J. Wu: None. D.S. Lakomy: None. B. Fellman: None. M.P. Salcedo: IGCS-International Gynecologic Cancer Society, Brazilian Federation of Gynecology and Obstetricians (FEBRASGO)/BRAZIL (Federação Brasileira de Ginecologia e Obstetricia, FEBRASGO). A. Sood: Consultant; Kiyatec, Astra Zeneca. Stock; BioPath. sponsored research; M-Trap. A. Jhingran: American Board of Radiology. A.H. Klopp: Research Grant; MD Anderson Cancer Center SPORE Grant. R.B. Iyer: None. C. Jimenez: None. L. Colbert: None. K. Schmeler: None. P.J. Eifel: Travel Expenses; National Cancer Center Network. Stock; Apple Computer. L.L. Lin: Employee; VA Hospital. Research Grant; AstraZeneca. Travel Expenses; AstraZeneca.
AIM: To assess the ability of dual-energy computed tomography (DECT) to distinguish benign from malignant ovarian tumours (OTs). MATERIALS AND METHODS: Following approval of the institutional review board, the institutional database was mined for treatment-naive patients who underwent primary cytoreduction for OT. Thirty-seven patients were included and divided into those with benign OTs (n = 11) and malignant OTs (n = 26), including high-grade (n = 20) and low-grade (n = 6) malignant OTs. Advanced processing and region of interest delineation on the ovarian mass were performed using the preoperative staging DECT examination using the Advantage Workstation. The pixel-level data of the CT attenuation values at 50, 70, and 120 keV and the effective atomic number (Z(eff)), water content (WC), and iodine content (IC) in the ovarian mass were recorded. The Wilcoxon rank-sum test was used to compare CT attenuation data at different voltages, Z(eff), and WC and IC levels between benign and malignant OTs and between high- and low-grade malignant OTs. Simple logistic regression was used to correlate the imaging characteristics with malignant status and grade. RESULTS: Malignant OTs had significantly higher Z(eff) and IC compared with benign OTs. The threshold values for the diagnosis of malignant OT were IC >= 9.74 (100 mu g/cm(3)) with 81% sensitivity and 73% specificity and Z(eff) >= 8.16 with 85% sensitivity and 73% specificity. High-grade OTs had significantly higher WC compared with low-grade OTs, and a threshold of >= 1,013.92 mg/cm(3) differentiated them with 80% sensitivity and 83% specificity. CONCLUSION: DECT may be a tool to help distinguish malignant and benign OTs and predict tumour grade. (C) 2020 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Objective. To describe the clinical outcomes associated with the use of checkpoint inhibitor therapy in recurrent ovarian malignancy. Methods. Women with recurrent ovarian cancer treated with an immune checkpoint inhibitor between 1/ 2012 and 8/2017 were included. RECIST criteria determined disease status, and immune related adverse events (irAE) were graded per trial protocols. Predictors of response, irAE, progression free survival (PFS) and overall survival (OS) were investigated. Results. Forty-four women were included with a median age of 53 years, median of 4 prior lines of chemotherapy, and most commonly high grade serous pathology (59.1%). 3 patients had partial response and 3 had pseudoprogression, for a response rate of 142%. In subset analysis of high grade serous (HGSOC) pathology, platinum sensitivity at time of checkpoint inhibitor therapy was correlated with response (p = 0.01). There were 28 grade 3/4 irAEs in 21 patients (47.7%). Combination therapy rather than monotherapy predicted irAE (OR 5.7, CI 1.6-20.9, p = 0.02). The most common severe irAE was elevation in hepatic or pancreatic enzymes in 12 total patients (13.6% each). Interestingly, the number of genes mutated was protective from hepatic/pancreatic AE (p = 0.02). Conclusions. While response rate was similar to prior literature, in patients with HGSOC platinum sensitivity at time of checkpoint inhibitor initiation was correlated to response. Grade 3/4 hepatic and pancreatic enzyme elevations were more common in ovarian cancer patients than has been previously reported in other tumor types. The number of genes mutated was inversely correlated to risk of this type of irAEs but not to total irAEs. (C) 2018 Elsevier Inc. All rights reserved.
