15th European Molecular Imaging Meeting
supported by:

To search for a specific ID please enter the hash sign followed by the ID number (e.g. #123).
*.ics

Imaging inflammatory response

Session chair: Johannes Schwenck (Tuebingen, Germany); Laura Mezzanotte (Rotterdam, Netherlands)
 
Shortcut: PW08
Date: Wednesday, 26 August, 2020, 5:30 p.m. - 7:00 p.m.
Session type: Poster

Contents

Abstract/Video opens by clicking at the talk title.

400

High resolution SPECT imaging of the inflammatory component of atherosclerotic plaque

Eric J. Meester1, 2, Hilary Barrett1, 2, Frank Gijsen2, Boudewijn J. Krenning4, Erik de Blois1, Marion de Jong1, Kim van der Heiden2, Monique Bernsen1, 3

1 Erasmus MC, Radiology & Nuclear Medicine, Rotterdam, Netherlands
2 Erasmus MC, Biomedical Engineering, Rotterdam, Netherlands
3 Erasmus MC, AMIE Core Facility, Rotterdam, Netherlands
4 Erasmus MC, Cardiology, Rotterdam, Netherlands

Introduction

Inflammation is the key driver in progression of atherosclerotic disease. Nuclear imaging may be used to longitudinally assess extent and composition of the inflammatory response in atherosclerotic plaque. Using a preclinical imaging platform with high-resolution and high-sensitivity capabilities we evaluated a panel of leukocyte-targeting tracers (LFA-1, Somatostatin receptor type 2 (SST2), CXC4 and folate receptor) as a monitoring platform for atherosclerotic plaque progression. The studies were performed with human plaque samples and a mouse model for atherosclerosis.

Methods

Patient-derived endarterectomy tissue sections were incubated with 111In-labeled or 99mTc-labelled tracers for the targets mentioned above. Binding specificity of the tracers was verified via a competition-binding assay with unlabeled tracer. Binding of the tracers within plaque sections was visualized and quantified by ultrahigh-resolution focused SPECT/CT imaging using a high-energy, multi-pinhole collimator or by autoradiography. SPECT/CT data were compared to immunohistological profiles of the tissue samples. In vivo performance of selected tracers was assessed by injection of 200 pmol  labeled (~50 MBq) tracer in atherosclerotic mice and ultrahigh-resolution SPECT/CT 2 h p.i. Target specific uptake was confirmed by ex vivo imaging and (immune)histological staining of excised arteries.

Results/Discussion

There is a need for non-invasive assessment of the inflammatory state of AS plaque to better identify patients at risk of an ischemic event and to monitor treatment efficacy in patients with symptomatic atherosclerosis. We tested a panel of tracers targeting different leukocyte subtypes. For each radiolabeled tracer, specific binding to leukocytes within AS plaque was confirmed. The level of binding and the inter and intra-plaque distribution of each tracer differed substantially. Overall, binding was most pronounced in plaque lesions generally characterized as vulnerable, specifically in rupture-prone areas. Uptake of tracers for LFA-1, SST2 or CXCR4 showed stronger association with features associated with rupture risk compared to the folate receptor.
With 111In-labeled tracers for LFA-1 and SST2 we showed that, despite the very small size of the lesions in mice, the presence of inflamed plaque could be adequately imaged in vivo with target-to-background ratios of around 2.

Conclusions

High resolution SPECT/CT imaging is a powerful tool to test the potential value of leukocyte targeting radiotracers for non-invasive imaging of atherosclerotic plaque inflammation. Our studies indicate that a combination of specific tracers may be of additional value in characterizing the risk level of atherosclerotic plaque in patients.

AcknowledgmentThis project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No.707404. Imaging was performed in and with help from the AMIE facility at Erasmus MC.
References
[1][ Libby, P et al. 2016 Leukocytes Link Local and Systemic Inflammation in Ischemic Cardiovascular Disease,67,  1091.
Figure 1
Keywords: SPECT/CT, Atherosclerotic plaque, Inflammation
402

Monotrack – a modular platform for combined molecular imaging of monocyte function and phenotype in an in vivo model of inflammation

Dominic A. Depke1, Stefanie Zenker2, Michael Claesener3, Sven Hermann1, Thomas Vogl2, Johannes Roth2, Michael Schäfers1, 3

1 WWU Münster, European Institute for Molecular Imaging, Münster, Germany
2 WWU Münster, Institute of Immunology, Münster, Germany
3 University Hospital Münster, Department of Nuclear Medicine, Münster, Germany

Introduction

In a recent publication1 we have shown that combined near infrared dye and 111In or 99mTc-HMPAO labeling of immortalized ER-HoxB8 monocytes allowed to study cell migration dynamics and receptor functions using knockout cell lines in an in vivo model of sterile inflammation. However, the use of unspecific radiolabeling and SPECT imaging limited the spatiotemporal resolution and did not allow for deeper analysis of the immunological phenotype of the migrated cells. To overcome these limitations, we have developed genetic reporter cell lines and a protocol to gently isolate the migrated cells.

