SS 316 - Female genitourinary cancers
Purpose: The objective was to evaluate the value of magnetic resonance imaging with quantitative DCE and DWI analysis for differential diagnosis of complex ovarian tumours (OT).
Methods and Materials: MRI was performed in 141 patients (173 OT) using the following protocol: Sg T2, Ax T2, Co T1, DCE using dynamic T1-FS, and DW MRI (b-values 0, 1000 v/vv2). Following quantitative parameters were evaluated: amplitude of contrast agent accumulation (ACAA), period of rise of signal intensity (HRSI), the maximum curvature (MC), apparent diffusion coefficient (ADC). All MRI results were compared with histological verification.
Results: The tumour grade distribution: benign tumours (BTs)—49%, borderline (BrTs)—12%, malignant (MTs)—39%. ACAA was signiﬁcantly higher in MT—165% (118.5-211.1%) than in BTs—68.2% (40.5-96.2%) (P<0.001) and BrTs—82.7% (59.1-141.5%), (P=0.05); HRSI was signiﬁcantly higher in BTs—35.1 s (31.0-41.2 s) than the BrTs—27.6 s (23.1-29.4 s), (P=0. 05), and BTs—23.6 s (21.9-30.2 s) (P = 0.01). MC was 1.75 (1.0-2.4), 2.8 (2.04-3.7) and 6.04 (4.25-9.26) for BTs, BrTs and MTs, respectively, and was signiﬁcantly higher in MTs (P<0.01). Mean ADC values of MT's solid component were signiﬁcantly lower than in BTs (1.013 ± 0.12 and 1.41 ± 0.3, respectively (P<0.05)). The diagnostic efﬁciency of multiparametric MRI was as follows: the accuracy of 93.4%, sensitivity of 92.9% and speciﬁcity of 95.1%.
Conclusion: Quantitative multiparametric MRI should be obligatorily included into preoperative assessment of indeterminate adnexal masses, allowing clear distinction the degree of OT malignancy, to optimise the treatment strategy.
Delayed detection of recurrence on postoperative PET-CT in patients with advanced ovarian cancer: a study of initially missed lesions on MDCT
Purpose: To describe patterns of erroneous postoperative multi-detector row CT (MDCT) interpretation of recurred peritoneal lesions in patients with advanced ovarian cancer (AOC).
Methods and Materials: Between 2011 and 2016, we reviewed postoperative follow-up PET-CT images of 223 patients with suspected recurrence based on images, who had a history of initially diagnosed AOC (International Federation of Gynaecology and Obstetrics (FIGO) stage IIIC-IV) and underwent primary cytoreduction. Under the awareness of the recurred location, we re-evaluated MDCT images taken right before the recurrence was detected on PET-CT and categorised the cases as either absence of lesion or presence of lesion (missed case). We performed region-based comparisons of the missed cases according to predefined peritoneal locations. In the same period, preoperatively neglected lesions in AOC patients were analysed for comparison.
Results: Re-evaluation of follow-up MDCT revealed that 42.6 % (95/223) of recurrent lesions were missed. According to region-based comparisons of the missed cases, the most commonly missed site was pelvic cavity (18.9%, 18/95) followed by porta hepatis (14.7%, 14/95) and para-aortic/aortocaval lymphadenopathy (12.6%, 12/95). Compared to postoperative images, the most commonly neglected site in preoperative staging evaluation was porta hepatis followed by small bowel mesentery and pelvic cavity.
Conclusion: The most commonly missed location on post-op CT imaging was pelvic cavity followed by upper abdominal lymphadenopathy while porta hepatis was equally overlooked on both pre- and postoperative evaluations. Compared to preoperative evaluations, missed rate of mesenteric implants was lower in relapsed cases. Familiarity with these imaging features may aid in detection of recurrence.
Dedicated diffusion-weighted MR imaging for staging peritoneal metastases: a preoperative selection tool for cytoreduction surgery (CRS) candidates
Purpose: The peritoneal cancer index(PCI) quantifies the extent of peritoneal metastases found at surgery. To prevent unnecessary surgical staging procedures in patients whose peritoneal tumorburden is too extensive to benefit from CRS/HIPEC, patient selection based on preoperative imaging would be ideal. Computed tomography is inaccurate in predicting the PCI. Therefore, we compare the PCI estimated preoperatively by DW-MRI with the PCI found at surgery to assess whether DW-MRI can be used to select CRS/HIPEC candidates.