AIM: To identify potential magnetic resonance imaging (MRI) biomarkers to predict the aggressiveness of endometrial cancer. MATERIALS AND METHODS: Seventy-one patients with endometrial cancer who underwent MRI staging were analysed retrospectively. The signal intensity (SI) of the tumours was assessed on sagittal T2-weighted imaging (WI) and sagittal T1WI sequences). The depth of myometrial invasion, tumour grade and subtype, lymphovascular invasion, and microsatellite stability status were assessed histopathologically, and these findings were compared with MRI findings using logistic regression. The log-rank test was used to assess differences in survival among groups defined by different MRI measurements. RESULTS: Tumours with qualitative higher signal than that of normal myometrium on the late T1WI DCE image sequences were more likely to have lymphovascular space invasion (p<0.001). Tumours that had a higher SI tumour ratio (T1 post-contrast arterial/T1 precontrast) had a higher chance of being microsatellite stable (odds ratio 2.36). The SI ratio of the tumour to the myometrium showed that lower T2 tumour/T2 myometrial ratio correlated with >= 50% depth of myometrial invasion as determined by imaging (p=0.006). Endometrial tumours showing a SI of >209 on delayed T1WI sequences had longer recurrence-free survival than those with tumours showing a SI <= 209 (p=0.014). Tumour subtype and grade were not associated with MRI findings. CONCLUSION: The SI of endometrial cancer on MRI may be used to predict the aggressiveness of the tumour and microsatellite stability status. Further studies are needed to confirm these findings. (C) 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
AIM:To determine the ability of magnetic resonance imaging (MRI) in detecting tumour-free margins from the internal os (IO). MATERIALS AND METHODS:A database search yielded 79 women with early-stage cervical cancer who underwent radical hysterectomy and preoperative MRI. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of MRI in assessment of ≤5 and >5 mm IO involvement were calculated with histopathological surgical specimen findings considered to be the reference standard. A main and subset analysis was performed. The subset analysis included only those patients who would have been considered for radical trachelectomy. RESULTS:For predicting a distance between the tumour and the IO of ≤5 mm, MRI had a sensitivity of 73%, a specificity of 98.3%, a PPV of 95%, a NPV of 88.1%, and an accuracy of 89.8% for the main analysis, and sensitivity of 81.8%, a specificity of 93.2% a PPV of 69.2% a NPV of 96.5% and an accuracy of 91.4% for the subset analysis. CONCLUSION:MRI has high specificity, NPV, and accuracy in detecting tumour from the IO, making MRI suitable for treatment planning in patients desiring trachelectomy to preserve fertility.
Objectives: To determine the ability of computed tomography (CT) to characterize peritoneal disease distribution and determine resectability in a prospective cohort of patients with advanced ovarian cancer (OC). Methods: All patients with suspected stage 3 or 4 OC undergo a triage algorithm to determine who is likely to achieve R0 resection. Each patient undergoes preoperative CT followed by diagnostic laparoscopy (LS) according to the Anderson algorithm (those with Fagotti score >8 are triaged to neoadjuvant chemotherapy). Trained radiologists, blinded to the clinical outcome of the patient, prospectively evaluated each CT scan. A structured radiology report (SRR) outlines the presence or absence of disease on the liver, mesentery, gastric wall, subdiaphragm, supracolic or infracolic omentum, and bowel. Agreement between individual and composite SRR and LS points of disease was determined using the k statistic. Concordance between surgeons and radiologists with respect to assessment of the extent of disease was estimated. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the SRR with respect to LS was evaluated. Results: A total of 120 patients who were suspected to have advanced OC had both SRR and LS. Composite radiology scores correlated with surgeon assessment of disease distribution 66.7% (95% CI 57.5%–75%) of the time (k = 0.35 [95% CI 0.21–0.49]), with disease being classified as resectable, based on composite LS and SRR scores, 73.3% and 46.7% of the time, respectively. The sensitivity, NPV, specificity, and PPV of the SRR were 87.5%, 92.9%, 59.1%, and 43.8%, respectively. Correlation of LS and SRR for specific sites (particularly opposing surfaces, i.e., subdiaphragm and liver or bowel and mesentery) of disease demonstrated poor agreement, with k < 0.4 for all. In general, the SRR overestimated disease presence compared with LS in every area except bowel infiltration, for which the SRR appeared to underestimate the disease presence. Conclusions: CT imaging is useful for gross disease detection but is limited in determining overall resectability of advanced OC. Although SRR is important for overall patient care, alternative modalities are important to determine disease resectability. LS for peritoneal disease assessment remains a critical component of personalizing surgical therapy for advanced OC.
