British Journal of Radiology (2005) 78, 1075-1077
© 2005 British Institute of Radiology
doi: 10.1259/bjr/36116150
Parametrial invasion in carcinoma of cervix: role of MRI measured tumour volume
A Jena, DNB (NM)
1
R Oberoi, MD
1
S Rawal, MS, Mch
2
S K Das, MD
2 and
K K Pandey, MS
3
Departments of 1 MRI, 2 Gynae Oncology and 3 Surgical Oncology, Rajiv Gandhi Cancer Institute & Research Centre, Sector-5, Rohini, New Delhi-110085, India
 |
Abstract
|
|---|
The aim of the study was to determine the correlation between MRI measured tumour volume and parametrial invasion on histology in the evaluation of carcinoma of the cervix showing full thickness stromal invasion (FTSI). Original MR images of 159 surgical cases of carcinoma of the cervix retrieved from the MR image bank of the department were analysed retrospectively. Tumour volume from the maximum tumour diameters in each case was computed using the standard formula for an ellipse and correlated with parametrial invasion on histology. Of 105 cases showing FTSI on axial T2 weighted MRI, tumour volume between the two groups of cases, with parametrial invasion (n=27) and without parametrial invasion (n=78), shows a statistically significant difference (p=0.006). Best accuracy of 60.95%, sensitivity 59.26% and specificity 61.54% was found for tumour volume of 36.39 cm3. The study concludes that MRI measured tumour volume is associated with low accuracy in the evaluation of parametrial invasion in carcinoma of cervix showing FTSI in axial T2 weighted MR images and may not help as an additional diagnostic criterion to predict parametrial invasion pre-operatively.
|
Introduction
|
|---|
Invasion of the parametrium is an important factor in the pre-operative evaluation of carcinoma of the cervix that significantly influences staging and treatment [1, 2]. Presence of an intact low signal intensity rim surrounding the cervical tumour in high-resolution axial T2 weighted MR images is a generally accepted indicator of negative parametrial invasion [2–4]. On the other hand, 30–73% of cases showing full thickness stromal invasion (FTSI) on MRI with loss of the low signal intensity rim of the cervix are associated with parametrial invasion [3, 4]. MR tumour volume has been shown to be associated with other risk factors such as tumour invasion depth, locoregional extent and lymph node involvement [5, 6], and is primarily correlated with clinical FIGO staging and patient outcome [6, 7]. Attempts have been made to improve the role of MRI in carcinoma of the cervix with FTSI by using tumour volume as an additional parameter to predict parametrial invasion pre-operatively [8]. We observed that many large size tumours seen on MRI were associated with negative parametrium on histology [9], thereby indicating the poor predictive role of tumour volume contrary to earlier reports. This prompted us to carry out an analysis of our cases to find out if a correlation exists between MRI measured tumour volume and parametrial invasion on histology in cases showing FTSI on MRI.
|
Patients and methods
|
|---|
A retrospective analysis of original MR images from 159 surgical cases of cancer of the cervix during 3
years from January 1999, retrieved from the MR image bank of the department that stores all DICOM MR images of studies in original, form the material of this work. MR studies in all cases are performed on 1.0 T scanner (Magnetom; Siemens, Erlangen, Germany) and all patients had undergone surgery within 15 days of MRI examination. Imaging protocol includes turbo spin echo (TSE) T2 weighted sagittal images: repetition time (TR)/echo time (TE) 4000 ms/99 ms, 5 mm slice thickness with 0.5 mm interslice gap, 350 x 500 field of view (FOV) and matrix size 256 x 256; and TSE T2 weighted high resolution axial images: TR/TE 5136 ms/132 ms, 5 mm slice thickness with 0.5 mm interslice gap, 350 x 500 FOV and matrix size 512 x 256. Two trained observers evaluated the MR images independently, with particular reference to tumour extension involving full thickness of the stroma of cervix to either partial or complete circumference, besides measuring the (a) transverse, (b) anteroposterior and (c) craniocaudal diameter of the tumour in centimetres using both T2 weighted TSE sagittal and T2 weighted TSE axial MR images. Tumour volume was computed using the formula for ellipse as abc
/6 cm3. Mann-Whitney test and STATA software were used for statistical analysis.
|
Results
|
|---|
Out of a total of 159 surgical cases of carcinoma of the cervix, 54 cases showed intact rim of the normal stroma of the cervix around the tumour on axial T2 weighted images and were not included for further analysis.
105 cases showed FTSI of the cervix on axial T2 weighted MR images, 91 of which were shown to be squamous cell carcinoma, 9 were adenocarcinoma and 5 were adenosquamous type on histology. The age of patients ranged from 28 years to 82 years. FIGO staging classified 53 of these cases as 
Ib, 6 as IIa and 46 as 
ge; IIb. 27 out of 105 cases with FTSI showed parametrial infiltration (25.71%). Interobserver agreement was excellent, with Kappa coefficient of 0.89 for MR findings (assessment of tumour extent) and 0.85 for measuring tumour volume.