Methods and Materials: In this ongoing study eighteen consecutive patients (April-September 2016,M/F=3/15) with histologically proven peritoneal carcinomatosis from either colorectal(n=13) or ovarian(n=5) origin were included. Patients were scheduled for exploratory laparoscopy and/or CRS/HIPEC and underwent preoperative dedicated DW-MRI(scan time=30min). Two independent readers prospectively determined the PCI on DW-MRI. Patients were categorized as low-risk(PCI 0-21) versus high-risk(PCI 22-39); in our center considered operable versus non-operable. Reference standard was PCI found at surgery. Quadratic weighted kappa was used to evaluate the interobserver agreement.
Results: At surgery the mean PCI was 11.3(range 0-26). For reader 1 the mean PCI was 11.5(0-29) and for reader 2 9.8(0-29). Both readers categorized 17 out of 18 patients correctly(accuracy 94%) when compared to surgical findings. Both readers understaged the same patient.
The interobserver agreement was perfect(k=1.0).
Conclusion: These data suggest that DW-MRI is an accurate selection tool to noninvasively select patients who could potentially benefit from CRS/HIPEC. No overstaging occurred with DW-MRI, meaning no patients would have been denied potential curative surgery if DW-MRI would have been used as a selection tool.
The usefulness of preoperative radiological assessment of peritoneal cancer index (PCI) as predictive factor for optimal debulking and survival: a preliminary study
Purpose: To evaluate if the radiological assessment of PCI on preoperative CT of patients with ovarian cancer can be predictive of complete surgical debulking and whether the radiological PCI (rPCI) is correlated with Overall Survival (OS) and Progression Free Survival (PFS).
Methods and Materials: We considered all patients with a diagnosis of ovarian cancer, a preoperative CT, up-front cytoreductive surgery at a single institution between 2004 and 2009 and a complete clinical follow-up after surgery, to December 2015 (51 patients). We recorded the surgical outcome and the OS and PFS after surgery. An expert radiologist retrospectively evaluated the CT examinations and assigned rPCI following Sugarbaker's diagram. We correlated rPCI score with the results of surgery using ROC curve analysis and with OS and PFS using Kaplan-Meier curve.
Results: 32 (62.7%) patients had complete cytoreduction and 31 (60.8%) died due to the cancer; the median value of OS was 42 months (2-113) and of PFS was 15 months (7-56). The median value of rPCI was 11 (0-31). ROC curve analysis identified a PCI of 18 as significant threshold useful to predict complete surgical cytoreduction (p=0,05). Patients with a rPCI >16 has a shorter OS (p=0.02; Hazard Ratio=2.13). Similarly, patients with a rPCI >7 has a shorter PFS (p=0.03; HR=1.9).
Conclusion: rPCI on preoperative CT could help in predicting surgical outcome and seems to correlate positively with both OS and PFS. This may be helpful in better evaluating prognosis in patients with high FIGO stage who may have very different peritoneal extension.
Purpose: Assess diagnostic performance and accuracy of diffusion weighted imaging with background signal suppression (DWIBS) compared with 18Fluorodeoxyglucose positron emission tomography (FDG-PET) in detection of peritoneal metastases.
Methods and Materials: The study included 64 patients with ovarian and colonic malignancies and lymphomas with peritoneal deposits. DWIBS imaging was performed on a 1.5T scanner (Achieva, Philips Medical Systems, Netherlands) within mean period of two days after 18FDG-PET/CT study. b-value of 800 s/mm2 was used. 18FDG-PET/CT was performed on Philips, Gemini PET/CT system after intravenous injection of FDG. Axial images and coronal MIP images of MR and PET CT images were obtained and blinded review of PET/CT and DWIBS images was performed separately to avoid bias and all lesions detected were recorded. DW MR images were also assigned colour scale based on diffusion restriction and ADC values to display PET like images. The findings detected on DWIBS and PET/CT images were compared on a per-lesion basis using PET/CT as gold-standard.
Results: FDG-PET/CT revealed 380 malignant lesions in 64 patients. DWIBS showed 368 lesions with a detection rate of 96% compared with 18F FDG-PET/CT which was the gold standard. The study exhibited relatively good concordance between detection rates of DWIBS and FDG-PET/CT irrespective of primary pathology.
