Korean J Obstet Gynecol.
2001 Jul;44(7):1296-1306.
Analysis of in vitro, ex vivo and in vivo 1H Magnetic Resonance
Spectroscopy Pattern in Squamous Cell Carcinoma of Uterine Cervix
- Affiliations
-
- 1Department of Obstetrics and Gynecology, College of Medicine,
Kangwon National University, Chunchon, Korea.
- 2Department of Obstetrics and Gynecology, College of Medicine,
University of Ulsan, Asan Medical Center, Seoul, Korea.
- 3Department of Diagnostic Radiology, College of Medicine,
University of Ulsan, Asan Medical Center, Seoul, Korea.
- 4Department of Diagnostic Radiology, NMR Laboratory,
Asan Institute for Life Science, Seoul, Korea.
Abstract
OBJECTIVE
To evaluate the potential value of in vitro, ex vivo, and localized in vivo 1H Magnetic resonance spectroscopy(MRS) for detecting and characterizing squamous cell carcinoma of uterine cervix.
MATERIALS AND METHODS
Eleven cases of squamous cell carcinoma and 6 cases of normal cervical
tissue underwent in vitro and ex vivo 1H MRS. In vitro analysis was carried out after processing the specimen
with dual phase extraction method. In ex vivo test, at least 500microliter of tissue specimen was directly analyzed
by 100.14 MHz NMR spectrometer (Bruker 11.7T AMX, Germany) without any processing procedure. In
addition, 29 patients with squamous cell carcinoma of uterine cervix and 6 normal women were examined by
in vivo 1H MRS. Localized in vivo 1H MR spectra of the human uterine cervix were obtained using PRESS
sequence with homebuilt endovaginal surface coil. Surface coil T2 weighted images were used as the localizer
images where voxel (1-3 mm3) was located to contain the cancer area only. The following spectral acquisition
parameters were used: SW=2500 Hz, SI=2048, TR=3.0 sec, TE=20/135 msec, NEX=2, and NS=36. H2O
signal suppressed down to 96-99% by 3-pulse CHESS technique. The spectra were processed with 0.5 Hz of
exponential line broadening, zerofilling of 8K followed by linear phasing.
RESULTS
The intense resonance signal at 0.9 ppm and 1.3 ppm by lipid component of triglyceride was
visible in all cases (11/11) of squamous cell cancer of cervix both in vitro and ex vivo test. But in 6 normal
cervical tissues, such peaks were not detected at all. The in vivo spectrum of the cervical cancer was
comparable with that in in vitro and ex vivo spectrum. In the in vivo spectra of the cervical cancer, the
resonance peaks at 0.9 ppm, 1.3 ppm, 3.0 ppm, and 3.2 ppm from CH3 and CH2 groups of neutral mobile
lipid (triglyceride), creatine, and choline containing residues respectively, were consistent with results reported
earlier. The resonance peak at 1.3 ppm which is from acyl chains of triglyceride was characteristically intense
and seen in 25 of 29 cancer patients. But triglyceride peak at 1.3 ppm was not detected in normal cervix
(sensitivity 86.2%; specificity 100%; positive predictive value 100%; negative predictive value 60%).
CONCLUSION
The study shows that 1H MRS can effectively discriminate squamous cell cancer from
normal cervix in most cases. It also demonstrates the feasibility of localized in vivo 1H MRS technique as a
new diagnostic modality in the detection of squamous carcinoma of uterine cervix.