Search

8012 Background: Early and sensitive detection of bone marrow disease and stratified patient management according to clinical risk can confer survival advantages in multiple myeloma (MM). Whole body MRI (WB MRI) and Fluorodeoxyglucose (FDG) PET/CT are included in international guidelines for imaging in patients with a suspected diagnosis of MM. However prospective studies comparing detection of MM by contemporary WB MRI as per recent MY-RADS consensus against FDG PET/CT are lacking. We report here protocol-defined endpoints from the prospective iTIMM (NCT02403102) study, comparing WB MRI and PET/CT, their relationship with serum and bone marrow estimates of disease burden, as well as molecular tumor characteristics. Methods: Patients with newly diagnosed MM or at first relapse planned to receive chemotherapy and autologous stem cell transplantation were enrolled in iTIMM. Matched baseline WB MRI and FDG PET/CT were performed and baseline clinical data including tumor genetics collected. Scans were double reported for presence of focal and diffuse disease by expert MRI and PET/CT radiologists, blinded to each other’s assessment. Paired methods were used to compare burden and patterns of disease on WB MRI compared to FDG PET/CT at baseline. Primary and secondary trial endpoints include relationship between post-treatment WB MRI response and progression-free survival, for which follow-up is ongoing. Exploratory endpoints include comparison of baseline WB MRI and PET/CT and their correlation with laboratory parameters, for which data is complete and reported here. Results: From May 2015 to March 2018, sixty patients (35 male; mean age 60 years) underwent baseline WB MRI as per MY-RADS consensus and FDG PET/CT. At least one focal lesion was detected in 50/60 patients (83.3%) by WB MRI and in 36/60 patients (60%) by PET/CT. WB MRI was more sensitive ( P< 0.05) across anatomical regions except for ribs and cervical spine. Four patients in our study showed two or more focal lesions ≥5 mm only on WB MRI but not PET/CT. All lesions detected by WB MRI but not PET/CT resolved in follow-up scans after treatment, excluding false positives. In 49/60 (81.7%) patients, diffuse disease was detected by WB MRI, compared to 10/60 (16.7%) by PET-CT; WB MRI was more sensitive across all anatomical areas ( P< 0.05). Plasma cell infiltration and paraprotein levels were significantly higher for patients with diffuse disease on WB MRI, but not on PET/CT. All genetically high-risk tumours, defined by t(4;14), t(14;16), del(1p), gain(1q) or del(17p), showed diffuse infiltration on WB MRI. Conclusions: WB MRI increases detection of focal and diffuse disease compared with FDG PET/CT, including improved detection of focal lesions meeting criteria for active disease as per International Myeloma Working Group diagnostic criteria, proposing it as a gold standard for tumor imaging in MM. Clinical trial information: NCT02403102.

e16513 Background: Ra-223 improves overall survival in patients with mCRPC, but is associated with increased risk of fracture when used in combination with abiraterone and prednisolone. We have assessed time to first new fracture in men with mCRPC receiving Ra-223 monotherapy, imaged using whole body MRI. Methods: In a prospective phase II study, patients with chemotherapy-naïve, bone mCRPC were randomised (1:1) to receive Ra-223 at 55 or 88 kBq/kg for 6 cycles, at 4-week intervals (REASURE; ISRCTN17805587). Whole body MRI was done at baseline, at cycles 2 and 4, and 1 month post-treatment. Subsequent imaging was done as clinically indicated. All scans were centrally reviewed to describe the time to first new fracture and characteristics of fractures. Time to new fracture was analysed using Kaplan Meier methods; comparisons of time to fracture between groups were based on log-rank tests. Results: 36 evaluable patients received a median of 6 cycles of Ra-223; scans were available for 35 patients. For these 35 patients, median age was 75 years (IQR: 73-80). 3 (9%) were receiving bone health agents at trial entry; 8 (23%) had received prior abiraterone or enzalutamide. 21 (60%) patients developed new fractures between randomisation and 2 years after completing treatment with radium-223. The spine was the most common location for fractures (14 patients). Median time to new fracture from randomisation was 10.8 months (95% CI: 6.2-18.2). Freedom from new fracture at 2 years post-randomisation was 21% (95% CI: 7 – 39%). Time to first new fracture was not associated with Ra-223 dose (55 kBq/kg: 10/18 patients experienced a fracture vs 88kBq/kg: 11/17; p = 0.93) or baseline use of bone health agents (3/3 vs 18/32; p = 0.09). Patients with Extent of Disease (EOD) ≥2 (defined as ≥6 metastatic sites/superscan) had earlier time to first new fracture compared to patients with EOD 1 (defined as < 6 metastatic sites) (11/18 vs 10/17; hazard ratio = 2.7, 95% confidence interval = 1.0 – 6.8; p = 0.04). Conclusions: New fractures are commonly seen on imaging in men with mCRPC during and after treatment with Ra-223 monotherapy. This observation supports treatment guidelines that recommend the use of bone health agents in men with mCRPC. Clinical trial information: ISRCTN17805587.

