783-787, 1997 Copyright 8 1997 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/97 $17.00 + .OO
Clinical Investigation ROLE OF DYNAMIC MRI IN THE EVALUATION OF HEAD AND NECK CANCERS TREATED WITH RADIATION THERAPY M.D.,* MITSUHIRO FURUSAWA, M.D.,* RYUJI MURAKAMI, M.D.,* TOSHIMI YOKOYAMA, M.D.,* YUJI SAKAMOTO, M.D.,* RYUICHI NISHIMURA, M.D.,* YASUYUKI YAMASHITA, M.D.,* MUTSUMASA TAKAHASHI, M.D.* AND TAKERU ISHIKAWA, M.D.+ YUJI BABA,
Departments of *Radiology and ‘Otolaryngology, Kumamoto University School of Medicine, Kumamoto, Japan
Purpose: To study the usefulness of dynamic magnetic resonance imaging in the evaluation of head and neck cancers treated with radiation therapy. Methods and Materials: Seventy-six patients (58 males and 18 females; ages 20-82) with head and neck cancers (10 nasopharyngeai carcinomas, 22 mesopharyngeal carcinomas, 10 hypopharyngeal carcinomas, 16 oral cavity carcinomas, 11 lingual carcinomas, and 7 laryngeal carcinomas) were treated by radiation therapy combined with concomitant low-dose cispIatinum. Magnetic resonance imaging (MRI) was performed before and 2 weeks after the irradiation in all cases.After bolus administration of gadopentetate dhneghtmine (GCDTPA) (0.1 mmoY kg), images were obtained every 30 s (repetition time 200 ms, echo time 16 ms) using a 1.5 or 0.5-T superconductive unit. Biopsy or surgery was performed after radiation therapy and the histologic findings were correlated with the MRI findings (T,, T,, dynamic, and enhanced T1). Results: Complete remission, partial response, and no response were obtained in 18,36, and 7 cases,respectively. eic MRI correctly diagnosed 17 of the 18 complete remission cases,33 of the 36 partial response cases,and all of the 7 no-response cases.The accuracy of dynamic MRI, T1-weighted image, T2-weighted image, and Gdenhanced T1-weighted image was 94.4%) 68%) 82%, and 86%) respectively. Conclusion: Dynamic MRI proved to be useful in the evaluation of the radiation therapy of head and neck cancers. 0 1997 Elsevier ScienceInc. Dynamic MRI, Head and neck cancers, Radiation therapy.
Seventy-six patients (58 males and 18 females; ages 2082) with head and neck cancers (10 nasopharyngeal carci-
nomas, 22 mesopharyngeal carcinomas, 10 hypopharyngeal carcinomas, 16 oral cavity carcinomas, 11 lingual carcinomas, and 7 laryngeal carcinomas) were treated by radiation therapy combined with concomitant low-dose cisplatinum (CDDP). A total dose of 39.6-70.2 Gy (1.8 Gy, 5 times/ week) was given by 3-MV X ray. Five mg/body of CDDP was administered intravenously (i.v.) within 2 h from irradiation, l-3 times/week throughout radiation therapy. In all patients, MRI was performed 2 weeks after irradiation. Spin-echo images, T,-weighted (TiWI) (repetition time in milliseconds per echo time in milliseconds: 600/ 15) and T1-weighted (T*WI, 200/90), were obtained using a 1.5-T (Magnetom 15T; Siemens, Erlangen, Germany) or a 0.5-T (Gyroscan T5; Philips Medical Systems International, Eindhoven, The Netherlands) superconductive magnet. After a bolus injection of Gd-DTPA (0.1 mmoY kg), dynamic images (TiWI spin-echo images: 200/16) of a 192 x 256 matrix were obtained every 30 s for 5 min. Conventional enhanced T,WI (600/15) were obtained immediately after the dynamic series was completed.
Reprint requests to: Yuji Baba, M.D., Department of Radiology, Kumamoto Univ. School of Medicine, Honjo l-l-l, Ku-
mamoto-city 860, Japan. Accepted for publication 6 December 1996.