Methods

ER-HoxB8 monocytic gene reporter cell lines that constitutively express variants of the human and mouse sodium iodide symporters were generated by lentiviral transduction. Using our subcutaneous granuloma model, a small amout of Biogel containing LPS or a mock control containing saline was injected subcutaneously into the dorsal flanks of a mouse. Intravenous injection of dye labeled reporter cell lines was performed on the next day and cells were allowed to migrate for 18-24 hours. Then 18F Tetrafluoroborate (TFB) was injected and mice underwent PET/MRI or PET/CT imaging. The cells were isolated by injection of PBS into the granulomas and aspiration of the dissolved gel without enzymatic treatment, leaving the cells virtually untouched. Then cells were analyzed by Flow cytometry.

Results/Discussion

In a first series of experiments, we observed accumulation of 18F-Tetrafluoroborate within the LPS containing granuloma following the injection of mouse sodium iodide symporter (mNIS) expressing DiR labeled ER-HoxB8 monocytes (figure 1). Fluorescence reflectance imaging confirmed accumulation of the cells at the LPS containing site, whereas the mock control site showed approximately only half of the signal of the LPS pellet. Unfortunately, analysis of the PET scans showed only a 10-20 % higher TFB uptake when compared to control sites or injection of mNIS negative cells. Further analysis indicated that there was a high background signal within the granulomas caused by unspecific binding of the tracer to the gel matrix. FACS analysis of the isolated cells showed that cells were positive for the monocyte/macrophage markers CD11b and Ly6c. Using an anti-DYKDDDDK antibody isolated cells were confirmed to be mostly ER-HoxB8 monocytes.

Conclusions

We have shown that the use of myleoid genetic reporter cell lines is a powerful tool to study immune cell function in an in vivo model of sterile inflammation. This model allowed us to visualize the migrated cells by whole-body PET imaging and enabled further analysis of the phenotype of the migrated cells by Flow cytometry. We will combine this approach with knockout cell lines and additional PET reporter systems in the near future.

Acknowledgment

We thank Nina Kreienkamp, Roman Priebe, Christine Bätza and Stefanie Bouma for technical assistance.

References
[1] Gran S, Honold L, Fehler O, Zenker S, Eligehausen S, Kuhlmann MT, Geven E, van den Bosch M, van Lent P, Spiekermann C, Hermann S, Vogl T, Schäfers M, Roth J. Imaging, myeloid precursor immortalization, and genome editing for defining mechanisms of leukocyte recruitment in vivo. Theranostics 2018; 8(9):2407-2423. doi:10.7150/thno.23632.

Figure 1.
left: PET-CT scan of biogel granulomas of a mouse that has been injected with mNIS+ ER-HoxB8 monocytes and 18F-TFB. right: FRI scan of the same mouse.
Keywords: PET/MRI, Genetic reporter systems, Immune cell imaging, Models of sterile inflammation, Monocytes
403

[18F]Fluoride PET-CT imaging for detecting disease activity of spondyloarthritis patients

Jerney de Jongh1, Ben J.C. Zwezerijnen1, Robert Hemke3, Maqsood Yaqub1, Irene E. van der Horst-Bruinsma1, Marleen G. H. van de Sande3, Arno W. G. van Kuijk2, Irene E.M. Bultink1, Lot T. Burgemeister2, Nancy M. A. van Dillen2, Alexandre E. Voskuyl1, Conny J. van der Laken1

1 Amsterdam UMC, location VUmc, Department of Rheumatology and Clinical Immunology, Amsterdam, Netherlands
2 Reade, Department of Rheumatology, Amsterdam, Netherlands
3 Amsterdam UMC, location AMC, Department of Rheumatology and Clinical Immunology, Amsterdam, Netherlands

Introduction

Bone formation in Spondyloarthritis (SpA) is presumably related to local enthesitis/peri-articular inflammation and can lead to functional limitation (1,2). X-rays only allow long-term monitoring of bone formation (3). Short term evaluation of treatment effects on bone formation would be valuable. Positron Emission Tomography (PET) using [18F]Fluoride can visualize and quantify bone formation at molecular level (4). In this study we have investigated the feasibility of [18F]Fluoride to assess new bone formation at axial and peripheral enthesial sites in SpA patients.

