DOI:10.2214/AJR.05.0799
AJR 2007; 188:W480-W484
© American Roentgen Ray Society
High-Resolution MRI of Basal Cell Carcinomas of the Face Using a Microscopy Coil
Hubert Gufler1,
Folker E. Franke2 and
Wigbert S. Rau1
1 Department of Diagnostic Radiology, Radiology Center, University of Giessen,
Klinikstrasse 36, Giessen 35385, Germany.
2 Department of Pathology, University of Giessen, Giessen, Germany.
Received May 10, 2005;
accepted after revision April 11, 2006.
Address correspondence to H. Gufler
(h.gufler{at}ccb.de).
WEB This is a Web exclusive article.
Abstract
OBJECTIVE. The objective of this article is to evaluate the
diagnostic accuracy of highresolution MRI using a microscopy surface coil to
stage basal cell carcinomas of the face.
CONCLUSION. High-resolution MRI using a microscopy surface coil is a
promising method to determine the extension of basaliomas of the facial region
and to exclude infiltration of bone by the tumor.
Keywords: head and neck imaging • high resolution • MRI • MR technique • soft-tissue neoplasms
Introduction
Noninvasive imaging and characterization of skin abnormalities are
of great interest in dermatology. MRI is an established method for this
purpose when low-noise coils with minimized volume and customized imaging
sequences are applied to optimize the signal-to-noise ratio (SNR)
[1]. Thin and contiguous slices
are needed for adequate MRI of the skin to differentiate between the dermal
and subcutaneous layers
[2-4].
Therefore, high-resolution MRI is essential to discriminate normal from
abnormal features of the skin. Recent developments in MR technology allow the
use of a combination of a microscopy surface coil with a small field of view
and a high-field clinical 1.5-T system with customized pulse sequences, which
together provide high-resolution images of the skin.
The purpose of this prospective study was to compare high-resolution MR
images of basal cell carcinomas (basaliomas) of the face with the results of
histology.
Subjects and Methods
Seven patients with basaliomas of the face were examined between March 2003
and January 2004. The patients were two women 54 and 85 years old and five men
who ranged in age between 51 and 89 years. MRI of each patient was performed
with a 1.5-T system (Gyroscan Intera, Philips Medical Systems) with high-power
gradients (amplitude, 60 mT; slew rate, 150 mT/m/s). We used a microscopy coil
with a diameter of 47 mm that was fixed by straps so that the region examined
and the radiofrequency coil were rigidly locked together to prevent motion
artifacts. The MRI protocol included axial T1-weighted (TR/TE, 450/24; field
of view, 60 mm; matrix, 224 x 256; turbo factor, 3; voxel size, 0.27
x 0.27 x 1.5 mm; excitations averaged, 3; acquisition time, 5.3
minutes) and T2-weighted turbo spin-echo (TSE)
(TRrange/TEeff, 3,000-4,500/100; turbo factor, 13; field
of view, 60 mm; matrix size, 224 x 256; voxel size, 0.27 x 0.27
x 1.5 mm; excitations averaged, 3; acquisition time, 
3.3 minutes)
images. Optionally, T1-weighted sagittal or coronal images were also acquired.
After the administration of gadopentetate dimeglumine, axial, sagittal, and
coronal T1-weighted sequences with or without fat suppression were
performed.
The MRI studies were evaluated prospectively by two radiologists in
consensus who assessed the extension of the tumor and decided whether there
was involvement of the bone, nasal cartilage, subcutaneous tissue, or muscles.
Signal intensities were measured in operator-defined regions of interest that
contained at least 50 pixels. Regions of interest were drawn on the tumor,
subcutaneous fat, dermal layer, bone, cartilage, and muscle, and SNRs were
calculated. These measurements were performed as close to the tumor as
possible because the signal intensities of the tissue diminished toward the
borders of the field of view.
The radiologists had no knowledge of the results of other imaging methods
such as sonography or CT, and because the evaluation was performed
preoperatively, they had no information about the histologic results. In a
final step, the radiologic findings were correlated with the histologic
results.
Results
The tumors were localized in the nasolabial fold (n = 4), cheek
(n = 1), temporal region (n = 1), and infraorbital region
(n =1). In one of the seven patients, MR examination was normal
without evidence of a recurrent tumor. All patients were operated on after MRI
had been performed. Four of the seven patients underwent MRI because
recurrence of a basal cell carcinoma was suspected.