INTRODUCTION
In the diagnosis of head and neck tumors, the role of magnetic resonance imaging (MRI) has been established (l-7). To improve tumor detection, Gd-DTPA enhancement with a fatsuppression technique (8-13) and dynamic MRI were recently iutroduced and their usefulness has been reported (14-16). In the evaluation of radiation therapy of such tumors, decreased tumor volume often makes tumor detection difficult. Associated fibrotic changes will influence the signal intensity of tumor and surrounding normal tissue and their enhancement by Gd-DTPA. Note that fibrotic tissue has previously shown delayed enhancement in dynamic MRI (15). In this study, we reviewed the findings of dynamic MRI after radiation therapy of head and neck cancers and compared the usefulness of dynamic and static MRI pulse sequences. METHODS
AND MATERIALS
783
184 Slgnal
I. J. RadiationOncology0 Biology 0 Physics intensity
determined as MRI tumor size (area). The postradiotherapy MRI tumor size was compared to pretreatment MRI tumor size (determined by enhanced T,WI) and the results of radiotherapy evaluated by MRI were also graded (MRI grading) as CR, PR, NR, and PD. The accuracy of each pulse sequence (TrWI, T,WI, dynamic, enhanced T,WI) was calculated (number of correct MRI diagnosis at each pusle sequence per number of total cases) and compared.
I Noise
t
tumor
t
normal
Volume 37, Number 4. 1997
RESULTS 01
pre
30
60
90
Time after
120
150
Gd-DTPA
180
210
injection
240
270
(8)
(4 Signal intensity
I Noise
at
35
20
In Fig. 1, the enhancement patterns of tumor and normal tissue after injection of Gd-DTPA is shown. The signalto-noise ratio (signal intensity/noise) of the tumor peaked at 60 or 90 s, while that of the normal tissue increased slowly. The tumor to normal tissue contrast also peaked
1
t
tumor
t
normal
meaniSD n=17
7
o?pre ,30 , 60 , 90 ,120 ,150 1180 ‘210 ,240 ,270 , Time after Gd-DTPA
injectlon
(8)
60 or 90 s.
In the clinical evaluation, CR, PR, and NR were obtained in 29, 40, and 7 cases, respectively. The clinical evaluations and MRI evaluations are shown in Table 1. The diagnostic accuracy of TrWI, T2WI, dynamic images, and enhanced TrWI was 69.7%, 76.3%, 93.4%, and 80.2%, respectively (Table 1). There was a significant difference between the diagnostic accuracy of dynamic MRI and all other MRI sequences (p < 0.05, chi-square test). In Fig. 2, supraglottic carcinoma (T2NOMO) is shown before irradiation; the tumor is demonstrated in all images. In dynamic MRI, early enhancement of the tumor was visualized. After irradiation at 40 Gy (Fig. 3), tumor is not detected in T,WI, T,WI, and postenhanced TrWI. In dy-
(W Fig. 1. Time-intensitycurve of tumor and normal tissueafter injection of Cd-DTPA. Both in 1.5-T unit images (a) and 0.5-T unit images (b), the signal-to-noiseratio (signal inteasity/noise)of the tumor peakedat 60 or 90 s, while that of the normal tissueincreasedslowly. The tumor to normal tissue contrastalso Peakedat 60 or 90 s.
In all 76 patients, biopsy (41 patients) or surgery (35 patients) was performed after radiation therapy and the macroscopic and histologic findings were correlated with the MRI (T,WI, T,WI, dynamic, and enhanced T,WI) findings. The results of radiotherapy were clinically evaluated by a otolaryngologist and graded as complete remission (CR), proven by biospy or surgery; partial remission (PR), ~50% reduction to the tumor; ~50% reduction of the tumor (NR); and progression of disease (PD). In the evaluation of radiation effects in dynamic MRI, early enhancement (30-90-s image) of the region where tumor initially existed was diagnosed as residual tumor. Delayed enhancement (120 s image or later) was diagnosed as normal tissue or fibrotic changes. All MRI images were retrospectively analyzed with discussion by two radiologists without knowledge of clinical radiation effect. The size of the tumor in MRI image was measured in two dimensions and the maximum data were
Table 1. Clinical and MRI evaluations MRI evaluation Clinical evaluation T,WI Diagnostic accuracy 53/76 CR PR NC T,WI Diagnostic accuracy 58/76 CR PR NC Dynamic images Diagnostic accuracy 71/76 CR PR NC Enhanced T,WI Diagnostic accuracy 61/76 CR PR NC
CR
PR
NR
22 16 0
7 24 0
0 0 7
22 11 0
7 29 0
0 0 7
27 3 0
2 37 0
0 0 7
21 7 0
8 33 0
0 0 7
(69.7%)
(76.3%)
(93.4%)
(80.2%)
There was a significant difference between the diagnostic accuracy of dynamic MRI and all other pulse sequences(p < 0.05).