Methods

Patients with clinically active ankylosing spondylitis (AS) (according to modified New York criteria and BASDAI ≥4) or with psoriatic arthritis (PsA) (according to CASPAR criteria and ≥1 clinically active enthesitis) were included. Of each patient, a whole body [18F]Fluoride PET-CT scan was performed. All scans were scored dichotomously for PET-positive lesions in the spine, peripheral enthesis sites and joints.

Results/Discussion

Preliminary data of this ongoing study are presented. Until now, whole body [18F]Fluoride PET-CT scans were performed in five AS-patients and six PsA patients. With the use of low dose CT for anatomical reference, enthesial sites were identified. In 4/5 AS scans, ≥1 PET positive lesions were found in the cervical, thoracic and/or lumbar vertebrae. These were mainly found in anterior corners of vertebra and bridging syndesmophytes (Fig. 1A). In all six PsA patients, at least 1 PET positive lesion was visualized, projected either intra-articularly or at the site of a tendon attachment. In addition, PET positive lesions were found at the fascia plantaris, achilles- and patella tendon (Fig 1B). [18F]Fluoride uptake at joint level was most frequently present in the ankle or wrist joint.

Conclusions

[18F]Fluoride PET uptake, reflecting new bone formation, can be visualized at enthesitis and arthritis sites in AS- and PsA patients. The technique therefore is sensitive to visualize bone formation and may reflect local disease activity. Additional scans will be collected and analyzed quantitatively, also after treatment, to further investigate the applicability of [18F]Fluoride PET for monitoring of therapeutic effects on bone formation in SpA.

AcknowledgmentWe thank EULAR Foreum, Pfizer and Novartis for financial support of this investigator initiated study.
References
[1] Maksymowych WP, Mallon C, Morrow S, Shojania K, Olszynski WP, Wong RL, et al., 2009, Development and validation of the Spondyloarthritis Research Consortium of Canada (SPARCC) Enthesitis Index, Ann Rheum Dis, 68(6), 948-53
[2] Rezvani A, Bodur H, Ataman S, Kaya T, Bugdayci DS, Demir SE, et al. 2014, 'Correlations among enthesitis, clinical, radiographic and quality of life parameters in patients with ankylosing spondylitis', Mod Rheumatol, 24(4), 651-6
[3] Rudwaleit M, Khan MA, Sieper J. 2005, 'The challenge of diagnosis and classification in early ankylosing spondylitis: do we need new criteria?', Arthritis Rheum, 52, 1000-8
[4] Bruijnen ST, Verweij NJF, van Duivenvoorde L, Bravenboer N, Baeten D, van Denderen JC, et al. 2017, '[18F]Fluoride PET-CT imaging of bone formation in ankylosing spondylitis before and after 12 weeks of anti-TNF treatment', Rheumatology, 57(4), 631-638
Figure 1

[18F]Fluoride uptake in the cervical, thoracic and lumbar spine in a clinically active AS patient (A) and in the patella tendon of the right knee in a clinically active PsA patient (B)

Keywords: positron emission tomography, spondyloarthritis, nuclear imaging, bone formation
404

Characterization of the hepatic lipid composition in a murine model of acute and chronic carbon tetrachloride induced hepatitis by in vivo magnetic resonance spectroscopy

Johannes Schwenck1, 2, 3, Bianca Blöchl1, Andreas Schmid1, Manuela Martella4, Natalie Mucha1, Leticia Quintanilla-Fend4, Bernd J. Pichler1, 2, Manfred Kneilling1, 2, 5

1 Eberhard Karls University Tübingen, Werner Siemens Imaging Center, Tübingen, Germany
2 Eberhard Karls University Tübingen, Cluster of Excellence iFIT (EXC 2180) "Image-Guided and Functionally Instructed Tumor Therapies", Tübingen, Germany
3 Eberhard Karls University Tübingen, Department of Nuclear Medicine and Clinical Molecular Imaging, Tübingen, Germany
4 Eberhard Karls University Tübingen, Institute of Pathology and Neuropathology, Tübingen, Germany
5 Eberhard Karls University Tübingen, Department of Dermatology, Tübingen, Germany

Introduction

High caloric diet causes liver steatosis and induces non-alcoholic steatohepatitis (NASH) in a significant number of patients, which is causing a growing number of hepatocellular carcinomas (HCC). Currently, invasive biopsies and subsequent histopathological analysis are needed for a reliable diagnostic of NASH. Here, the feasibility of non-invasive in vivo 1H magnetic resonance spectroscopy (MRS) was investigated in a murine model of acute and chronic carbon tetrachloride (CCl4)-induced hepatitis, which shows similar features, like steatohepatitis, in histology.

