CTiR 11 - Clinical Trials in Radiology 3

Purpose:

A considerable proportion of ischemic stroke patients do not recover despite fast and complete recanalisation after endovascular therapy (EVT). It is unknown whether periprocedural antithrombotic therapy can improve clinical outcome. This study assesses the effect of acetylsalicylic acid (ASA) and unfractionated heparin (UFH), alone, or in combination, in patients who undergo EVT.

Methods and materials:

MR CLEAN-MED is a multicenter, prospective, randomised, open-label, blinded-endpoint trial using a 2x3 factorial design. We aim to enroll 1500 patients. Study interventions are intravenous treatment with ASA (300 mg) and/or UFH in a low-dose (5000 IU bolus, followed by 500 IU/hour x 6 hours) or moderate-dose (5000 IU bolus, followed by 1250 IU/hour x 6 hours). Primary outcome is the score on the modified Rankin Scale at 90 days. Safety endpoints include the occurrence of symptomatic intracerebral hemorrhage (sICH) and mortality.

Results:

Sixteen Dutch and five French sites participated in the study. By October 2019, 270 patients were randomised. Following on receipt of the 4^th safety report in April 2019, after inclusion of 132 patients, the data safety monitoring board recommended to stop enrollment in the moderate-dose UFH arm based on an increased sICH risk compared to the other treatment arms (22.7% vs. 5.7%; OR 4.88 95%CI 1.55-15.55[Table]). No patients were included in the moderate-dose UFH arm after receipt of this recommendation.

Conclusion:

In MRCLEAN MED the effect of ASA and/or UFH during EVT on functional outcome will be evaluated further and the inclusion of patients in the ASA and low-dose UFH arms continues. Periprocedural treatment with moderate-dose UFH (5000 IU bolus, followed by 1250 IU/hour during 6 hours) is associated with increased sICH risks and should be avoided.

Limitations:

N/A

Ethics committee approval

Medisch Ethische Toetsings Commissie Erasmus MC (Medical Ethical Committee Erasmus MC), 09/10/2017, ref: MEC-2017-366.

Funding:

None.

Purpose:

The RETomo trial was a two-arm test-and-treat randomised controlled trial comparing tomosynthesis (DBT) plus digital mammography (DM) versus DM alone for breast cancer screening. We investigate interim analysis on recall, detection, interval cancer rate in the first round by breast density.

Methods and materials:

Women (45-70yo) presenting for a screening mammography in ReggioEmilia, and previously screened with DM, were randomised to the experimental arm (DBT+DM) or to the control arm (DM), both with two projections and double reading (NCT02698202). Density was automatically assessed with DenSeeMammo software on the 4 DM projections. Recall, detection, and interval cancer rates were reported. All women were followed up to 24 months from recruitment or up to the second round.

Results:

From March 2014-2016, 19566 women were recruited. Density assessment was available for 8651 women in the DM+DBT-arm and 8544 in the DM-arm. Recall rate was similar in both arms and increased with density from 1.9% in BI-RADS A to 4.2% in BI-RADS D. Detection rate, including ductal carcinoma in situ (DCIS), was 9.1/1000(79) and 4.7/1000(40) respectively. DBT+DM detected about two fold more cancers in all BI-RADS classes except A, but interaction could be due to chance (p=0.65). Interval cancers rate was similar in both arms (2.0/1000[17] and 1.9/1000[16] in DBT+DM and DM arm, respectively). Interval cancer rate increases with density in both arms (OR for the increase of one BI-RADS class 1.6, 95%CI 1.0-2.5).

Conclusion:

DBT+DM detected more cancers independently from density, but without impact on interval cancer rate. Density increased recall rate and interval cancer rate in both arms.

Limitations:

Interval cancers are only one of the outcomes to assess the efficacy of DBT in screening.

Ethics committee approval

ProvincialEthical Committee.

Funding:

Partially funded by the Regione Emilia-Romagna and the Local Health Authority-IRCCS.

Purpose:

To assess the concordance of WB-MRI and FDG-PET/CT for staging in children with Hodgkin lymphoma (HL) in order to contribute to the development of evidence-based ‘radiation reduced’ imaging protocols in paediatric HL in the future.

Methods and materials:

A total of 68 children with HL were included in this prospective, multicenter, international study. All participants underwent WB-MRI and FDG-PET/CT at staging. Two radiologist independently evaluated all WB-MR images in two separate readings: with and without DWI. The FDG-PET/CT examinations were evaluated by a nuclear medicine physician. An expert panel assessed all discrepancies between WB-MRI and FDG-PET/CT to derive the FDG-PET/CT-based reference standard. Concordance for correct classification of all disease sites between WB-MRI (both with and without DWI) and the reference standard was calculated, as well as PPV, NPV and Cohen’s kappa statistics. Inter-observer agreement for the WB-MRI including DWI reading was calculated using kappa statistics.