Dynamic MRI in head and neck cancer
(4
0 Y. BABA
et al.
(b)
(4 Fig. 2. Supraglottic carcinoma T2NOMO before radiation therapy. Supraglottic tumor is demonstrated on TIWl (a), T,Wl (b), postcontrast T,WI (c), and dynamic images (d). Early enhancement of the tumor is shown on dynamic images.
namic MRI, early enhancement was still visualized and residual tumor was proved by biopsy. Dynamic MRI failed to make correct evaluation in five cases. In two (both hypopharyngeal carcinoma), early enhancement of surrounding inflammatory changes was diagnosed as residual tumor. In a very small residual mesopharyngeal tumor, no early enhancement was detected. In another mesopharyngeal carcinoma, early enhancement was not detected because of prolonged injection (technical failure of bolus injection) of Gd-DTPA. In a case of lingual carcinoma, associated leukoplakia showed early enhancement and was misdiagnosed as residual carcinoma.
DISCUSSION Often, MRI is useful in the detection and delineation of head and neck cancers (1-16). Both TIWI and T2WI provide useful information. Usually, good tumor-to-normaltissue contrast is obtained on T,WI, but in small tumors, detection becomes difficult because of lower spatial resolution of T2WI (due to motion artifacts and lower signals). The value of contrast enhancement for the qualita-
tive evaluation of neoplasms in the head and neck has been well documented (5-9). However, in some cases, lower contrast enhancement is seen in head and neck tumors compared with the enhancement of lesions in the central nervous system (13). To increase the tumor-to-normal-tissue signal contrast, dynamic MRIs or enhancement with Gd-DTPA combined with fat-suppression techniques have been useful (13-16). After successful radiation therapy, tumors are remarkably reduced in size and sometimes replaced by the fibrotic tissue. Good tumor-to-normal-tissue contrast in dynamic images, produced by early enhancement of tumors, can identify small residual tumors. In a recent report, the detectability of head and neck tumor in TIWI, T2WI, enhanced TIWI, and dynamic images was 67%, 83%, 87%, and 93%, respectively (16). In our study, the diagnostic accuracy of T,WI and postenhanced TIWI was lower (76 and 80%) compared to the previous report. In contrast, the diagnostic accuracy of dynamic images remained almost same. In addition to high tumor-to-normal-tissue contrast, dynamic MRI can provide good tissue characterization. In dynamic MRI, head and neck tumors are usually enhanced in
I. J. Radiation Oncology 0 Biology 0 Physics
Volume 37, Number 4, 1997
(4
(4 Fig. 3. Same case as in Fig. 2 (supraglottic carcinoma T2NOMO) after irradiation of 40 Gy. Supraglottic tumor is remarkably reduced in size and not demonstrated on TrWI (a), TzWI (b), or postenhanced T,WI (c). Early enhancement is still demonstrated on dynamic images (d).
the early phase like other squamous cell carcinomas (15 17). In contrast, the fibrotic tissue shows enhancement slowly in the late phase (16). Early enhancement in dynamic images can demonstrate small residual tumors even within the fibrotic tissue, but it is difficult to separate tumors from slowly enhancing fibrotic changes in postcontrast T,WI. Good tumor-to-normal-tissue contrast and the difference of enhancement pattern between tumor, normal tissue, and fibrotic changes all contribute to the high diagnostic accuracy of dynamic MRI after radiotherapy. Inflammation of the normal tissue may be present even 2 weeks after radiation in some patients and show high signal intensity in T2WI. Inflammatory changes also show early enhancement in dynamic images and may mask the
tumor enhancement. It is difficult to visualize small tumors within inflammatory changes, and performance of dynamic MRI after inflammation diminishes is strongly recommended. In our hospital, a local tumor is operatively resected after irradiation of 40 Gy except remarkable tumor remission (complete remission, or 90% decrease of the tumor) was obtained. After a full dose (60-70 Gy) of irradiation, salvage surgery is performed if complete remission of the tumor is not obtained. It is important to evaluate the results of radiation; such evaluation may change the therapy. The result of our study suggests that dynamic MRI is essential to evaluate head and neck tumors after radiotherapy.
REFERENCES 1. Vogl, T.; Dresel, S.; Bilaniuk, L. T.; Crevers, G.; Kang, K.; Lissner, J. Tumors of the nasophaqnx and adjacent areas: MR imaging with Gd-DTPA. AJR 1.54585-592; 1990.