Target definition is crucial for the definitive treatment of cervix cancer with intensity modulated radiation therapy (IMRT). Consensus guidelines for the clinical target volume (CTV) have been established using magnetic resonance imaging (MRI). However, treatment planning MRI is not as readily available as computed tomography (CT) and image fusion carries inherent uncertainties. Therefore, the aim of this study was to evaluate the variability in CTV definition amongst physicians experienced in IMRT for the definitive treatment of cervical cancer in preparation for a collaborative NRG clinical trial. A consensus working group that had participated in prior CTV definition was convened to contour on two treatment planning CT scans. Observers were blinded to the corresponding MRI scans. One case was an early cervical cancer and the other a loco-regionally advanced case. Clinical vignettes for the two cases were distributed and each participant was asked to draw CTV contours which included a CTV1 contour for the uterus/cervix and a CTV 2 contour for the vagina/parametria. Participants contoured on CT images of the pelvis using their own treatment planning software. Nodal CTV contours have been well described and were not included in this study. The CTV contours were then analyzed for consistency and clarity of target delineation using an expectation-maximization algorithm for simultaneous truth and performance level estimation (STAPLE, CERR), with Kappa statistics as a measure of agreement between observers. Contoured datasets were merged and analyzed for agreement. CTV1 contours showed almost perfect agreement (Kappa > 0.8), while CTV2 showed moderate agreement (0.4 < Kappa < 0.6) among observers (see Table 1). Agreement among the experienced gynecologic radiation oncologists was excellent for CTV delineation in two representative intact cervical cancer cases. Consensus demonstrated near perfect agreement for the uterus and cervix and moderate agreement for the vagina and parametria. The variability seen in vaginal contours was primarily due to the vaginal length included in the CTV. The value of this data, building on previously published guidelines for IMRT in the post-operative setting and MRI guidance in the intact setting, provides clinically valuable information to promote safety and quality among radiation oncologists treating cervical carcinoma. Furthermore, this atlas will be used for future trials utilizing IMRT for the definitive management of intact cervical cancer.Tabled 1Abstract 28; Table 1STRUCTURE MEASURECase 1Case 2CTV1CTV2CTV1CTV2Vol. Mean/Min/Max (SD in cc)225.1/189.4/259.3 (22.4)166.4/96.4/238.0 (49.4)322.3/283.6/348.2 (21.1)197.5/71.2/365.1 (75.3)STAPLE/Intersection/Union Vol. (cc)225.3/152.2/305.6224.9/18.2/416.0332.0/226.2/423.3253.1/10.56/596.5Kappa0.820.560.870.50Conformity Index (Mean Vol./Union Vol.)0.740.400.760.33 Open table in a new tab
OBJECTIVE:To evaluate treatment outcomes for patients with vulvar cancer with grossly positive pelvic lymph nodes (PLNs). METHODS:From a database of 516 patients with vulvar cancer, we identified patients with grossly positive PLNs without distant metastasis at initial diagnosis. We identified 20 patients with grossly positive PLNs; inclusion criteria included PLN 1.5cm or larger in short axis dimension on CT/MRI (n=11), FDG-avid PLN on PET/CT (n=3), or biopsy-proven PLN disease (n=6). Ten patients were treated with chemoradiation therapy (CRT), 4 with RT alone, and 6 with various combinations of surgery, RT or CRT. Median follow-up time for patients who had not died of cancer was 47months (range, 4-228months). RESULTS:Mean primary vulvar tumor size was 6.4cm; 12 patients presented with 2009 AJCC T2 and 8 with T3 disease. All patients had grossly positive inguinal nodes, and the mean inguinal nodal diameter was 2.8cm. The 5-year overall survival and disease specific survival rates were 43% and 48%, respectively. Eleven patients had recurrences, some at multiple sites. There were 9 recurrences in the vulva, but no isolated nodal recurrences. Four patients developed distant metastasis within 6months of starting radiation therapy. CONCLUSIONS:Aggressive locoregional treatment can lead to favorable outcomes for many patients with grossly involved PLNs that is comparable to that of grossly involved inguinal nodes only. We recommend modification of the FIGO stage IVB classification to more accurately reflect the relatively favorable prognosis of patients with PLN involvement.
