| Background and PurposeOvarian tumors are the leading indication for gynecological surgery.Trans-vaginal ultrasound assessment is the initial investigation of choice in suspected ovarian cancer and has been found to be sensitive for the detection of malignancy,Its advantages are high efficiency and low price,it allows realtime evaluation of the region of interest,including functional information on tissue vascularity,but it is less accurate for complex or indeterminate masses.MR imaging offers several advantages in the imaging of the pelvis:It provides excellent soft-tissue contrast;it does not require the use of ionizing radiation.MRI has proved to be superior to computed tomography for the assessment of complex and indeterminate ovarian tumors.Several authors have suggested that certain morphological features detected on nonenhanced MR sequences,such as a solid portion,papillary projections,or thickened irregular septa,are suggestive of malignancy.Although some morphological characteristics have been described,there are still some overlapping features on conventional MRI among ovarian tumors.As an advanced technique,dynamic contrast enhanced MRI(DCE-MRI)could be used not only to assess the microcirculatory perfusion and vascular permeability in ovarian tumors noninvasively,but provide a more comprehensive discrimination of ovarian tumors according to the time-intensity curves(TIC)?some semi-quantitative and quantitative parameters.In several other malignancies,including breast,prostate and renal cancer,some semi-quantitative and quantitative analysis of the dynamic contrast-enhanced sequence has proved useful.DCE-MRI has also been shown to correlate with tumor angiogenesis biomarkers in ovarian cancer.The semi-quantitative analysis include the time intensity curve analysis,and.the semi-quantitative parameters based also on time intensity curve.Thomassin-Naggara et al presented three shapes of the time-intensity curve,when solid tissue is enhanced with a weak and progressive curve in comparison to the myometrium,the curve is named "type 1."When solid tissue is enhanced with a moderate enhancement in comparison to myometrium with a plateau,the curve is named "type 2." when solid tissue is enhanced with a curve steeper than that of myometrium,whatever the intensity of enhancement,the curve is named "type 3".Curve type 3 appeared specific for invasive tumors.Semi-quantitative parameters:With a higher temporal resolution acquisition data,Thomassin-Naggara et al.determined three semi-quantitative parameters:enhancement amplitude(EA),time of half rising(THR),and maximal slope(MS)of the curve.Using myometrium as internal reference,the enhancement amplitude ratio(EAratio),time of half rising ratio(THRratio),and maximal slope(MSratio)were significantly higher in invasive malignant tumors than in benign and borderline tumors.With a lower temporal resolution acquisition data of 30 seconds,Dilks et al.demonstrated the usefulness of other parameters including maximum actual enhancement(SImax),maximum relative enhancement(SIrel),wash-in rate(WIR);Bernardin et al.demonstrated that benign lesions displayed a lower mean maximal signal intensity(SImax),a lower relative signal intensity(SIrel),and a lower wash-in rate(WIR)than borderline and malignant tumors.These authors underlined the simplicity of this technique because the software used to calculate enhancement parameters is widely available on MR imaging workstations.DCE-MRI has also been shown to be associated with biomarkers of tumor angiogenesis in ovarian cancer.Ovarian cancer is characterized by a large number of immature microvessels.These microvessels are characterized by the lack of high expression of vascular endothelial growth factor(VEGF)receptor(VEGFR-2),which is involved in endothelial cells and epithelial cells.These pathophysiological features have been shown to be consistent with changes in MR perfusion parameters.Semi-quantitative analysis are based on signal intensity evaluation.However,using MR imaging,signal intensity depends mainly on the type of acquisition parameters,such as flip angle and TR.Image contrast provided by the administration of contrast agent and linearity between signal intensity variation and contrast agent concentration are highly dependent on these parameters.Thus,many authors argue that we need to obtain reproducible perfusion parameters independent of acquisition conditions.We needed to develop perfusion parameters that are not expressed according to signal intensity but to gadolinium concentration.Then,a more recent approach consists in a quantitative analysis which is based on a pharmacokinetic modeling allowing conversion of signal intensity into gadolinium concentration.Depending on the temporal resolution of the acquisition,different pharmacokinetic models may be applied.Tofts-Kety model is the most used pharmacokinetic model in this context thanks to its good reproducibility.Quantitative parameters include:volume transfer constant(Ktrans),per unit volume in tissues outside the vascular space(EES)volume fraction(Ve),EES and plasma between the rate constant Kep;Kep = Ktrans/Ve theory.The quantitative analysis of DCE-MRI has become the focus of clinical research,has been applied in brain tumor,breast,prostate,cervical and other parts of the assessment,the main research directions for the identification of benign and malignant tumor,malignant tumor classification,efficacy evaluation etc.At present,there are few studies on the value of quantitative analysis in the diagnosis of ovarian tumors.The purpose of our study was:first,to evaluate whether semi-quantitative and quantitative DCE-MRI can differentiate benign from malignant ovarian masses based on the measured enhancement characteristics using histology as the gold standard;second,to establish the threshold criteria to differentiate benign and malignant tumor;third,to compare the diagnostic