2. Ditlon, W. P. The pharynx and oral cavity. In: Som, P. M.; Bergeron, R. T., eds. Head and neck imaging, 2nd. St. Louis, MO: Mosby-Yearbook; 1991:407-466.
Dynamic
MRI
in
head and neck cancer
3. Tabor, E. K.; Curtin, H. D. MRI of the salivary glands. Radiol. Clin. North. Am. 27:379-392; 1989. 4. Teresi, L. M.; Lufkin, R. B.; Wortham, D. G.; Abemayor, E.; Hanafee, W. N. Parotid masses: MR imaging. Radiology 163:405-409; 1987. 5. Mandelblatt, S. M.; Braun, I. F.; Davis, P. C.; Fry, S. M.; Jacobs, L. H.; Hoffman, J. C. Parotid masses: MR imaging. Radiology 163:411414; 1987. 6. Robinson, J. D.; Crawford, S. C.; Teresi, L. M.; Schiller, V. L.; Lufkin, R. B.; Harnsberger, H. R.; Dietrich, R. B.; Crim, J. R.; Duckwiler, G. R.; Spickler, E. M.; Hanafee, W. N. Extracranial lesions of the head and neck: preliminary experience with Gd-DTPA-enhanced MR imaging. Radiology 172:165-170; 1989. 7. Vogl, T.; Dresel, S.; Bilaniuk, L. T.; Grevers, G.; Kang, K.; Lissner, J. Tumors of the head and neck and adjacent areas: MR imaging with Gd-DTPA. American Journal of Neuroradiology 11:187-194; 1990. 8. Crawford, S. C.; Hamsberger, H. R.; Lufkin, R. B.; Hanafee, W. N. The role of gadolinium-DTPA in the evaluation of extracranial head and neck mass lesions. Radiol. Clin. North. Am. 27:219-242; 1989. 9. Tien, R. D.; Hesselink, J. F.; Chu, P. K.; Szumowski, J. Improved detection and delineation of head and neck lesions With fat-suppression spin-echo MR imaging. American Journal of Neuroradiology 12: 19-24; 1991. 0. Barakos, J. A.; Dillion, W. P.; Chew, W. M. Orbit, skull base, and pharynx: contrast-enhanced fat-suppression MR imaging. Radiology 179:191-198; 1991.
0 Y. BABA
et
al.
787
11. Zoarski, G. H.; Mackey, J. K.; Anzai, Y.; Hanafee, W. N.; Melki, P. S.; Mulkern, R. V.; Jolesz, F. A.; Lull&, R. B. Head and neck: Initial clinical experience with fast spin-echo MR imaging. Radiology 188:323-327; 1993. 12. Panush, D.; Fulbright, R.; Sze, G.; Smith, R. C.; Constable, R. T. Inversion-recovery fast spin-echo MR imaging: Efficacy in the evaluation of head and neck lesions. Radiology 187:421426; 1993. 13. Ross, M. R.; Schemer, D. F.; Phyliss, C.; Enzmann, D. R. MR imaging of head and neck tumors: Comparison of T,weighted contrast-enhanced fat-suppressed images with conventional T,-weighted and fast spin-echo T,-weighted images. AJR 163:173-178; 1994. 14. Arakawa, A.; Tsuruta, J.; Nishimura, R.; Sakamoto, Y.; Korogi, Y.; Baba, Y.; Furusawa, M.; Ishimaru, Y.; Uji, Y.; Taen, A.; Ishikawa, T.; Takahashi, M. MR imaging of linguinal carcinoma: comparison with surgical staging. Radiation Medicine 14:25-29; 1996. 15. Arakawa, A.; Tsuruta, R.; Nishimura, R.; Sakamoto, Y.; Korogi, Y.; Baba, Y.; Furusawa, M.; Ishimaru, Y.; Uji, Y.; Taen, A.; Ishikawa, T.; Takahashi, M. Lingual carcinoma: Correlation of MR imaging with histopathologic findings. Acta Radiologica 37:700-707; 1996. 16. Takashima, S.; Noguchi, Y.; Okumura, T.; Aruga, H.; Kobayashi, T. Dynamic MR imaging in the head and neck. Radiology 189:813-821; 1993. 17. Yamashita, Y.; Takahashi, M.; Sawada, T.; Miyazaki, K.; Okamura, H. Carcinoma of the cervix: dynamic MR imaging. Radiology 182:643-648; 1992.