246 Background: Percutaneous biopsy of suspected liver metastases is a common practice for diagnostic purposes. Particularly, in the setting of oncological clinical trials, it is a relatively non-invasive method to obtain sufficient tissue for molecular analyses at regular set time points. However, various complications may occur, including seeding of the tumour along the biopsy tract. Only few reports exist on the actual incidence of seeding, on a limited number of tumor types. The aim of this study was to evaluate the technique’s safety and risk of seeding. Methods: All patients with an ultrasound or CT-guided liver biopsy between 2012-2016 were included. Medical records were reviewed retrospectively for post-biopsy complications and all follow-up imaging was re-assessed for the presence of seeding, defined as tumoral deposits in the biopsy needle tract. Results: In total 782 biopsies were performed in 550 patients (282 women, 268 men; mean age of 61 years), 43.9% (343/782) for trials and 56.1% (439/782) for diagnostic/molecular purposes, 93.7% (733/782) were diagnostic, revealing malignancy in 96.9% (710/733). Number of biopsies per patient ranged between 1 (n=387) to 7 (n=1), a co-axial system was used in 70.6% (552/782) and multiple passes in 29.4% (230/782). Complications were reported in 8.8% (69/782), more often pain (4.7%) and hypotension/vasovagal (2.3%). Admission and/or re-intervention were needed for more severe complications as bleeding (1.0%), sepsis/fever (1.1%), pulmonary embolism (0.3%) and pneumothorax (0.4%). Seeding was seen in 1.1% (8/782) of cases (2/44 melanoma, 1/11 GIST, 1/39 cholangiocarcinoma, 1/247 colorectal, 1/14 oesophagus, 1/97 breast, 1/31 prostate). Mean time for seeding was 208 days (range 43-469 d); mean post-biopsy survival time was 495 days in the seeding and 349 days in the non-seeding group. Conclusions: Percutaneous liver biopsy is a highly effective and safe method for tissue collection, with only a minimal risk of seeding.

401 Background: Microwave (MWA) and radiofrequency (RFA) ablation are commonly used for the treatment of unresectable liver metastases from colorectal cancer (CRC). MWA has a few theoretical advantages over RFA (active heating, less affected by heat sink effect, larger ablation zones, faster) however limited data exists on this field. Our aim was to analyse the safety and efficacy of MWA and RFA in the treatment of CRC liver metastases. Methods: All patients with unresectable CRC liver metastasis who were treated with RFA or MWA in a single center between March 2006 and December 2016 were included. Medical records and imaging studies were reviewed retrospectively for demographic data, to assess for local tumour recurrence and overall survival and to evaluate ablation-related adverse events. Results: 456 ablations were performed in 193 patients (123 men) with a median age of 66 (range 32-91). The majority of patients had a history of surgery of the primary tumour (n=170, 88,1%) and systemic chemotherapy (n=183, 94,8%). 68 patients had a history of liver surgery (35,2%). RFA was used in 343 procedures (75,2%) and MWA in 113 (24,8%). Median number of procedures per patient was 2 (range 1-10). Median lesion size was 17mm (range 3-80mm). Mean follow-up per lesion was 2.0 years in the RFA group and 1.3 years in the MWA group. Local tumour recurrence rate was 45% for RFA and 28% for MWA, with a hazard ratio of 0.6 in favour of MWA (95% CI 0.4-0.9). Two and five year overall survival rates from diagnosis of liver metastasis were 88% and 35% for RFA and 89% and 66% for MWA, respectively. Complications were reported in 43 of 456 procedures (9,4%); In 8,2% of RFAs (n=28) and 13,3% of MWAs (n=15). Complications classified as CTCAE grade 3 or higher were: haemothorax requiring surgery (n=1 for RFA), biloma (n=2 for RFA & MWA), biliary stricture (n=2 for RFA and n=1 MWA), liver failure (n=2 for RFA), liver abscess (n=1 RFA), pneumonia (n=2 for RFA) and pulmonary embolism (n=1 RFA). Segmental portal vein occlusion was only seen with MWA (n=2; 1,8%). One patient died due to multi-organ failure post RFA. Conclusions: MWA could achieve better local tumour control compared with RFA in the treatment of CRC liver metastasis, with slightly higher complication rate.