Results:

WB-MRI including DWI was concordant with the FDG-PET/CT-based reference standard for determining disease stage in 97% of the patients (66/68). Agreement between WB-MRI and the reference standard was very good for both nodal (κ=0.97, 95%CI 0.94-0.99) and extra-nodal (κ 1.00) staging. Kappa values for correct disease stage increased significantly with the addition of DWI (p=0.04). Inter-observer agreement between WB-MRI readers was good (κ=0.74, 95%CI 0.58-0.91).

Conclusion:

WB-MRI with DWI showed excellent agreement with the FDG-PET/CT-based reference standard, but did not reach complete concordance for stage. The addition of DWI to the WB-MRI protocol in staging of paediatric HL was shown to improve staging agreement with the FDG-PET/CT-based reference standard.

Limitations:

Main limitation: lack of a true gold standard.

Ethics committee approval

The local institutional review boards approved the study. Written informed consent was obtained.

Funding:

Financially supported by Kinderen Kankervrij (projectnumber 87).

Purpose:

The PRECISION trial was a multi-centre randomised study showing the superiority of MRI-targeted to standard TRUS-guided biopsy in 500 biopsy-naïve men across 23 centres from 11 countries. We report the results from the quality assurance work assessing scans from each trial site.

Methods and materials:

Men were scanned using either a 3T or 1.5T scanner. Three centres used an endorectal coil and 4 did not administer antiperistaltic agents before multiparametric MRI.

64 out of 252 scans were chosen randomly and included scans from each site. Two expert radiologists reviewed the images in consensus using the PI-RADS v.2 guidelines. The quality of each scan was assessed using a 1-to-5 Likert scale, where 1 meant no MR sequences were of diagnostic quality and 5 meant each sequence was independently of diagnostic quality.

Results:

Four scans (6%) had a score of 2, 16 (25%) of 3, 27 (42%) of 4, and 17 (27%) of 5. MR quality was deemed of at least sufficient diagnostic quality (scoring ≥3) for 60 (94%) scans, with 44 (69%) being of good/excellent quality (scoring ≥4).

Adherence to minimum standards was highest for T2-WI, lower for DWI (slice thickness, number of b-values and lack of dedicated long b value sequences) and lowest for DCE (temporal resolution and absence of fat saturation/subtraction).

Conclusion:

MRI and targeted biopsies outperform standard biopsy when sufficient attention is paid to high-quality scan acquisition and reporting. Whilst adherence to PI-RADS v2 minimum technical requirements in PRECISION was good overall, the greatest variability was in the DCE sequences.

Limitations:

Further assessment of the utility of DCE imaging is warranted in the diagnostic setting.

Ethics committee approval

National Research Ethics Committee East Midlands, Leicester (15/EM/0188)

Funding:

NIHR through a doctoral fellowship award (DRF-2014-07-146, to Dr. Kasivisvanathan) and a grant (2015001) from the EAU Research Foundation.

Purpose:

To evaluate the incident vs. prevalent round of supplemental MRI screening for women with extremely dense breasts within a population-based screening program.

Methods and materials:

Between 2011-2015, we randomised 40,373 screening participants (aged 50-75) with a negative mammography and extremely dense breasts (assessed with Volpara software) to supplemental 3.0-T MRI (n=8,061) or mammography screening only (n=32,312). In the prevalent (first) screening round, 4,783 women underwent MRI. Here we compare the results of the incident (second) MRI screening round with those of the prevalent round (presented earlier).

Results:

After the prevalent screening round, 4,308 women (90%) had again undergone mammographic screening with a negative result and were eligible for the second round. Of them, 3,389 (79%) underwent supplemental MRI screening again. Reasons for discontinuation were time constraints, other health concerns, or MRI-related (e.g. claustrophobia).

Supplemental cancer detection rate with MRI was 5.9/1000 screens [95%CI:3.6-9.1], compared to 16.5/1000 [95%CI:13.3-20.5] in the prevalent round. For this, 3.2% [95%CI:2.6%-3.8%] of women were recalled for further diagnostic work-up, compared to 9.5% [95%CI:8.7%-10.4%] in the prevalent round.

Conclusion:

As expected from an incident screening round, the cancer detection rate in this second round of supplemental MRI screening of women with extremely dense breasts is lower than in the prevalent round. Still, by adding MRI, the number of cancers detected almost doubled compared to normal mammography screening only. The false-positive rate decreased to one-third of the rate of the prevalent screening round.

Limitations:

The incident screening round data on interval cancers are not yet available.

Ethics committee approval

Dutch Minister of Health, Welfare and Sport (2011/19); participants gave written informed consent.

Funding:

UMC Utrecht, ZonMw, Dutch Cancer Society, Dutch Pink Ribbon/A Sister’s Hope, Stichting Kankerpreventie Midden-West, Bayer AG Pharmaceuticals, and Volpara Health Technologies.