PII SO360-3016(96)00595-O
ELSEVIER
l
Biol. Phys., Vol. 37, No. 4, pp. 783-787, 1997 Copyright 8 1997 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/97 $17.00 + .OO
Clinical Investigation ROLE OF DYNAMIC MRI IN THE EVALUATION OF HEAD AND NECK CANCERS TREATED WITH RADIATION THERAPY M.D.,* MITSUHIRO FURUSAWA, M.D.,* RYUJI MURAKAMI, M.D.,* TOSHIMI YOKOYAMA, M.D.,* YUJI SAKAMOTO, M.D.,* RYUICHI NISHIMURA, M.D.,* YASUYUKI YAMASHITA, M.D.,* MUTSUMASA TAKAHASHI, M.D.* AND TAKERU ISHIKAWA, M.D.+ YUJI BABA,
Departments of *Radiology and ‘Otolaryngology, Kumamoto University School of Medicine, Kumamoto, Japan
Purpose: To study the usefulness of dynamic magnetic resonance imaging in the evaluation of head and neck cancers treated with radiation therapy. Methods and Materials: Seventy-six patients (58 males and 18 females; ages 20-82) with head and neck cancers (10 nasopharyngeai carcinomas, 22 mesopharyngeal carcinomas, 10 hypopharyngeal carcinomas, 16 oral cavity carcinomas, 11 lingual carcinomas, and 7 laryngeal carcinomas) were treated by radiation therapy combined with concomitant low-dose cispIatinum. Magnetic resonance imaging (MRI) was performed before and 2 weeks after the irradiation in all cases.After bolus administration of gadopentetate dhneghtmine (GCDTPA) (0.1 mmoY kg), images were obtained every 30 s (repetition time 200 ms, echo time 16 ms) using a 1.5 or 0.5-T superconductive unit. Biopsy or surgery was performed after radiation therapy and the histologic findings were correlated with the MRI findings (T,, T,, dynamic, and enhanced T1). Results: Complete remission, partial response, and no response were obtained in 18,36, and 7 cases,respectively. eic MRI correctly diagnosed 17 of the 18 complete remission cases,33 of the 36 partial response cases,and all of the 7 no-response cases.The accuracy of dynamic MRI, T1-weighted image, T2-weighted image, and Gdenhanced T1-weighted image was 94.4%) 68%) 82%, and 86%) respectively. Conclusion: Dynamic MRI proved to be useful in the evaluation of the radiation therapy of head and neck cancers. 0 1997 Elsevier ScienceInc. Dynamic MRI, Head and neck cancers, Radiation therapy.
Seventy-six patients (58 males and 18 females; ages 2082) with head and neck cancers (10 nasopharyngeal carci-
nomas, 22 mesopharyngeal carcinomas, 10 hypopharyngeal carcinomas, 16 oral cavity carcinomas, 11 lingual carcinomas, and 7 laryngeal carcinomas) were treated by radiation therapy combined with concomitant low-dose cisplatinum (CDDP). A total dose of 39.6-70.2 Gy (1.8 Gy, 5 times/ week) was given by 3-MV X ray. Five mg/body of CDDP was administered intravenously (i.v.) within 2 h from irradiation, l-3 times/week throughout radiation therapy. In all patients, MRI was performed 2 weeks after irradiation. Spin-echo images, T,-weighted (TiWI) (repetition time in milliseconds per echo time in milliseconds: 600/ 15) and T1-weighted (T*WI, 200/90), were obtained using a 1.5-T (Magnetom 15T; Siemens, Erlangen, Germany) or a 0.5-T (Gyroscan T5; Philips Medical Systems International, Eindhoven, The Netherlands) superconductive magnet. After a bolus injection of Gd-DTPA (0.1 mmoY kg), dynamic images (TiWI spin-echo images: 200/16) of a 192 x 256 matrix were obtained every 30 s for 5 min. Conventional enhanced T,WI (600/15) were obtained immediately after the dynamic series was completed.
Reprint requests to: Yuji Baba, M.D., Department of Radiology, Kumamoto Univ. School of Medicine, Honjo l-l-l, Ku-
mamoto-city 860, Japan. Accepted for publication 6 December 1996.
