Systemic sclerosis (SSc) is a multi-system disease characterized by immune dysregulation, fibrosis and vascular damage that affects the gastrointestinal tract (GIT) in up to 90% of patients. Baseline GIT involvement is an independent predictor of 2-year mortality in patients with early diffuse SSc. An urgent unmet need in the management of SSc GIT involvement is the availability of non-invasive investigations to facilitate early diagnosis, monitoring of disease and treatment response. In this study, we investigate simultaneous FDG-PET/MR imaging for the diagnosis of SSc GIT involvement, hypothesizing that (i) FDG-PET detects inflammation in the bowels associated with early involvement of the GI tract, and (ii) T1 MOLLI mapping is able to distinguish healthy from fibrotic GIT tissue as already demonstrated on cardiac tissue [2].The authors are unaware of any other studies assessing the GI involvement in SSc using FDG-PET or T1 MOLLI maps to investigate bowel fibrosis.

Image Acquisition:

n this institutional review board approved pilot study, 2 SSc patients with symptomatic GIT symptoms, as determined by a validated questionnaire were recruited (female, aged 24 and 62). Subjects fasted for at least 10 hours prior and had a non-spicy low residue diet 3 days prior. Subjects underwent a PET-MR scan (Siemens mMR), 60 minutes after injection of 6mCi of FDG tracer and immediately after injection of 10ml of Buscopan to reduce bowel motion. PET images were acquired on four 10 minute beds to cover the abdomen and thorax. PET data were reconstructed using the OP-OSEM algorithm. The reconstructed images were smoothed with a 6mm FWHM Gaussian filter, and FDG uptake in the gut was quantified by SUV. Simultaneously, the bowel area was imaged using T2 weighted half-Fourier acquisition single-shot turbo spin-echo (HASTE) in coronal view. Then, breath-hold native T1 MOLLI mapping was acquired. Sequence parameters, presented in Table 1, yielded a resolution of 1.48mmx1.48mmx7.2mm. One healthy volunteer underwent the same MR protocol as a control to compare the T1 MOLLI values.

Image Analysis:

From the anatomical images, small and large bowels were manually segmented using MITK software (www.mitk.org). These contours were superposed on the T1 MOLLI map and the PET image as shown Figure 1. SUVs above 1 were considered abnormal while those above 2 were considered “hot”. The PET and MOLLI images were overlaid to measure the correlation between hot spots and high T1 values.

PET image analysis:

Both patients presented PET activity above 1 SUV in a large part of the organs and some hot spots, signs of bowels inflammation with PET SUV above 2. Table 2 shows the percentage of each organ that has a PET SUV above 1 or 2, for both patients.

T1 MOLLI analysis:

At the time of this writing, no ground truth T1 value had been reported for the bowels, so only a comparison with the control data is presented. Table 3 reports the mean T1 value per organ as well as the percentage of the organ that has a high T1 (T1 above 1500 ms which corresponds to the mean water T1 value found in healthy bowels.) Water is normally present in the small bowels. For our subjects it represented 7%, 31% and 27% for patient 1, patient2 and the volunteer respectively, measured on the T2 HASTE.

PET/MOLLI comparison:

Hot spots and high T1 regions overlay in 1% of the bowels for patient 1 and 7% for patient 2, not necessarily within the water areas of the bowels. An example is shown Figure 1 where yellow arrows point to hot spots that correlate with high T1 values, where no water is visible on the anatomy image.

Our preliminary results suggest that GIT inflammation was identified by high SUV on FDG-PET. Both patients with GIT symptoms showed increased uptake over most of their GIT and hot spots in approximately 2-8% of their GIT. Bowel inflammation is not common in healthy subjects [3], so our results suggest that this may be related to SSc. Only some hot spots (1-7%) correspond to high T1 values which could be due to edema caused by the inflammation. Better evaluation of the water content will be added in future studies to differentiate between high T1 values due to water and high T1 values due to fibrosis. Whether hot spots would later lead to fibrotic tissue cannot yet be confirmed. Recruitment of a larger cohort of patients and healthy controls is required to confirm these findings in a longitudinal study with serial scans to tract the inflammation and high T1 value areas.