AIM: To evaluate utility of magnetic resonance imaging (MRI) in determining the primary site of endometrial versus cervical tumours when the biopsy results are inconclusive.MATERIAL AND METHODS: Forty-eight patients who underwent a total hysterectomy for unknown primary adenocarcinoma of the uterus after endometrial and/or endocervical biopsies were included in the study. The 48 available pelvic MRI images were reviewed by two body radiologists independently and jointly to resolve discordance, blinded to any clinical and pathological information. The clinical information and histopathology were reviewed by a radiology fellow and a pathologist specializing is gynaecological oncology. The final surgical pathology was used as the reference standard to confirm the origin of the primary tumour.RESULTS: The radiologists correctly identified the primary sites in 85% of the cases (41/48). There was substantial agreement between the two readers (kappa statistics = 0.79). Both radiologists found that T2 and dynamic T1-weighted images (WI) were most helpful in making the diagnosis, and dynamic T1WI helped resolve problematic cases. The sensitivity and specificity for detecting endometrial and cervical cancer on MRI were 88% and 88% and 75% and 93%, respectively.CONCLUSION: MRI has high sensitivity and specificity in determining the origin of the primary endometrial versus cervical tumours when endometrial/endocervical curettage is inconclusive. (C) 2013 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Background: Obstructive coronary disease is a leading cause of out of hospital ventricular fibrillation (OOH VF) arrest. However, in the absence of ST segment elevation there is no consensus on the role of immediate diagnostic angiography. Methods: Retrospective analysis identified 154 consecutive patients admitted following OOH VF arrest between July 2005 and December 2010. 111 (73%) were investigated with coronary angiography with the remainder treated conservatively due to poor neurological status. Sixty-four patients had ST elevation (STE), 63 undergoing immediate angiography. Forty-seven patients with no ST elevation (NSTE) also underwent angiography (7 immediate and 40 deferred until neurological recovery). ECG findings were correlated with coronary anatomy as well as survival to hospital discharge. Results: Acute arterial occlusion was found in 88% of patients with STE and 11% of patients with NSTE.Tabled 1No ST elevationST elevationNo acute arterial occlusion427Acute arterial occlusion557Patients undergoing angiography following OOH VF arrest P < 0.01. Open table in a new tab Patients undergoing angiography following OOH VF arrest P < 0.01. An occluded artery was associated with poorer LV function. Survival to discharge was 87.5% with STE and 72% with NSTE. Absence of ST elevation predicted arterial patency in OOH VF arrest survivors with sensitivity 86% and specificity 92%. Conclusion: A significant minority of patients with OOH VF and NSTE still has an acute arterial occlusion. Immediate angiography should also be considered in this cohort, especially in the absence of contraindications such as advanced age, prolonged down time, obvious serious neurological insult or other serious comorbidity.