INTRODUCTION
In the diagnosis of head and neck tumors, the role of magnetic resonance imaging (MRI) has been established (l-7). To improve tumor detection, Gd-DTPA enhancement with a fatsuppression technique (8-13) and dynamic MRI were recently iutroduced and their usefulness has been reported (14-16). In the evaluation of radiation therapy of such tumors, decreased tumor volume often makes tumor detection difficult. Associated fibrotic changes will influence the signal intensity of tumor and surrounding normal tissue and their enhancement by Gd-DTPA. Note that fibrotic tissue has previously shown delayed enhancement in dynamic MRI (15). In this study, we reviewed the findings of dynamic MRI after radiation therapy of head and neck cancers and compared the usefulness of dynamic and static MRI pulse sequences. METHODS
AND MATERIALS
783
184 Slgnal
I. J. RadiationOncology0 Biology 0 Physics intensity
determined as MRI tumor size (area). The postradiotherapy MRI tumor size was compared to pretreatment MRI tumor size (determined by enhanced T,WI) and the results of radiotherapy evaluated by MRI were also graded (MRI grading) as CR, PR, NR, and PD. The accuracy of each pulse sequence (TrWI, T,WI, dynamic, enhanced T,WI) was calculated (number of correct MRI diagnosis at each pusle sequence per number of total cases) and compared.
I Noise
t
tumor
t
normal
Volume 37, Number 4. 1997
RESULTS 01
pre
30
60
90
Time after
120
150
Gd-DTPA
180
210
injection
240
270
(8)
(4 Signal intensity
I Noise
at
35
20
In Fig. 1, the enhancement patterns of tumor and normal tissue after injection of Gd-DTPA is shown. The signalto-noise ratio (signal intensity/noise) of the tumor peaked at 60 or 90 s, while that of the normal tissue increased slowly. The tumor to normal tissue contrast also peaked
1
t
tumor
t
normal
meaniSD n=17
7
o?pre ,30 , 60 , 90 ,120 ,150 1180 ‘210 ,240 ,270 , Time after Gd-DTPA
injectlon
(8)
60 or 90 s.
In the clinical evaluation, CR, PR, and NR were obtained in 29, 40, and 7 cases, respectively. The clinical evaluations and MRI evaluations are shown in Table 1. The diagnostic accuracy of TrWI, T2WI, dynamic images, and enhanced TrWI was 69.7%, 76.3%, 93.4%, and 80.2%, respectively (Table 1). There was a significant difference between the diagnostic accuracy of dynamic MRI and all other MRI sequences (p < 0.05, chi-square test). In Fig. 2, supraglottic carcinoma (T2NOMO) is shown before irradiation; the tumor is demonstrated in all images. In dynamic MRI, early enhancement of the tumor was visualized. After irradiation at 40 Gy (Fig. 3), tumor is not detected in T,WI, T,WI, and postenhanced TrWI. In dy-
(W Fig. 1. Time-intensitycurve of tumor and normal tissueafter injection of Cd-DTPA. Both in 1.5-T unit images (a) and 0.5-T unit images (b), the signal-to-noiseratio (signal inteasity/noise)of the tumor peakedat 60 or 90 s, while that of the normal tissueincreasedslowly. The tumor to normal tissue contrastalso Peakedat 60 or 90 s.
In all 76 patients, biopsy (41 patients) or surgery (35 patients) was performed after radiation therapy and the macroscopic and histologic findings were correlated with the MRI (T,WI, T,WI, dynamic, and enhanced T,WI) findings. The results of radiotherapy were clinically evaluated by a otolaryngologist and graded as complete remission (CR), proven by biospy or surgery; partial remission (PR), ~50% reduction to the tumor; ~50% reduction of the tumor (NR); and progression of disease (PD). In the evaluation of radiation effects in dynamic MRI, early enhancement (30-90-s image) of the region where tumor initially existed was diagnosed as residual tumor. Delayed enhancement (120 s image or later) was diagnosed as normal tissue or fibrotic changes. All MRI images were retrospectively analyzed with discussion by two radiologists without knowledge of clinical radiation effect. The size of the tumor in MRI image was measured in two dimensions and the maximum data were
Table 1. Clinical and MRI evaluations MRI evaluation Clinical evaluation T,WI Diagnostic accuracy 53/76 CR PR NC T,WI Diagnostic accuracy 58/76 CR PR NC Dynamic images Diagnostic accuracy 71/76 CR PR NC Enhanced T,WI Diagnostic accuracy 61/76 CR PR NC
CR
PR
NR
22 16 0
7 24 0
0 0 7
22 11 0
7 29 0
0 0 7
27 3 0
2 37 0
0 0 7
21 7 0
8 33 0
0 0 7
(69.7%)
(76.3%)
(93.4%)
(80.2%)
There was a significant difference between the diagnostic accuracy of dynamic MRI and all other pulse sequences(p < 0.05).