View larger version (120K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —65-year-old man with basal cell carcinoma of left temporal region.
Axial T1-weighted unenhanced turbo spin-echo (TSE) image shows superficially
ulcerated tumor penetrating superficial layer of fascia temporalis
(arrow).
|
|
View larger version (119K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —65-year-old man with basal cell carcinoma of left temporal region.
Coronal T1-weighted image shows that tumor reaches zygomatic bone
(arrowheads), but there is uncertainty whether there is infiltration
of bone.
|
|
View larger version (121K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D —65-year-old man with basal cell carcinoma of left temporal region.
Coronal fat-suppressed T1-weighted image shows infiltration of zygomatic bone
can be excluded. This finding was confirmed by histology.
|
|
The tumors were slightly hyperintense on the T1-weighted images (SNR, 23.8
± 10.3 [SD]) compared with the muscle tissue (SNR, 13.3 ± 5.1)
and isointense compared with the skin (SNR, 21.3 ± 3.6). In the
T2-weighted images, the lesions were markedly hyperintense (SNR, 21.3 ±
13.5) compared with the muscle (SNR, 7.4 ± 2.8) and moderately
hyperintense compared with the skin (SNR, 13.0 ± 4.3). Subjectively,
unenhanced T1-weighted images seemed to be more useful than T2-weighted images
in delineating the contour of the masses and in providing a higher signal
contrast between the tumor and the surrounding tissues.
Calculations of the tumor-to-tissue ratio showed that T2-weighted images
are obviously more helpful in distinguishing the tumor from the normal skin,
but T1-weighted images yielded higher contrast between the tumor and the
subcutaneous fat. Signal enhancement of the tumor was calculated in four
patients for whom contrast-enhanced T1-weighted TSE sequences without fat
suppression were available. The increase in signal intensity after
administration of contrast material ranged between 58% and 99% compared with
unenhanced images. The normal skin enhanced within a wider range, between 10%
and 74%. Involvement of the bone by the tumor could be excluded reliably in
three patients by using T1-weighted fat-suppressed contrast-enhanced sequences
(Fig. 1A,
1B,
1C,
1D). The size of the basaliomas
ranged between 6.2 and 32.0 mm (mean, 15.9 mm). The extensions and the volumes
of the tumors measured on MRI correlated highly with those measured on
histology (R2 =0.98 and R2 = 0.98,
respectively).
On MRI, periosteal infiltration was seen in one patient, and infiltration
of muscle in four patients. These findings were confirmed by histology. A
beginning infiltration of the tumor into the nasal cartilage could not be
ruled out on the unenhanced MR images in one patient. On the contrast-enhanced
images, however, this finding could not be corroborated. The histologic
examination showed that there was a 2-mm gap between the tumor and the nasal
cartilage, without any sign of tumorous infiltration (Fig.
2A,
2B,
2C).
View larger version (158K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —51-year-old man with recurrence of basal cell carcinoma of left
nasal region. Axial T1-weighted turbo spin-echo image shows tumor that extends
into depth of skin. Thin layer of fatty tissue separates tumor from nasal
cartilage. Interruption of continuity of this layer (arrows) suggests
tumor infiltration of nasal cartilage.
|
|
View larger version (136K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —51-year-old man with recurrence of basal cell carcinoma of left
nasal region. Axial T1-weighted contrast-enhanced fat-suppressed image does
not corroborate finding in A. There is no interruption of small layer
of fat between tumor and cartilage (arrow).
|
|
Discussion
A basal cell carcinoma (basalioma) is a semimalignant tumor of the skin
that derives from incompletely differentiated immature keratinocytes of the
epidermis or appendages of the skin
[5]. The natural history of
this tumor is characterized by a slow local infiltration with invasion and
destruction of the underlying structures. Metastases of the tumor are
extraordinarily rare. Basal cell carcinomas develop most likely on sun-exposed
body sites. Areas of predilection on the face are the nose, cheeks, and
nasolabial folds. On clinical examination, a basal cell carcinoma is a nodule
with telangiectases and multiple small satellite nodules. The tumor may be
pigmented, sclerotic (basalioma cicatricans), or ulcerated (ulcus rodens), and
it may infiltrate deeper structures such as bone or cartilage (ulcus
terebrans). Basal cell carcinomas are prone to recurrence. Especially in the
case of a recurrent tumor, it is difficult to estimate the depth of
infiltration clinically and determine if bone, cartilage, muscles, or vessels
are involved. Such cases require an imaging method that is able to
discriminate the superficial structures from the deeper structures of the
skin.