Methods

Acute hepatitis (AH) was induced in mice by a single intraperitoneal (i.p.) CCl4 injection while repetitive CCl4 injections (3x weekly over five weeks) were performed to cause chronic hepatitis (CH). We assessed hepatic lipid composition in AH and CH using MRS and performed extensive ex vivo analysis after the experiment employing histopathology (H&E) as well as RT-PCR analysis focusing on the expression patterns of proinflammatory cytokines (TNF, IL-1b, IL-6), hypoxia inducible factor 1α (HIF-1α) and reactive oxygen species-associated superoxide dismutase 1 (SOD-1). Furthermore, the liver enzymes in the blood serum of mice with AH and CH were measured. Finally, flow cytometry analysis was performed to investigate immune cells in livers and spleens of mice with CH.

Results/Discussion

We revealed a tendency to increased fatty acids as well as elevated fractional lipid mass by MRS in the livers of AH mice compared to sham-treated control mice. In contrast, no significant alterations in the hepatic lipid composition upon CH were detected. H&E histology in AH liver tissue uncovered necrosis, regenerative processes and microsteatosis, whereas CH liver tissue exhibited weak signs of inflammation but remarkable necrosis and steatosis. We determined higher serum alanine transaminase only in mice with AH but not in CH. CH liver tissue revealed a strong expression of TNF in RT-PCR when compared to AH, while downregulation of SOD-1 was found in both, AH and CH, compared to liver tissue from sham-treated controls. The infiltrating CD4+ T cells in the liver tissue of CH mice were clearly decreased while an increase in regulatory T cells and PD-1+ CD4+ T cell was detected. In contrary, effector memory CD4+ T cells and PD-1+ CD4+ T cells were increased in the spleen of CH mice.

Conclusions

In vivo MRS is able to assess molecular changes in CCl4-induced AH and CH, which are characterized by distinct inflammatory and molecular mechanisms. However, a detailed characterization of the basic underlying inflammatory and molecular mechanisms is required to evaluate the diagnostic performance of MRS in NASH. Furthermore, clinical translation of the results could support detection and treatment stratification of NASH patients.

Keywords: MRS, hepatitis, NASH
405

Targeted 89Zr-labelled microbubble-based contrast agents for imaging inflammation in a murine model of inflammatory bowel disease

Una Goncin1, Viswas R. Solomon2, Eric Price2, Steven Machtaler1

1 University of Saskatchewan, Department of Medical Imaging, College of Medicine, Saskatoon, Canada
2 University of Saskatchewan, Department of Chemistry, Saskatoon, Canada

Introduction

Inflammatory bowel disease is characterized by inflammation and ulceration of bowel segments, resulting in fibrosis and diseased bowel resection. Endoscopic monitoring is the gold standard for diagnosing and monitoring activity, but is invasive, can’t reach all areas of bowel, and has difficulty identifying submucosal inflammation. There is an unmet need to rapidly detect and diagnose active inflammation in the entire bowel. We aim to develop radiolabelled microdroplet-based contrast agent, RadioDroplet (RD), for molecular imaging of inflammation using PET in a murine model of IBD.

Methods

RDs were prepared by extruding dodecafluoropentane with solubilized phospholipids containing two functionalized lipids (PE-PEG2000-Biotin, PE-PEG2000-Azide that was prelabeled with DFO-DBCO at 4°C overnight; Figure 1A). RDs were sequentially labelled with neutralized zirconium-89 (30 min RT), streptavidin, and biotinylated CD62P antibody, washing between each step.
Biodistribution: 89Zr-RDs (100 uL, 0.6 MBq) were injected intravenously into 9 healthy FVB mice. Tissues were harvested at 5 min, 30 min, and 60 min (n=3 per time point), and analyzed using a gamma counter.
In vivo PET imaging: Acute IBD was induced using 4% dextran sulfate sodium (DSS) over 6 days in 1 of 4 female FVB mice. 89Zr-RDs were injected intravenously (150 mL, 1-2 MBq) and imaged for 60 min using micro-PET (Sophie G4).