Dynamic MRI in head and neck cancer
(4
0 Y. BABA
et al.
(b)
(4 Fig. 2. Supraglottic carcinoma T2NOMO before radiation therapy. Supraglottic tumor is demonstrated on TIWl (a), T,Wl (b), postcontrast T,WI (c), and dynamic images (d). Early enhancement of the tumor is shown on dynamic images.
namic MRI, early enhancement was still visualized and residual tumor was proved by biopsy. Dynamic MRI failed to make correct evaluation in five cases. In two (both hypopharyngeal carcinoma), early enhancement of surrounding inflammatory changes was diagnosed as residual tumor. In a very small residual mesopharyngeal tumor, no early enhancement was detected. In another mesopharyngeal carcinoma, early enhancement was not detected because of prolonged injection (technical failure of bolus injection) of Gd-DTPA. In a case of lingual carcinoma, associated leukoplakia showed early enhancement and was misdiagnosed as residual carcinoma.
DISCUSSION Often, MRI is useful in the detection and delineation of head and neck cancers (1-16). Both TIWI and T2WI provide useful information. Usually, good tumor-to-normaltissue contrast is obtained on T,WI, but in small tumors, detection becomes difficult because of lower spatial resolution of T2WI (due to motion artifacts and lower signals). The value of contrast enhancement for the qualita-
tive evaluation of neoplasms in the head and neck has been well documented (5-9). However, in some cases, lower contrast enhancement is seen in head and neck tumors compared with the enhancement of lesions in the central nervous system (13). To increase the tumor-to-normal-tissue signal contrast, dynamic MRIs or enhancement with Gd-DTPA combined with fat-suppression techniques have been useful (13-16). After successful radiation therapy, tumors are remarkably reduced in size and sometimes replaced by the fibrotic tissue. Good tumor-to-normal-tissue contrast in dynamic images, produced by early enhancement of tumors, can identify small residual tumors. In a recent report, the detectability of head and neck tumor in TIWI, T2WI, enhanced TIWI, and dynamic images was 67%, 83%, 87%, and 93%, respectively (16). In our study, the diagnostic accuracy of T,WI and postenhanced TIWI was lower (76 and 80%) compared to the previous report. In contrast, the diagnostic accuracy of dynamic images remained almost same. In addition to high tumor-to-normal-tissue contrast, dynamic MRI can provide good tissue characterization. In dynamic MRI, head and neck tumors are usually enhanced in
I. J. Radiation Oncology 0 Biology 0 Physics
Volume 37, Number 4, 1997
(4
(4 Fig. 3. Same case as in Fig. 2 (supraglottic carcinoma T2NOMO) after irradiation of 40 Gy. Supraglottic tumor is remarkably reduced in size and not demonstrated on TrWI (a), TzWI (b), or postenhanced T,WI (c). Early enhancement is still demonstrated on dynamic images (d).
the early phase like other squamous cell carcinomas (15 17). In contrast, the fibrotic tissue shows enhancement slowly in the late phase (16). Early enhancement in dynamic images can demonstrate small residual tumors even within the fibrotic tissue, but it is difficult to separate tumors from slowly enhancing fibrotic changes in postcontrast T,WI. Good tumor-to-normal-tissue contrast and the difference of enhancement pattern between tumor, normal tissue, and fibrotic changes all contribute to the high diagnostic accuracy of dynamic MRI after radiotherapy. Inflammation of the normal tissue may be present even 2 weeks after radiation in some patients and show high signal intensity in T2WI. Inflammatory changes also show early enhancement in dynamic images and may mask the
tumor enhancement. It is difficult to visualize small tumors within inflammatory changes, and performance of dynamic MRI after inflammation diminishes is strongly recommended. In our hospital, a local tumor is operatively resected after irradiation of 40 Gy except remarkable tumor remission (complete remission, or 90% decrease of the tumor) was obtained. After a full dose (60-70 Gy) of irradiation, salvage surgery is performed if complete remission of the tumor is not obtained. It is important to evaluate the results of radiation; such evaluation may change the therapy. The result of our study suggests that dynamic MRI is essential to evaluate head and neck tumors after radiotherapy.
REFERENCES 1. Vogl, T.; Dresel, S.; Bilaniuk, L. T.; Crevers, G.; Kang, K.; Lissner, J. Tumors of the nasophaqnx and adjacent areas: MR imaging with Gd-DTPA. AJR 1.54585-592; 1990.