1520 Background: In the United Kingdom, current screening guidelines for TP53 germline mutation carriers solely recommends annual breast MRI, despite the wide spectrum of malignancies typically seen in this group. Some centres employ family specific screening with most centres having an “open-door” policy. Recent evidence suggests there may be a survival benefit with more intensive screening, including whole-body MRI. There are no published data on the psychological impact of such screening programmes. Methods: This study sought to investigate the role of one-off non-contrast whole-body MRI (WB MRI) in the screening of asymptomatic TP53 mutation carriers aged over 18 years. Clinical and psychological endpoints were assessed. Scans were read by radiologists blinded to participant carrier status. The incidence of malignancies diagnosed in TP53 mutation carriers against general population controls was calculated. The incidences of non-malignant relevant disease and irrelevant disease were measured, as well as the number of investigations required to determine relevance of findings. A series of questionnaires is used to assess the psychological impact of screening. Results: 44 TP53 mutation carriers and 44 population controls were recruited. In TP53 mutation carriers, six of 44 (13.6%, 95% CI: 5.2%-27.4%) participants were diagnosed with cancer during the study, all of which would be considered life threatening if untreated. Two were found to have two primary cancers. Two participants with cancer had abnormalities on the MRI which were initially thought to be benign (a pericardial cyst and a uterine fibroid) but transpired to be sarcomas. No controls were diagnosed with cancer. Fifteen carriers (34.1%, 95% CI: 20.5%-49.9%) and 7 controls (15.9%, 95% CI: 6.6%-30.1%) underwent further investigations following the WB MRI for abnormalities that transpired to be benign (p = 0.049). To date, there is no evidence of clinically significant adverse psychosocial effects. Conclusions: The cancer detection rate in this group justifies a minimum baseline non-contrast WB MRI in germline TP53 mutation carriers. Clinical trial information: NCT01737255.

613 Background: REG demonstrated efficacy in pre-treated mCRC pts. Lack of predictive biomarkers, potential toxicities and cost/effectiveness concerns highlight the unmet need for better patient selection. Methods: RAS mutant mCRC pts with biopsiable metastases were enrolled in this phase II trial. Tissue biopsies (6-12 cores) were obtained at baseline (BL), after 2 months if stable disease (SD) and at disease progression (PD). Dynamic contrast enhanced (DCE) MRI was acquired pre and at day 15 post-treatment. Median values of volume transfer constant (Ktrans) and enhancing fraction (EF) [K-trans*EF/100] were generated. Circulating tumour (ct)DNA was collected monthly until PD and tested for clonal RAS mutations by digital droplet PCR. PDOs derived from responders and non-responders pts were implanted orthotopically in the liver of mice and treated with REG for 5 days. Changes in tumour and fractional blood volume (fBV) were monitored by oxygen-enhanced MRI. Results: mCRC pts (n = 27) with paired MRI scans were analysed; a single target lesion per pt was chosen (25 liver and 2 pelvic metastases). Median K-trans*EF/100 product decrease was 58.2%. In the 23 analysable pts (4 received < 1 cycle of treatment due to toxicities), > 70% drop in K-trans*EF/100 (8/23) was associated with higher disease control rate (6/6 vs. 0/6, p = 0.048) measured by RECIST 1.1 at 2 months, improved progression free survival (PFS) [HR = 0.24 (0.07-0.86), p = 0.03], and 4-month PFS (58.3% VS 21.2%). Sequential tissue biopsies analysis confirmed reduction in CD31 in pts with K-trans*EF/100 drop. RAS mutant clones decay in ctDNA after 8 weeks of treatment was associated with better PFS [HR = 0.25 (0.08 - 0.83), p = 0.02] independently of K-trans*EF/100 drop. PDOs xeno-transplants treated with REG compared to controls had significant lower tumour fBV (4.5 VS 10.6, p = 0.03) and lower microvascular density measured by CD31 staining (4.3 VS 8.9, p = 0.02). Conclusions: Combining DCE MRI and ctDNA predicts depth and duration of anti-angiogenic response to REG monotherapy and may improve pt selection with potential health/economic implications. Clinical trial information: 201400357951.