Results/Discussion

RD production and labelling: Validation of successful RD-antibody labelling was carried out using confocal microscopy, where a positive fluorescent signal was detected (Figure 1B). There was high labelling efficiency of zirconium-89 (95-97%) on the RDs based on iTLC (Figure 1C).
Biodistribution: 89Zr-RDs rapidly cleared the blood pool and accumulated in the lungs (mean, 31.8 ± 12.5 % injected dose [ID]/g at 60 min), liver (mean, 16.2 ±  8.1 % ID/g at 60 min), and spleen (mean, 26.3 ± 16.0 % ID/g at 60 min).
In vivo PET imaging: 89Zr-RDs accumulated within the lungs, liver and spleen, consistent with ex vivo biodistribution. Within 60 min, there was a distinct accumulation within a bowel segment in the mouse with induced acute IBD. This bowel signal was not present in any control mice (no inflammation). At the conclusion of the 60 min, most mice had an accumulation of activity within the bladder, consistent with clearance via the reticuloendothelial system (RES).

Conclusions

These preliminary data show RDs have potential to be a useful PET contrast agent for detecting inflammation in a mouse model of IBD. We are currently modifying our design to elongate blood pool half-life to increase target binding.  If successful, this would provide a diagnostic imaging tool that would enable clinicians to visualize inflammation in the full digestive tract in a short period of time, guiding decisions on monitoring and therapy.

Fig. 1. 89Zr-labelled RadioDroplets (RD) for PET imaging of acute murine IBD.

A) Schematic of dodecafluoropentane-filled RDs (Shell: DSPC, PEG40S, PE-PEG2K-Biotin, and PE-PEG2K-Azide) labelled with aP-selectin and 89Zr. B) Fluorescent confirmation of RDs labelled with aP-selectin (red). C) iTLC chromatograms of purified (left) and bound 89Zr (right) on RD in 50 mM EDTA. D) Biodistribution of RDs at 5 min (blue), 30 min (red), 60 min (green; n=3 FVB mice per time point). E) PET images of control (n=3) and acute IBD mice (n=1) at 30 min and 60 min after i.v. injection of RDs. White arrow indicates accumulation in bowel only present in acute IBD mouse.

Keywords: contrast agent, P-selectin, inflammation, ex vivo biodistribution, in vivo microPET
406

[68Ga]Ga-NOTA-folate detects atherosclerotic plaque inflammation: comparison with [18F]FDG

Senthil Palani1, Jenni Virta1, Heidi Liljenbäck1, Xiang Guo Li1, Olli Moisio1, Tiit Örd2, Merja Heinäniemi3, Minna Kaikkonen-Määttä2, Juhani Knuuti1, Philip Low4, Antti Saraste1, Anne Roivainen1

1 University of Turku/ Åbo Akademi University/Turku University Hospital, Turku PET Centre, Turku, Finland
2 University of Eastern Finland, A.I.Virtanen Institute for Molecular Sciences, Turku, Finland
3 University of Eastern Finland, Biomedicine, Kuopio, Finland
4 Purdue University, Department of Chemistry, West Lafayette, United States of America

Introduction

Atherosclerosis is a chronic inflammatory disease characterized by the accumulation of macrophage subtypes defined by distinct markers and functions. Folate receptor b(FR-β ) is one of many markers selectively expressed on macrophage subsets. We have previously shown that [18F]AlF-NOTA-folate targeting FR-β positive macrophages enables in vivo detection of atherosclerotic plaque inflammation by PET/CT (1). Here, we evaluated another FR-β targeted tracer, [68Ga]Ga-NOTA-folate, to investigate atherosclerotic plaque inflammation and compared the results to [18F]FDG.

Methods

First, we determined the affinity of [68Ga]Ga-NOTA-folate using CHO cells expressing FR-β . Then, we studied low-density lipoprotein receptor deficient mice expressing only apolipoprotein B100 fed with high-fat diet to induce atherosclerotic lesions. The mice were injected with [68Ga]Ga-NOTA-folate or [18F]FDG for PET/CT. The tracers were compared in a head-to-head PET/CT setting and results were expressed as standardized uptake values (SUV) and target-to-background ratios (TBR). After sacrifice, the tissues excised were measured with a γ-counter for ex vivo biodistribution. Further, the tracer distribution and co-localization to macrophages on aorta cryosections were studied using autoradiography, hematoxylin-eosin staining and immunostaining with anti-Mac-3 or anti-FR-β antibody.

