BRCA1 and BRCA2 carriers are at increased risk of pancreatic cancer, with a 5% lifetime risk and a relative risk of 1.5-9.0¹. Population-based screening for pancreatic cancer has not been recommended² due to the low incidence of pancreatic cancer in the general population and because screening is either invasive (e.g. endoscopic ultrasound (EUS)) and/or expensive (e.g. magnetic resonance cholangiopancreatography (MRCP)). Although pancreatic cancer screening is recommended for high-risk individuals^2,3, no established screening tool is available. BRCA positive patients do however undergo MRI screening for breast cancer, which is recommended annually between the ages of 25 and 75³.

The purpose of this study was to develop and optimize a rapid MR pancreas screening protocol to be performed in conjunction with breast MRI screening in BRCA-positive individuals.

This prospective study was approved by the local IRB. Inclusion criteria: BRCA1 or BRCA2 mutation, age 25 years or older, female, and undergoing screening breast MRI.

The rapid screening protocol was designed to be completed in less than 10 minutes at the completion of the breast MRI examination. The images were acquired with the patient in the prone position, with the breast coil still in place, but using the built-in body coil without placement of a dedicated body surface coil. All studies were performed on a 3T GE MR 750 magnet.

Sequences performed included T2 SSFSE axial and coronal, T1 axial post-contrast, and reduced field of view (rFOV) DWI (b=0, 20, 50, 80, 250, 500 and 800 s/mm²); sequence details are reported in Table 1. Variable NEX were used for rDWI. Image quality was qualitatively assessed by two radiologists in consensus with a grading score system of 1-5, with 1 = all diagnostic information present, and 5 = non-diagnostic. Following subject 1, parameters for axial T1-post-contrast and rFOV DWI were modified to improve image quality (Table 1). The imaging protocol was modified until diagnostic images (score of 1) were achieved for 5 consecutive patients, at which point the protocol was considered optimized. Images were reviewed by the same two radiologists in consensus for detection of pancreatic lesions.

Signal-to-noise-ratio (SNR), contrast-to-noise-ratio (CNR) and ADC were assessed with manually placed regions-of-interest (ROI) in the pancreatic head, body, and tail. As not all sequences included background air, estimates were made using direct measurement of the pancreas:

$$SNR_{pancreas} = \frac {S_{pancreas}}{SD_{pancreas}}$$

$$CNR_{pancreas} = \frac {S_{pancreas}-S_{peri-pancreatic fat}}{SD_{peri-pancreatic fat}}$$

Between July and October 2015, 15 BRCA-positive individuals undergoing routine screening breast MRI underwent a rapid pancreatic screening protocol at the conclusion of their breast MRI examination. The rapid pancreas protocol was successfully completed in all patients. Average time per sequence, SNR, and CNR measurements are reported in Tables 2 and 3. Image quality was rated as 1 (all diagnostic information present) for all patients for the axial and coronal T2 SSFSE and T1 post-contrast sequences. Excellent image quality was achieved for low b-values; however, image quality at higher b-values was more variable. Calculated ADC values for the pancreatic head, body and tail were 1.42, 1.51, and 1.54x10-3 mm2/s respectively. In 3 of 15 subjects, small pancreatic cystic lesions were detected. In 1 subject an hepatic adenoma was identified.

This study demonstrates that rapid pancreas screening using the built-in body coil on a 3T magnet is both feasible and yields diagnostic image quality able to detect abnormalities within the pancreas. Combined with breast MRI screening, this is a rapid, safe, non-invasive, and potentially inexpensive screening tool. With minimal inconvenience to the patient, this screening protocol can be utilized in the BRCA population where currently there is no established screening tool available.

Image quality was still diagnostic even in a subject with a Body mass index of 34.1 kg/m² (Figure 1). The rate of observed positive subjects (3 of 15 subjects) in this small group of BRCA1 and BRCA2 mutation carriers is within the range of expected pancreatic lesions in this population⁴. 2 of 3 subjects positive for cystic lesions underwent dedicated MRCP which did not detect additional pancreatic lesions.

This study was not designed to assess the effectiveness or indications of pancreatic cancer screening itself, but rather to develop a new screening tool which could be further evaluated in larger clinical trials.

Rapid MR protocol for pancreatic cancer screening with subjects in the prone position and use of the built-in-body coil is feasible and provides diagnostic quality images. Ultimately 30 subjects will be enrolled into this pilot study, and in the final version of the protocol we plan to use 3 b-values, probably 0, 20, and 500 s/mm².