Subscribe to RSS
DOI: 10.1055/s-2005-931827
3 Tesla Magnetic Resonance (MR) Microimaging of Primary Cutaneous Melanoma
Purpose: This study investigates the efficacy of 3 Tesla magnetic resonance (MR) microimaging for characterisation of primary melanoma. Technologies currently in use for non-invasive characterization of melanoma (confocal microscopy, optical coherence tomography, or high frequency ultrasound) are limited by restricted penetration depth. In contrast, MR microimaging can determine the full depth of all tumors.
Methods: Five patients with primary melanomas of lateral dimension 1cm or greater were imaged with a 3.7cm surface coil using a 3D gradient echo sequence (1). Nominal spatial resolution was 0.18×0.12mm in-plane, with 1 to 2.0mm slice thickness. Tumour size was measured as the lateral an intradermal extent of hyperintense change in the epidermis.
Results: High quality images were obtained from all five tumours. There was close correlation between the tumor surface dimensions, Breslow thickness, and Clark levels estimated by MR microimaging and histopathology. Measurment of tumor cross-sectional area in successive MR microimaging slices enabled the estimation of tumor volume (Table).
|
|
|
MRI |
Pathology |
|||||||
n° |
sex/age |
site |
Volume(cm3) |
Length(mm) |
Width (mm) |
Breslow(mm) |
Clark |
Length(mm) |
Width(mm) |
Breslow(mm) |
Clark |
1 |
F/57 |
shoulder |
0.63 |
11 |
9 |
1.2 |
III |
- |
- |
0.9 |
III |
2 |
M/37 |
leg |
0.72 |
10 |
9 |
1.5 |
III |
11 |
8 |
1.6 |
III |
3 |
M/76 |
nail thumb |
0.23 |
11 |
8 |
5.6 |
IV |
10 |
9 |
5.5 |
IV |
4 |
M/36 |
scapula |
0.12 |
8 |
7 |
4.6 |
IV |
9 |
6 |
5.2 |
IV |
5 |
F/47 |
knee |
2.95 |
26 |
25 |
7.6 |
IV |
25 |
25 |
7.5 |
IV |
Conclusion: MR microimaging of primary melanoma demonstrates close correlation with the microscopic features documented by histopathology. Breslow thickness is currently considered to be the most reliable prognostic indicator. However, primary tumor volume as estimated by MR microimaging may prove to be a more reliable indicator than tumor thickness as tumor volume accounts for both lateral and vertical growth.
References: (1) King SB, Ryner LN, Tomanek B, Sharp JC, Smith IC. MR spectroscopy using multi-ring surface coils. Magnetic Resonance in Medicine, 1999. 42: p. 655–664