Purpose/Objective(s)Increasingly, conformal radiation treatment fields are being used in treatment of cervical cancer (CCa), making accurate delineation of regions at risk critical. Currently, the nodal basins at risk in CCa are defined by the location of the normal pelvic and paraaortic lymph nodes (LNs) as well as surgical data defining the distribution of LN metastasis. To define the spatial distribution of LN metastasis on CT imaging, we generated a map of 2-deoxy-2-[F-18] fluoro-d-glucose (FDG) avid LNs on PET/CT scans in patients with CCa.Materials/MethodsWe identified 50 patients with stage I-IV (I-16, II-20, III-11, IV-3) CCa treated in our department with definitive chemoradiation from 2006-2009 who had FDG avid LNs on pretreatment PET/CT. Forty-one patients were consecutively identified. Nine additional patients with positive paraaortic LNs were added to the analysis. PET/CT images were transferred to treatment planning system and involved LNs were contoured on individual scans. The nodal contours were mapped to a single CT image set using deformable image registration. The mapped contours were edited as necessary by a diagnostic radiologist and radiation oncologist specializing in gynecologic malignancies. A volume overlapping probability map was created to indicate the spatial likelihood distribution of positive LNs throughout the pelvic and paraaortic lymphatics. Percentage of positive LNs along major blood vessels (defined as within a 3 cm radial margin of vessel) and in presacral and perirectal regions were collected for the initial 41 patients.ResultsWe identified 190 PET positive LNs from 50 patients (122 LNs from initial 41 patients, 68 from the additional 9 patients). Deformable image registration was found to be reasonably accurate to map LNs to a template patient, in spite of drastically different body mass indices. The highest density of nodes was found along bilateral external iliac (EI) vessels, followed by the left common iliac (CI) vessels. A lower density of nodes was identified along presacral (4), perirectal (2), and medial inguinal (2) regions outside of conventionally described nodal drainage regions. The highest density of CI lymph nodes was between the CI vein and psoas muscle, extending posteriorly between the muscle and sacrum. The distribution for initial 122 LNs was as follows: left (L) and right (R), EI 43 (35%) and 38 (31%), respectively; L and R internal iliac 3 (2.5%), and 3 (2.5%); L and R CI 18 (15%,) and 3 (2.5%); presacral 2 (1.5%); perirectal 3 (2.5%); and paraaortic 9 (7%).ConclusionsOur study is the first to provide probabilistic distribution of involved LNs in CCa. This analysis provides a template to define LNs at risk, which can be used to improve target definition for conformal therapy. We identified regions anterior and posterior to the psoas muscle, which require careful attention during treatment planning. Purpose/Objective(s)Increasingly, conformal radiation treatment fields are being used in treatment of cervical cancer (CCa), making accurate delineation of regions at risk critical. Currently, the nodal basins at risk in CCa are defined by the location of the normal pelvic and paraaortic lymph nodes (LNs) as well as surgical data defining the distribution of LN metastasis. To define the spatial distribution of LN metastasis on CT imaging, we generated a map of 2-deoxy-2-[F-18] fluoro-d-glucose (FDG) avid LNs on PET/CT scans in patients with CCa. Increasingly, conformal radiation treatment fields are being used in treatment of cervical cancer (CCa), making accurate delineation of regions at risk critical. Currently, the nodal basins at risk in CCa are defined by the location of the normal pelvic and paraaortic lymph nodes (LNs) as well as surgical data defining the distribution of LN metastasis. To define the spatial distribution of LN metastasis on CT imaging, we generated a map of 2-deoxy-2-[F-18] fluoro-d-glucose (FDG) avid LNs on PET/CT scans in patients with CCa. Materials/MethodsWe identified 50 patients with stage I-IV (I-16, II-20, III-11, IV-3) CCa treated in our department with definitive chemoradiation from 2006-2009 who had FDG avid LNs on pretreatment PET/CT. Forty-one patients were consecutively identified. Nine additional patients with positive paraaortic LNs were added to the analysis. PET/CT