Conclusion: DWIBS allows for fast and robust imaging with low technical and operational efforts and is cost effective. Patients are comfortable as there is no injection of contrast and no exposure to radiation. DWIBS is as accurate as FDG PET/CT in assessment of peritoneal metastases.
Purpose: Assess the diagnostic accuracy of diffusion-weighted imaging (DWI) versus dynamic contrast-enhanced MRI (DCE-MRI) and T2 fast spinecho-weighted images (T2WI) in the preoperative staging of cervical carcinoma.
Methods and Materials: MRI pelvis using T2WI, DCE sequence and DWI were done for 50 cases with cancer cervix after initial biopsy confirmation for pre-management staging. In dynamic imaging we used T1 THRIVE (High Resolution Isotropic Volume Examination) technique that enables images acquisition at one pre-contrast, and six sequential post-contrast phases at 40, 80, 120, 160, 200 and 240sec. On DWI, scanning acquired by b values: 0, 500, 1000 and 1500.Analysis considered signal intensity (SI) at b1000 and the mean ADC values for the solid components of the masses. Surgical staging was the standard reference.
Results: The staging accuracy in all cases were 66.6% for T2WIs, 83.3% for DCE-MRI and 91% for DW-MRI.Upon assessment of parametrial infiltration, T2 detected 36%(n=18), DCE-MRI detected 35.4% (n=17) and DWI detected 32% (n=16). Assessment of locally advanced and advanced stages was comparable among T2WIs, DWI and DCE-MRI.While on assessing metastatic lymph nodes T2WIs was positive for lymphadenopathy in 15 cases (30%), DWI in 14 cases (28%) and DCE-MRI in 7 cases (14.6%).The overall accuracy of T2WIs, DW-MRI and DCE-MRI in staging cervical carcinomas was 74%, 88% and 83.3% respectively.
Conclusion: DWIs supported by conventional MRI data can improve detection of small tumours, increase overall accuracy, increase reader’s confidence in evaluation of parametrial infiltration and help in proper detection of locally advanced and advanced stages of cancer cervix.
Purpose: To show if combined MRI data can improve differential diagnostics of the histological type of cervical cancer (CC).
Methods and Materials: 91 patients with histologically verified diagnosis of CC (cervical squamous cell carcinoma, CSCC: 74; cervical adenocarcinoma, CAC: 17); for some grade of the tumour had been determined: SCCC: G1-6, G2-23, G3-24; CAC: G1-5, G2-5, G3-4) MRI (1.5 & 3T): T2WI TSE native/fat saturated (fs); DWIfs+ADC-maps; T1WIfs VIBE before and 4 min after i.v. injection of MRCA (1.0 M; 0.1 ml/kg); 35 series DCE T1WIfs TWIST after MRCA injection, time resolution=5s. Signal intensity (SI) and standard deviation (SD) in CC and the gluteus maximus muscle (15-30 pixels) were measured (SD,SI/SD for heterogeneity; SI(CC),SI(CC)/SI(muscle) for intensity).
Results: DCE: both types showed active SI increase at first 10-20 s after the occurrence of MRCA in pelvic arteries with following change of signal (inflection point), after relative SI (rSI) for DCE was calculated as rSI(t)= 100%* (SI(t)-SI(inflection))/SI(inflection);CAC was more heterogeneous at 110 s (p < 0.03) and had higher rSI (p < 0.003-0.04). Additionally areas under DCE curves were significantly different: CAC>CSCC, p < 0.005, CSCC curve had an additional inflection point before plateau at about 60-70th second while CAC continually accumulated MRCA, rSI (60s) had maximum accuracy (Sen/Sp=0.75/0.76) or Sen/Sp=0.27/0.95 T2WIfs: CAC was brighter, less heterogeneous (p<0.05). A regression model showed that T2WIfs and DCE combined data significantly differentiate CSCC and CAC: p<0.0015, Sen/Sp=0.80/0.85 or 0.67/0.95. SD of SI of the tumours on postcontrast T1WI-VIBE for CAC of G1 and G3 grades were significantly different from each other and from all other types of CC: SD(CAC(G1))>SD(SCCCs,CAC(G2))>SD(AC(G3)), p<0.02. Sp/Sen(CAC(CAC(G1))= 0.75/0.95;Sp/Sen(CAC(G3))= 1.00/0.83.
Conclusion: T1WI after MRCA injection, T2WI with fat saturation can be used to estimate histological type of CC.