Many studies have shown that MRI is a useful imaging tool to diagnose skin
abnormalities
[6-8].
Krug et al. [9] found that
high-resolution MRI on a 1.5-T system allows identification of the main
histologic patterns of inflammatory skin diseases noninvasively. However, the
structures of the epidermis could not be assessed adequately due to
limitations in spatial and contrast resolution. At 3 T, improved SNR and
higher spatial resolution of human nerves were achieved
[10]. Ashman et al.
[11] showed a marked increase
in SNR with 8 T. There is, however, an increase in magnetic susceptibility and
chemical shift artifacts at higher field strengths
[12,
13]. Magnetic susceptibility
artifacts are known to be most prominent at the bone-to-air and
bone-to-softtissue interfaces. Magnetic susceptibility increases if
gradient-recalled echo sequences are used. Although a better SNR is achieved
at high magnetic field strengths, Hawnaur et al.
[14] reported that even at 0.5
T, the MRI findings correlated well with histologic findings in patients with
skin tumors.
The main purpose of this study was to produce high-quality, high-resolution
images of basal cell carcinomas on a clinical 1.5-T system by means of a
microscopy coil. High-resolution images with 2D TSE sequences were obtained by
selecting 1.5-mm-thin slices, a 60-mm field of view, and an imaging matrix of
224 x 256 mm (in-plane resolution, 0.27 x 0.27 mm) for the
T1-weighted and T2-weighted images. On a subjective visual basis, the margins
of the tumor could be differentiated best from the surrounding tissue on the
unenhanced T1-weighted spinecho images. However, measurements of signal
intensity ratios between tumor and surrounding tissues showed that T1-weighted
and T2-weighted TSE images were almost equivalent in the differences in
contrast between tumor and bone and tumor and muscle.
T1-weighted images yielded a higher contrast between tumor and fat;
T2-weighted images had a superior contrast between tumor and normal skin
(Table 1). Fat-suppressed
contrast-enhanced T1-weighted images were essential to assess the infiltration
of the bone. An incipient involvement of the nasal cartilage by the tumor
could not be ruled out consistently in all sequences in one case. The
histologic examination, however, showed no infiltration of the cartilage.
The "lesion" of one patient with suspicion of a recurrent tumor
on clinical and MR examination proved to be scar tissue with acute and chronic
inflammation on histology (Fig.
3A,
3B). The marked signal
enhancement of the superficial layer of the scar with acute and chronic
inflammation on histology after the administration of the contrast agent made
the lesion indistinguishable from a recurrent tumor on MRI. Dynamic
contrast-enhanced scans might have helped to differentiate scar tissue from
tumor.
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —80-year-old man with suspicion of recurrent basal cell carcinoma of
left nasal region. Axial T2-weighted turbo spin-echo image shows scar tissue
in left nasolabial fold with extension to muscle layer. Surface is excavated,
mimicking ulcerated recurrence of tumor (white arrow). MS = maxillary
sinus. Black arrow = scar tissue. Star = mimic muscle.
|
|
View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —80-year-old man with suspicion of recurrent basal cell carcinoma of
left nasal region. Axial T1-weighted contrast-enhanced fat-suppressed image
shows enhancement of superficial region of scar (arrowheads) and lack
of enhancement of deeper scar tissue. Histology revealed acute and chronic
inflammation of superficial scar layer and no recurrence of basal cell
carcinoma.
|
|
Our examination influenced the planning of the operation in the three
patients with clinical suspicion of bone involvement that could be excluded by
MRI. A recurrent tumor was definitively excluded by MRI in one case; however,
an open biopsy was performed, confirming the diagnosis.
View larger version (113K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C —54-year-old woman with nodule on left cheek. Photomicrograph of
stained specimen shows close concordance with MR microscopy imaging. Solid
arrows = tumor borders, open arrow = ulcerated tumor surface, V = vessels. (H
and E)
|
|
A limitation of the 47-mm microscopy coil is that the signal intensity
decreases markedly from the center to the periphery of the 6-cm field of view,
so the determination of signal intensities represents a problem if the tumor
is not in the center of the field of view. Another limitation with using
microscopy coils is that the contralateral regions of the face are not
available for comparison. Therefore, it may be difficult to detect
superficially growing basaliomas. This shortcoming, however, could be avoided
by using multichannel, multicoil imaging. A further limitation of this study
is that 3D gradient-echo sequences were not used, which might have yielded
even better results regarding spatial resolution and SNR. Song et al.