images were transferred to treatment planning system and involved LNs were contoured on individual scans. The nodal contours were mapped to a single CT image set using deformable image registration. The mapped contours were edited as necessary by a diagnostic radiologist and radiation oncologist specializing in gynecologic malignancies. A volume overlapping probability map was created to indicate the spatial likelihood distribution of positive LNs throughout the pelvic and paraaortic lymphatics. Percentage of positive LNs along major blood vessels (defined as within a 3 cm radial margin of vessel) and in presacral and perirectal regions were collected for the initial 41 patients. We identified 50 patients with stage I-IV (I-16, II-20, III-11, IV-3) CCa treated in our department with definitive chemoradiation from 2006-2009 who had FDG avid LNs on pretreatment PET/CT. Forty-one patients were consecutively identified. Nine additional patients with positive paraaortic LNs were added to the analysis. PET/CT images were transferred to treatment planning system and involved LNs were contoured on individual scans. The nodal contours were mapped to a single CT image set using deformable image registration. The mapped contours were edited as necessary by a diagnostic radiologist and radiation oncologist specializing in gynecologic malignancies. A volume overlapping probability map was created to indicate the spatial likelihood distribution of positive LNs throughout the pelvic and paraaortic lymphatics. Percentage of positive LNs along major blood vessels (defined as within a 3 cm radial margin of vessel) and in presacral and perirectal regions were collected for the initial 41 patients. ResultsWe identified 190 PET positive LNs from 50 patients (122 LNs from initial 41 patients, 68 from the additional 9 patients). Deformable image registration was found to be reasonably accurate to map LNs to a template patient, in spite of drastically different body mass indices. The highest density of nodes was found along bilateral external iliac (EI) vessels, followed by the left common iliac (CI) vessels. A lower density of nodes was identified along presacral (4), perirectal (2), and medial inguinal (2) regions outside of conventionally described nodal drainage regions. The highest density of CI lymph nodes was between the CI vein and psoas muscle, extending posteriorly between the muscle and sacrum. The distribution for initial 122 LNs was as follows: left (L) and right (R), EI 43 (35%) and 38 (31%), respectively; L and R internal iliac 3 (2.5%), and 3 (2.5%); L and R CI 18 (15%,) and 3 (2.5%); presacral 2 (1.5%); perirectal 3 (2.5%); and paraaortic 9 (7%). We identified 190 PET positive LNs from 50 patients (122 LNs from initial 41 patients, 68 from the additional 9 patients). Deformable image registration was found to be reasonably accurate to map LNs to a template patient, in spite of drastically different body mass indices. The highest density of nodes was found along bilateral external iliac (EI) vessels, followed by the left common iliac (CI) vessels. A lower density of nodes was identified along presacral (4), perirectal (2), and medial inguinal (2) regions outside of conventionally described nodal drainage regions. The highest density of CI lymph nodes was between the CI vein and psoas muscle, extending posteriorly between the muscle and sacrum. The distribution for initial 122 LNs was as follows: left (L) and right (R), EI 43 (35%) and 38 (31%), respectively; L and R internal iliac 3 (2.5%), and 3 (2.5%); L and R CI 18 (15%,) and 3 (2.5%); presacral 2 (1.5%); perirectal 3 (2.5%); and paraaortic 9 (7%). ConclusionsOur study is the first to provide probabilistic distribution of involved LNs in CCa. This analysis provides a template to define LNs at risk, which can be used to improve target definition for conformal therapy. We identified regions anterior and posterior to the psoas muscle, which require careful attention during treatment planning. Our study is the first to provide probabilistic distribution of involved LNs in CCa. This analysis provides a template to define LNs at risk, which can be used to improve target definition for conformal therapy. We identified regions anterior and posterior to the psoas muscle, which require careful attention during treatment planning.