580 Background: REG that has single agent efficacy in patients (pts) with refractory mCRC, is known to have anti-angiogenic activities. The benefit of REG in unselected pts is modest. Thus, the identification of predictive biomarkers is critical for treatment stratification. PROSPECT-R study aims to identify genetic and radiological mechanisms of primary and acquired REG resistance in RAS mt mCRC patients. Methods: Multiparametric MRI studies including dynamic contrast enhancement (DCE)-MRI and diffusion weighted imaging (DWI) were acquired pre- and at day 15 post-treatment on a 1.5T Siemens Avanto MR scanner. Regions of interest of the entire chosen target metastatic lesion were drawn by a senior radiologist and the following imaging parameters were generated: volume transfer constant (Ktrans) derived from a pharmacokinetic analysis based on the extended Kety model and apparent diffusion coefficient (ADC) calculated using a mono-exponential fitting algorithm; median values of ADC and Ktrans were reported. Results: The first seven enrolled pts were analysed; a single target lesion per patient was chosen (5 liver and 2 pelvic metastases). At day 15 post treatment, a marked decrease (68-81%) of median tumour Ktrans was observed in 4 out of 7 pts; the remaining 3 patients showed no significant median Ktrans change (-36 to + 17%). Overall, the cohort average Ktrans decreased from 0.17 to 0.07 min-1(58%). No significant ADC changes were observed at day 15. Of the 4 pts with Ktrans reduction on day 15, 1 achieved RECIST 1.1 partial response (38% reduction in target lesions), 2 had stable disease and 1 progressed, based on CT assessments performed at 8 weeks of REG therapy; 3/7 pts with no Ktrans change progressed within 2 months of initiating REG. When modified Choi criteria were applied, 3/4 pts with Ktrans reduction were classified as responders. Conclusions: REG may have early anti-angiogenic affects; DCE-MRI could be a potential predicting biomarker in pts treated with REG. Further analysis within PROSPECT-R may elucidate genetic biomarkers to validate these findings. Clinical trial information: EUDRACT No: 2014- 003579-51.

5042 Background: Surgical evaluation of lymph node (LN) metastases has served as the gold standard for comparison of imaging modalities in locally advanced cervical cancers. It has been suggested t...

5035 Background: Ferumoxtran-10 MRI has been evaluated for lymph node (LN) metastasis. Objective of the study was to determine utility of ferumoxtran-10 MRI in detection of LN metastasis in advance...

TPS154 Background: The anti-angiogenic tyrosine kinase inhibitor, sunitinib has been shown to significantly improve progression-free (PFS) and overall survival in RCC (Motzer 2009). However assessi...

378 Background: Surgeons rely on imaging to plan exenterative pelvic surgery for locally advanced primary and recurrent rectal cancer. Accurate local staging is vital in this process, as it provides information about the extent of the disease, facilitating the planning of the operation required to achieve complete resection. DW-MR imaging is a functional radiological modality that can provide indirect informationabout the water proton mobility within biologic tissue, without the need of a contrast agent. The aim of the present study was to assess the diagnostic accuracy and added value of diffusion DW-MRI in detecting colorectal tumor invasion into seven intrapelvic compartments for planning exenterative pelvic surgery. Methods: Thirty three consecutive patients were preoperatively staged using DW-MRI to undergo exenterative surgery for locally advanced primary (n=12) and recurrent (n=21) colorectal pelvic cancer. Two radiologists reported tumor invasion for each of the seven anatomic surgical resection compartments and were blinded to histopathology and intraoperative reference standards. Accuracy, sensitivity, specificity and predictive values were calculated for the seven intrapelvic compartments. Interobserver agreement was assessed using Cohen's Kappa (k) coefficient for each compartment. Results: The sensitivity of DW-MRI when used alone was low for all the compartments except the central (92%). Its specificity was very high for all the compartment (≥89.5%). Combining DW-MRI with conventional MRI increased the sensitivity for the lateral compartment by 8.3% to 100% and the specificity for the anterior compartment below the peritoneal reflection by 9.1% to 90.9%. The overall incremental value for the lateral compartment was 3%. The interobserver agreement was either good or very good (k≥0.669; p Conclusions: DW-MRI can improve the diagnostic accuracy of conventional MRI when staging patients with advanced colorectal pelvic cancer. Diffusion images are cheap and easy to obtain. Therefore DW-MRI should always be used in combination with conventional MRI when considering patients for exenterative pelvic surgery. No significant financial relationships to disclose.