2. Ditlon, W. P. The pharynx and oral cavity. In: Som, P. M.; Bergeron, R. T., eds. Head and neck imaging, 2nd. St. Louis, MO: Mosby-Yearbook; 1991:407-466.
Dynamic
MRI
in
head and neck cancer
3. Tabor, E. K.; Curtin, H. D. MRI of the salivary glands. Radiol. Clin. North. Am. 27:379-392; 1989. 4. Teresi, L. M.; Lufkin, R. B.; Wortham, D. G.; Abemayor, E.; Hanafee, W. N. Parotid masses: MR imaging. Radiology 163:405-409; 1987. 5. Mandelblatt, S. M.; Braun, I. F.; Davis, P. C.; Fry, S. M.; Jacobs, L. H.; Hoffman, J. C. Parotid masses: MR imaging. Radiology 163:411414; 1987. 6. Robinson, J. D.; Crawford, S. C.; Teresi, L. M.; Schiller, V. L.; Lufkin, R. B.; Harnsberger, H. R.; Dietrich, R. B.; Crim, J. R.; Duckwiler, G. R.; Spickler, E. M.; Hanafee, W. N. Extracranial lesions of the head and neck: preliminary experience with Gd-DTPA-enhanced MR imaging. Radiology 172:165-170; 1989. 7. Vogl, T.; Dresel, S.; Bilaniuk, L. T.; Grevers, G.; Kang, K.; Lissner, J. Tumors of the head and neck and adjacent areas: MR imaging with Gd-DTPA. American Journal of Neuroradiology 11:187-194; 1990. 8. Crawford, S. C.; Hamsberger, H. R.; Lufkin, R. B.; Hanafee, W. N. The role of gadolinium-DTPA in the evaluation of extracranial head and neck mass lesions. Radiol. Clin. North. Am. 27:219-242; 1989. 9. Tien, R. D.; Hesselink, J. F.; Chu, P. K.; Szumowski, J. Improved detection and delineation of head and neck lesions With fat-suppression spin-echo MR imaging. American Journal of Neuroradiology 12: 19-24; 1991. 0. Barakos, J. A.; Dillion, W. P.; Chew, W. M. Orbit, skull base, and pharynx: contrast-enhanced fat-suppression MR imaging. Radiology 179:191-198; 1991.
0 Y. BABA
et
al.
787
11. Zoarski, G. H.; Mackey, J. K.; Anzai, Y.; Hanafee, W. N.; Melki, P. S.; Mulkern, R. V.; Jolesz, F. A.; Lull&, R. B. Head and neck: Initial clinical experience with fast spin-echo MR imaging. Radiology 188:323-327; 1993. 12. Panush, D.; Fulbright, R.; Sze, G.; Smith, R. C.; Constable, R. T. Inversion-recovery fast spin-echo MR imaging: Efficacy in the evaluation of head and neck lesions. Radiology 187:421426; 1993. 13. Ross, M. R.; Schemer, D. F.; Phyliss, C.; Enzmann, D. R. MR imaging of head and neck tumors: Comparison of T,weighted contrast-enhanced fat-suppressed images with conventional T,-weighted and fast spin-echo T,-weighted images. AJR 163:173-178; 1994. 14. Arakawa, A.; Tsuruta, J.; Nishimura, R.; Sakamoto, Y.; Korogi, Y.; Baba, Y.; Furusawa, M.; Ishimaru, Y.; Uji, Y.; Taen, A.; Ishikawa, T.; Takahashi, M. MR imaging of linguinal carcinoma: comparison with surgical staging. Radiation Medicine 14:25-29; 1996. 15. Arakawa, A.; Tsuruta, R.; Nishimura, R.; Sakamoto, Y.; Korogi, Y.; Baba, Y.; Furusawa, M.; Ishimaru, Y.; Uji, Y.; Taen, A.; Ishikawa, T.; Takahashi, M. Lingual carcinoma: Correlation of MR imaging with histopathologic findings. Acta Radiologica 37:700-707; 1996. 16. Takashima, S.; Noguchi, Y.; Okumura, T.; Aruga, H.; Kobayashi, T. Dynamic MR imaging in the head and neck. Radiology 189:813-821; 1993. 17. Yamashita, Y.; Takahashi, M.; Sawada, T.; Miyazaki, K.; Okamura, H. Carcinoma of the cervix: dynamic MR imaging. Radiology 182:643-648; 1992.