To determine the impact of implant quality on overall and disease-free survival; local, regional, para-aortic, and distant failure; and acute and late toxicity among patients with cervical cancer treated with brachytherapy and concurrent chemotherapy on RTOG prospective trials 0116/0128. All patients had individual brachytherapy parameters assessed for acceptability for each implant: symmetry of ovoids to tandem, displacement of ovoids in relation to cervical os, position of tandem in midpelvis on lateral film, tandem bisecting ovoids on lateral film, and appropriateness of packing. Multivariate (MV) Cox proportional hazard models for disease-free survival (DFS); overall survival (OS); and local (LR), distant (DR), and regional (RR) recurrence were constructed for each parameter. FIGO stage and pelvic/iliac and para-aortic nodal status were adjusted for in each model. A total of 109 patients were enrolled and included in this analysis. The median overall follow-up time is 24 months. Patients with unacceptable symmetry of ovoids to tandem have a significantly higher risk of LR than patients in the acceptable group (HR, 2.61; 95% CI, 1.05, 6.45; p = 0.04). Patients with unacceptable displacement of ovoids in relation to cervical os have a significantly greater risk of DFS (MV HR, 2.12; 95% CI, 1.16, 3.89; p = 0.015) and a higher risk of LR (HR, 2.54; 95% CI, 1.11, 5.80; p = 0.03) than patients in the acceptable group. Appropriateness of packing was associated with an increased risk of DFS (HR, 1.95; 95% CI, 1.08, 3.55; p = 0.03) and DR (HR, 2.07; 95% CI, 1.03, 4.20; p = 0.04). Position of tandem in midpelvis on lateral film and tandem bisecting ovoids on lateral film were not associated with any outcome on MV. None of the assessed parameters affected late ≥Grade 3 gastrointestinal or genitourinary toxicity. The quality of the brachytherapy implant is an important predictor of outcome. This is the first prospective multi-institutional trial to demonstrate that implant quality affects relapse rates. These parameters should be assessed for each brachytherapy implant and repacking is recommended when symmetry, displacement, or packing of the implant is not optimal. These parameters should be assessed for each brachytherapy implant and repacking is recommended when symmetry, displacement, or packing of the implant is not optimal.
5549 Background: VEGF blockade has proved to be a promising therapeutic strategy in solid tumors, including ovarian. Aflibercept, a novel fusion protein consisting of the extracellular domains of VEGFR1/2 binds VEGF A, B and PlGF. Aflibercept has been studied as a single agent in heavily pretreated ovarian cancer patients. We hypothesized that the combination of aflibercept and docetaxel could be safely administered to women with recurrent ovarian cancer. Correlative biomarker and imaging studies of anti-angiogenesis targeting, pharmacokinetics (PK) and preliminary efficacy were additional objectives. METHODS Eligible patients had measurable, recurrent disease with no more than 3 prior chemotherapeutic regimens. Study design was a "lead-in" phase I trial; cycle 0, administered aflibercept IV as a single agent in 1 of 3 dose levels (2, 4, or 6 mg/kg) in a 3+3 design. Aflibercept was given in subsequent cycles with docetaxel (75 mg/m2); each cycle was 21 days. Correlative studies in cycle 0 were: PK (single agent), circulating endothelial cells and precursors (CEC, CEP), and imaging FDG-PET, DCE-MRI (baseline, day 2 and day 21). Efficacy evaluation (RECIST) was conducted q2 cycles of combination therapy. RESULTS Nine patients were recruited, 3 at each dose level. All are evaluable. No DLTs were observed in cycles 0 and 1; The most common hematological toxicities were myelosuppression (1 Grade 4 ANC) and anemia (Grade 2). Non-hematological toxicities (Gr 3) included headache, hypertension, fatigue and ulceration. One patient each with hypertension and ulceration lead to treatment discontinuation after 4 and 13 cycles, respectively. FDG-PET defined SUVmax in target lesions within 25% of baseline in 7 patients; 2 others had >25% increase at 48 hours post treatment. CECs, CEPs, DCE-MRI and PK are being analyzed. Confirmed PR was observed in 2 (22%) with 1 additional near PR. Median number of cycles: 5 (range 3-15). All have now progressed, median time to progression: 15 weeks. CONCLUSIONS Aflibercept can be safely administered at 6 mg/kg with docetaxel repeatedly in this population of recurrent ovarian cancer patients. Preliminary efficacy supports phase II study, which is ongoing. No significant financial relationships to disclose.