Results/Discussion

The cell binding studies showed high (5.1±1.1 nM) affinity of [68Ga]Ga-NOTA-folate to FR-β . The myocardial uptake of [68Ga]Ga-NOTA-folate (SUV 0.43±0.06) was 20-folds lower compared to [18F]FDG (SUV 10.6±1.8, P=0.001). The autoradiography and immunohistochemistry of aorta revealed that [68Ga]Ga-NOTA-folate radioactivity co-localized with Mac-3- and FR-β-positive macrophage-rich plaques. The plaque-to-healthy vessel wall ratio of [68Ga]Ga-NOTA-folate (2.0±0.4) showed no difference than that of [18F]FDG (1.9±0.3, P=0.23).

Conclusions

The [68Ga]Ga-NOTA-folate results are in the line with our previous studies using [18F]AlF-NOTA-folate and corroborate FR-β as an imaging target for detection of inflamed atherosclerotic lesions.

References
[1] Silvola JMU, Li X-G, Virta J, Marjamäki P, Liljenbäck H, Hytönen JP, Tarkia M, Saunavaara V, Hurme S,  Palani S, Hakovirta H, Ylä-Herttuala S, Saukko P, Chen Q, Low PS, Knuuti J, Saraste A, Roivainen A. Aluminum fluoride-18 labeled folate enables in vivo detection of atherosclerotic plaque inflammation by positron emission tomography. Sci Rep 8:9720, 2018
[68Ga]Ga-NOTA-folate and [18F]FDG in vivo PET/CT

Figure 1. A) Coronal PET/CT image of a atherosclerotic mouse administered with [68Ga]Ga-NOTA-folate or [18F]FDG. White arrows shows aortic arch and pink arrows shows myocardium region, respectively. B) Quantification of myocardial PET data shows significant difference between the tracers.

[68Ga]Ga-NOTA-folate and [18F]FDG ex vivo autoradiography

Figure 2. A) Hematoxylin-eosin (H&E) staining and autoradiography image from representative aorta cryosections. Black arrow shows healthy vessel wall and red arrow shows plaque region, respectively. B) Quantification of autoradiography data shows no difference between the tracers.

Keywords: Atherosclerosis, Immune inflammation, PET/CT
407

FDG-PET of acute gastrointestinal Graft versus Host disease (GvHD): diagnostics and prediction of response to anti-inflammatory therapy

Wolfgang Roll1, Georg Evers2, Rebecca Strotmann1, Benjamin Noto1, Matthias Stelljes2, Michael Schäfers1, 3

1 University Hospital Münster, Nuclear Medicine, Münster, Germany
2 University Hospital Münster, Hematology, Oncology and Pulmonary Medicine, Münster, Germany
3 WWU Münster, European Institut for Molecular Imaging, Münster, Germany

Introduction

Allogenic stem cell transplantation (ASCT) is the only curative treatment option in numerous hematologic malignancies. Graft-versus-Host disease is a common complication significantly decreasing morbidity and mortality of ASCT. First studies indicate that FDG-PET can be used for detection and outcome prediction in GvHD patients. Objective of this retrospective study was to determine the prognostic value of FDG-PET in acute gastrointestinal GvHD patients.

Methods

101 patients with suspected acute gastrointestinal GvHD underwent FDG PET between 6/2011 and 2/2019. PET data sets were reviewed for presence of intestinal inflammation. 72 patients with proven acute GvHD and positive PET were included in the analysis of response to anti-inflammatory therapy and of outcome. Quantitative PET parameter SUVmax and SUVpeak were compared between patients with fast response to anti-inflammatory therapy and slow/no response.

Results/Discussion

Overall FDG PET demonstrated a per patient sensitivity of 93 % and, in comparison lower specificity (72 %) for the detection of acute GvHD. Patients with fast response to anti-inflammatory therapy showed significantly higher SUVmax (mean 13.4) and SUVpeak values compared to slow/non-responders (mean SUVmax 7.5) (p< 0.05).

Conclusions

FDG PET might contribute to detection of acute gastrointestinal GvHD and more important prediction of response to anti-inflammatory therapy. Prospective studies are needed before implementation into clinical routine.

Keywords: GvHD, FDG, gastrointestinal