efficacy of the semi quantitative parameters and quantitative parameters.Material and MethodsForty-five cases of ovarian lesions were evaluated in our study.All patients underwent MR dynamic enhanced scan and confirmed by pathology,including 13 cases of benign tumors,32 cases of malignant tumors;Following dynamic contrast enhanced imaging,regions of interest were drawn around the solid tumor component.The selected areas of interest were significantly enhanced in the soft tissue area,avoiding necrosis,bleeding,cystic change,and blood vessels,etc.1.Time-intensity curve and semi-quantitative parameters:the time-intensity curve(TIC)was generated automatically using Mean Curve software package.The TICs were classified into three types:(1)curve ? showed a rapid rise,seen as "fast-in and slow-out" pattern;(2)curve ? had a moderate and continued enhancement and displayed as "platform" pattern;(3)curve ? had a "slow-in and slow-out" pattern.Curves ?-? indicated malignant and benign lesions,respectively.According to the TIC curve,the semi-quantitative parameters were obtained;The semi-quantitative parameters include the signal intensity(SI)rate at 60s after enhancement(SI60%)and Time to peak within 200 s after Gd-DTPA injection(TTP200S);The signal intensity(SI)of the TIC peak 60 s after Gd-DTPA injection(SIpost60)was set as the standard,which was used to calculate the estimated SI rate at 60 s after enhancement(SI60)with the following formula:[(SIpost60-SIpre)/SIpre×100%];Slpre is the SI pre-enhancement.2.Quantitative measurements:With the Siemens TISSUE 4D software package,the software designed to use the two compartment model modeling of Tofts for the quantitative measurements.According two flip angle T1-mapping sequence before enhanced scan,calculate the baseline T1 value,and T1 enhancement value of the image after dynamic contrast-enhanced scan was calculated,select the solid composition of mass as a region of interest(ROI),which reflects the quantitative parameters of tumor microvascular permeability:?Ktrans:is the volume transfer constant between the blood plasma and extravascular extracellular space(EES)per unit volume of tissue(min-1);?Kep:the rate constant between the EES and blood plasma(min-1);?Ve:the volume of extravascula rextra cellular space per unit volume of tissue.Statistical analysis was performed with MedCalc 15.2.2 software.To analyze the differences between different types of TIC curves in benign and malignant ovarian tumors using chi-square test.Data were presented as mean±SD(continuous variables).The Mann-Whitney and U parameters were used to analyze the difference of semi quantitative and quantitative parameters between benign and malignant ovarian tumors.Receiver operating characteristic(ROC)curves were used to determine the dynamic contrast-enhanced MR image parameter cutoffs with the best sensitivity and specificity for distinguishing between benign and malignant tumors.A P value of<0.05 indicates statistical significance.ResultsAmong 32 malignant tumors(age range,32-76 years;mean 58 years),17 were serous cystadenocarcinoma,13 were mucinous cystadenocarcinoma,1 was sertoli cell tumor and 1 were metastatic carcinoma;15 benign lesions(age range,35-80 years,mean 55 years);included 5 Follicular fibrosis,8 Follicular tumor and 1 cystadenofibromas.12 malignant tumors(37.50%)displayed type ? TIC,20 malignant tumors(62.50%)displayed type ? TIC,type ? TIC can not been seen in the malignant tumor;9 benign tumors(69.23%)showed type ? TIC,3cases(23.08%)displayed type ? TIC,1 cases(7.69%)displayed type ? TIC,and significant differences were found in TIC type between malignant and benign ovarian lesions(P<0.0001).SI60%was higher for malignant tumors(126.43±17.90%)than for benign(69.62±33.12%),TTP200S was longer for benign tumors(141.11± 48.12s)than for malignant(72.93±22.24s);There was a significant difference in SI60%and TTP200S between the two groups,a threshold value of 5I60%104,62%,TTP200<129s had a high sensitivity(93.7%.96.87%respectively)for predicting malignant tumors,the specificity were 76.9%?69.23%respectively.Malignant tumors displayed higher Ktrans(0.352±0.114 mm-1)?Kep(0.535±0.182min-1)and Ve(0.761 ±0.428)than benign tumors(0.116±0.068mm-1?0.293±0.106 min-1 and 0.367±0.123 respectively).There was a significant difference in Ktrans?Kep and Ve between the two groups.A threshold value of Ktrans is 0.234min-1 have a high specificity for predicting the malignant,the sensitivity was 87.50%,specificity was 100%;Kep>0.414 min-1 had a sensitivity of 93.7%and specificity of 81%;Ve>0.512 had a sensitivity of 68.75%and specificity of 92.31%.There was no significant difference between semi-quantitative and quantitative parameters.The area under the curve of the quantitative parameter Ktrans(0.974)was higher than semi quantitative parameters(the area under the curve was 0.829 and 0.882,respectively).ConclusionOur preliminary study has confirm DCE-MRI has high efficiency in the differentiation of benign and malignant ovarian tumors according to time intensity curve(TIC),some semi-quantitative parameters and quantitative parameters.Based on the best threshold of TIC?semi-quantitative and quantitative parameters,there are high sensitivity and specificity to identify the benign and malignant ovarian tumors;the difference between the diagnostic efficacy of semi-quantitative parameters and quantitative parameters of ovarian tumors had no significant difference,the area under the curve of Ktrans was higher than semi quantitative parameters,The Ktrans ratio was the most relevant factor for discriminating benign from malignant tumors,In conclusion,these parameters were important supplement for conventional morphology MR diagnosis.For those cases where conventional MRI diagnosis is uncertain,quantitative and semi-quantitative parameters are particularly useful,which can avoid unnecessary radical surgery. |
PDF Full Text Request