Figure 1
shows the distribution of tumour volume in two groups of patients, those with parametrial involvement and those without parametrial involvement, using box and whisker plots. A statistically significant difference (p=0.006) in tumour volume was found between the two groups of patients. In the parametrium negative cases, tumour volume ranged from 0.5 cm3 to 97.43 cm3 with median tumour volume of 30.75 cm3 including one extreme value of 156.3 cm3. Tumour volume for parametrium positive cases was ranging from 17.30 cm3 to 99.20 cm3 with median value of 49.14 cm3.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 1. Distribution of tumour volume between cases showing parametrial involvement and without parametrial involvement by using box and whisker plots.
|
|
Sensitivity, specificity and accuracy assessed for each MR measured tumour volume for parametrial outcome on histology were calculated with the help of STATA software and shown in Figure 2. Best accuracy of 60.95%, sensitivity 59.26% and specificity 61.54% was found for tumour volume of 36.39 cm3.
|
Discussion
|
|---|
It is difficult to assess parametrial involvement when a tumour completely obliterates the normal hypointense rim of cervical stroma in axial T2 weighted MR images and there is no obvious thickening or nodularity in the parametrial tissue to indicate frank invasion [3]. 40–73% of these cases are reported to show parametrial invasion on histology [2–4]. 25.71% (27/105) of our cases with FTSI showed parametrial invasion on histology. Although lower than the reported value, a similar figure of 27.2% has been found by Okuno et al [8] in a series of 33 cases in which 9 had parametrium infiltration on histology. However, compared with 87% accuracy, we observed a much lower accuracy (less than 61%) for MRI measured tumour volume in predicting parametrial invasion in cases showing FTSI on MRI. Best accuracy for parametrial invasion in our cases found with tumour volume of 36.39 cm3 was much higher than the 13.9 cm3 observed by Okuno et al in their series. Although statistically significant differences in the tumour volume were noted in this series between parametrium positive and parametrium negative cases, distribution of volume showed a considerable overlap in parametrium positive and parametrium negative cases which accounts for low accuracy in the present study.
In conclusion, the study showed that because of low accuracy, MRI measured tumour volume may not help as an additional diagnostic criterion in cases of carcinoma of the cervix showing FTSI in predicting parametrial invasion pre-operatively.
|
Acknowledgments
|
|---|
We acknowledge the contribution Dr T Jeisobers for statistical analysis of data and Mr Pradeep Negi for helping in the preparation of the manuscript.
Received for publication June 2, 2004.
Accepted for publication May 26, 2005.
|
References
|
|---|
- Greco A, Mason A, Leung AWL, Dische S, McIndoe AJ, Anderson MC. Staging of carcinoma of the uterine cervix: MRI surgical correlation. Clin Radiol 1989;40:401–5.[CrossRef][Medline]
- Togashi K, Nishimura K, Sagoh T, Minami S, Noma S, Fujisawa I, et al. Carcinoma of the cervix: staging with MR imaging. Radiology 1989;171:245–51.[Abstract/Free Full Text]
- Hricak H, Lacey CG, Sandles LG, Chang YCF, Winkler ML, Stern JL, et al. Invasive cervical carcinoma: comparison of MR imaging and surgical findings. Radiology 1988;166:623–31.[Abstract/Free Full Text]
- Sironi S, Belloni C, Taccagni G, DelMaschio A. Carcinoma of the cervix: value of MR imaging in detecting parametrial involvement. AJR Am J Roentgenol 1991;156:753–6.[Abstract/Free Full Text]
- Hawnaur JM, Johnson RJ, Buckley CH, Tindall V, Isherwood I. Staging; volume estimation and assessment of nodal status in carcinoma of the cervix: comparison of magnetic resonance imaging with surgical findings. Clin Radiol 1994;49:443–52.[CrossRef][Medline]
- Chen AC, Sung WH, Wang PH, Sheu MH, Doong JL, Yuan CC. Correlation of three-dimensional tumor volumetry with cervical cancer prognostic parameters. Eur J Gynaecol Oncol 2002;23:401–4.[Medline]
- Wagenaar HC, Trimbos JB, Postema S, Anastasopoulou A, van der Geest RJ, Reiber JH, et al. Tumor diameter and volume assessed by magnetic resonance imaging in the prediction of outcome for invasive cervical cancer. Gynaecol Oncol 2001;82:474–82.[CrossRef][Medline]
- Okuno K, Joja I, Miyagi Y, Sakaguchi Y, Notohara K, Kudo T, et al. Cervical carcinoma with full thickness stromal invasion: relationship between tumor size on T2W images and parametrial involvement. J Comput Assist Tomogr 2002;26:119–25.[CrossRef][Medline]
- Jena A, Oberoi R, Rawal S, Das SK, Pandey KK. Influence of tumor type and location of growth on MRI in the staging of cancer cervix. Asian Oceanian J Radiol 2004;9:16–20.
Top
Abstract
Introduction