[15] produced high-resolution
images of the dermis by means of a customized 3D gradient, partial flip-angle
spin-echo pulse sequences, and very small transmit-receive coils on a 1.5-T
clinical scanner in vivo, obtaining voxel sizes as small as 19 x 78
x 800 µm. We preferred spin-echo sequences to gradient-echo sequences
to avoid susceptibility artifacts because many of the tumors were located near
the nasal cavity or the maxillary sinus. Furthermore, there are only a small
number of pulse sequences commercially available for microscopy coils.
In conclusion, high-resolution MRI with a commercially available microscopy
surface coil and a clinical high-field 1.5-T MR system with strong gradients
is a suitable method to determine the extension and depth of infiltration of
basal cell carcinomas of the facial region (Fig.
4A,
4B,
4C). In the future, the
diagnostic accuracy of high-resolution MRI of these tumors will improve with
the use of 3-T MR systems and their combination with dedicated sequences and
the microscopy coil.
References
- Liffers A, Vogt M, Ermert H. In vivo biomicroscopy of the skin with
high-resolution magnetic resonance imaging and high frequency ultrasound.
Biomed Tech 2003;48
: 130-134
- Mäurer J, Knollmann FD, Schlums D, et al. Role of
high-resolution magnetic resonance imaging for differentiating
melanin-containing skin tumors. Invest Radiol1995; 11:638
-643
- Bittoun J, Saint-Jalmes H, Querleux BG, et al. In vivo
high-resolution MR imaging of the skin in a wholebody system at 1.5 T.
Radiology 1990;176
: 457-460[Abstract/Free Full Text]
- Thali MJ, Dirnhofer R, Becker R, Oliver W, Potter K. Is virtual
histology the next step after the virtual autopsy? Magnetic resonance
microscopy in forensic medicine. Magn Reson Imaging2004; 22:1131
-1138[CrossRef][Medline]
- Schwartz RA. Skin cancer: recognition and
management. New York, NY: Springer-Verlag, 1988:57
-69
- Mirrashed F, Sharp JC. In vivo morphological characterisation of
skin by MRI micro-imaging methods. Skin Res Technol2004; 10:149
-160[CrossRef][Medline]
- Mirrashed F, Sharp JC, Krause V, Morgan J, Tomanek B. Pilot study
of dermal and subcutaneous fat structures by MRI in individuals who differ in
gender, BMI, and cellulite grading. Skin Res Technol2004; 10:161
-168[CrossRef][Medline]
- Pennasilico GM, Arcuri PP, Laschena F, et al. Magnetic resonance
imaging in the diagnosis of melanoma: in vivo preliminary studies with dynamic
contrast-enhanced subtraction. Melanoma Res2002; 12:365
-371[CrossRef][Medline]
- Krug B, Schulze HJ, Kugel H, Krahe T, Wesselmann C, Lackner K.
Correlation of MRI and histopathological findings in inflammatory skin
diseases [in German]. Rofo 1998;168
: 429-435[Medline]
- Hormann M, Traxler H, Ba-Ssalamah A, et al. Correlative
high-resolution MR-anatomic study of sciatic, ulnar and proper palmar digital
nerve. Magn Reson Imaging 2003;21
: 879-885[CrossRef][Medline]
- Ashman CJ, Farooki S, Abduljalil AM, Chakeres DW. In vivo
high-resolution coronal MRI of the wrist at 8T. J Comput Assist
Tomogr 2002; 26:387
-391[CrossRef][Medline]
- Constable RT, Henkelman RM. Contrast, resolution, and detectability
in MR imaging. J Comput Assist Tomogr1991; 15:297
-303[Medline]
- Abduljalil AM, Robitaille P-ML. Macroscopic susceptibility in ultra
high field MRI. J Comput Assist Tomogr1999; 23:832
-841[CrossRef][Medline]
- Hawnaur JM, Dobson MJ, Zhu XP, Watson Y. Skin: MR imaging findings
at middle field strength. Radiology 1996;201
: 868-872[Abstract/Free Full Text]
- Song HK, Wehrli FW, Ma J. In vivo MR microscopy of the human skin.
Magn Reson Med 1997;37
: 185-191[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