Purpose: The aim of the study was to assess the diagnostic merits of each of the applicable MRI pulse sequences in the evaluation of tumour residual, recurrence or post treatment complications of cervical cancer aiming to assess the possibility of use of Diffusion weighted imaging instead of Contrast enhanced studies especially in patients with deteriorated renal functions.
Methods and Materials: Our study included 48 patients with histopathologically proven cancer cervix and a control group of 20 patients. All patients underwent post treatment Contrast Enhanced and Diffusion weighted MRI examinations to assess, confirm or exclude the presence of residual/recurrence mass lesions or post therapy complications.The reported MRI findings of both studies were correlated with histopathology results and/or with follow -up imaging.
Results: The sensitivity and specificity of DWI were 96.8 % and 100% respectively as compared to sensitivity and specificity of 87.5 % & 87.5% respectively calculated for contrast enhanced MRI studies. The mean ADC value of malignant post-treatment tumour (1+/- 0.18 x 10-3 mm2/sec) was significantly different than those of benign post -treatment changes (1.56 +/- 0.03 x 10-3 mm2/sec) and from values calculated for the control group (1.57 +/- 0.2 x 10-3 mm2/sec) (P value< 0.0001). The initial mean ADC value of complete responders were also significantly less than that of partial responders and those with recurrent disease (P value< 0.0001).
Conclusion: Quantitative DW imaging provides a better assessment tool than Contrast enhanced MRI in the characterisation of post-treatment changes.
Purpose: The purpose of this study was to investigate the clinical value of DW-MRI as early predictor of treatment response in patients with locally advanced cervical cancer (LACC) receiving preoperative chemoradiotherapy (CRT) through the measurement of ADCm. Correlation with histopathological findings.
Methods and Materials: This prospective study was approved by our Ethical Committee. 88 patients receiving CRT for LACC ( FIGO staging IB2 or more advanced) were enrolled and underwent 1.5 T pelvic MRI including DWI before CRT (baseline) and after 4 weeks (early). ADCm (b-values: 0-800 s/mm2), determined at baseline and early MRI, were compared with histopathologic findings after radical hysterectomy (good responders (R) vs. poor-R). For ADCm values, differences between good-R and poor-R were assessed by T-test. Its discriminatory capability for detection of good-R was compared with receiver operating characteristics (ROC) analysis.
Results: Early ADCm values were significantly higher for good-R than for poor-R (1.12 × 10-3 mm2/sec vs 1.04 × 10-3 mm2/sec; p=0.02), but no significant differences were observed for ADCm at baseline MRI (p= 0.40) between the two groups. The area under the ROC curve for prediction in treatment response was higher for early ADCm comparing to baseline ADCm (0.63 vs 0.52). A cut-off value of 1.10 × 10-3 mm2/sec was determined for early ADCm with a sensibility, specificity, VPP, VPN and accuracy of 74%, 67%, 50%, 85% and 90%.
Conclusion: Early evaluation of ADCm might provide useful information for predicting outcome and selecting high-risk patients for more aggressive therapy.
Purpose: This study aims to evaluate factors related to IVC filter (IVCF) retrieval rate and factors that affect success rate in retrieval attempts.
Methods and Materials: All patients with IVCF insertion in a district hospital from 10/2011 to 8/2016 were retrospectively included. Data regarding patient demographics, comorbidities, indications for and type of IVCF, IVCF tilting, and time to retrieval were obtained.
Results: 75 patients with 76 IVCF placement were included. IVCF retrieval rate was 30% (23 out of 76). IVCF retrieval success rate was 82% (23 out of 28 attempts). In patients with DVT and contraindicated for anticoagulation due to operation, the rate of IVCF retrieval was significantly higher than that of patients with IVCF placed for other indications (63% vs 17%, p<.05). Patients with malignancy had a significantly lower rate of IVCF retrieval compared to patients without malignancy (17% vs 39%, p<.05). IVCF angulation >15 degree was observed more in patients with failed retrieval attempt (3/5), compared to patients with successful retrieval (0/23). This difference was statistically significant (p<.05). There were no significant differences between successful and failed attempts in terms of age, sex, history of malignancy, placement route, type of filter, and time to filter retrieval.
Conclusion: IVCF retrieval is more likely in patients with IVCF placed for DVT and contraindication for anticoagulation due to operation, and in patients with no known malignancy. Unsuccessful IVCF retrieval attempts are more likely to occur in IVCF which are angulated.