Introduction
Functional Imaging Of Pancreatic Islet Hemodynamics
Beta Cell Surface Markers
Beta Cell Metabolism
Specific Applications for Assessing Type 1 Diabetes
Specific Applications for Assessing Amyloid Deposits in Type 2 Diabetes
Specific Applications for Islet Transplantation
Generic Recommendations
Workshop Participants

Diabetes Mellitus can be described as the result of an inadequate mass of functional pancreatic islet cells. The endocrine pancreas, composed of the b cells as well as the other three endocrine cell types of the islets of Langerhans, is normally only 1-2% of the adult pancreas. The low insulin secretion seen in diabetes can be due to selective autoimmune destruction of b cells as in type 1 diabetes, to functional defects as in MODY, or to inadequate compensation for increased insulin demand due to insulin resistance as in type 2 diabetes. Because there is no current way to visualize pancreatic b cells, many questions of the natural history of the disease are still to be answered. For instance, in type 1 diabetes, interventional strategies may depend on a clear understanding of whether there is a slow gradual loss of b cells, or maintenance of b cell mass until a certain point and then a sudden loss. Similarly, new treatments for type 2 diabetes may arise from the knowledge of whether the observed functional failure is due to altered rates of b cell formation or loss.

The functional capacity of islets can be assessed in vivo in humans or animals by measuring insulin secretion during a hyperglycemic clamp or an intravenous glucose tolerance test. However, the question stands as to whether insulin secretion correlates well with the actual number of b cells or b cell mass in diabetic patients. At present, determinations of cell mass can only be done on excised pancreas. Data collected in the perfused pancreas from animal models of diabetes show that a loss of glucose-induced insulin secretion results from a reduction in the b cell mass. Both loss of first phase glucose-induced insulin release and loss of the normal oscillations of insulin secretion are seen in type 2 diabetic patients. In animal models, it has been shown that such loss is secondary to sustained hyperglycemia, and often can be reversed by normalization of glycemic levels. Non-invasive means of assessing b cell mass could allow longitudinal studies to map the evolution of disease both in animal models and, more importantly, in humans.

Although significant progress has been made in imaging anatomic, physiologic, biological and genetic markers in other organs and disease states, there is a paucity of imaging research currently addressing the endocrine pancreas. Over the last decades a variety of different imaging techniques have been introduced and developed with unprecedented spatial, contrast and temporal resolution. Several of these techniques are relatively non-invasive, and can therefore be used to monitor people and animals, especially valuable transgenic and knock out mice, longitudinally over time. Although there is no single technique that fulfills all three parameters, existing techniques are complementary and may be used to address different biological questions. Techniques most suited for endocrine pancreatic imaging include magnetic resonance (MR) imaging, MR spectroscopy (NMR), positron emission tomography (PET), single photon emission computed tomography (SPECT), fluorescence imaging, intravital microscopy, bioluminescence imaging and potentially computed tomography (CT) and ultrasound (US). Currently achievable spatial resolutions for anatomic imaging of deep tissues range from 10 microns to 500 microns in isolated cells, animals and humans using state-of-the-art MR, and contrast resolution can be extraordinarily high for nuclear imaging techniques.

Thus, there is an exciting opportunity to apply the new imaging techniques to the pancreatic b cell to further increase our knowledge of the evolution of a devastating disease, diabetes. Imaging is also expected to have a major impact in the clinical management of diabetes. There are few if any investigators in this area as yet, and this meeting provided an interface between experts in the two separate fields of pancreatic b cell biology and imaging. There were many topics discussed and many obvious feasibility experiments were identified that need to be done. The lack of previous work in this important field necessitates measures to jump start the area by giving ear-marked funding in an appropriate grant mechanism that will allow exploration of different aspects. The following areas of physiology, metabolism and cell biology were discussed and identified as potential targets for imaging to further our insight into b cell mass and function in vivo.

The islets of Langerhans are highly vascularized with fenestrated capillaries. Islet blood flow is highly regulated in response to changes in function (such as response to elevated glucose) and has been measured in experimental animals to be as high as 6 ml/min/gm tissue. Transplantation of islets encourages growth of these specialized capillaries in the region of engraftment. Functional imaging techniques that assess blood volume, blood flow, vascular permeability and oxygenation hold promise for visualizing the islet and defining its physiology. Areas that should be encouraged for investigation include:

• High resolution functional MR and PET imaging of the pancreas to assess regional blood flow, regional blood volume, vascular permeability and oxygenation.
• Assessing whether contrast that arises from these functional imaging techniques can be used to assess islet volume and mass under normal and pathophysiological states. The question would be whether a "visual cast" of the islet microvasculature would yield estimates of the islet b cell mass.
• Defining normal and pathophysiological changes in regional blood flow, blood volume, vascular permeability, and oxygenation of islets in animal models and humans to assess the specificity of these measures for visualizing islets and defining their role in normal and pathophysiological responses.
• Defining fruitful protocols to cause changes in islet hemodynamics that assess function, such as high glucose challenges, pharmacologic challenges, etc.
• Develop quantitative blood flow techniques (ml/min/gm) because longitudinal studies are needed to understand the evolution of disease. Additionally, these data will be needed to complement the study of receptor binding by separating the flow dependence from the receptor concentration dependence.
• Validation of these flow methods relative to microsphere studies or other "gold standard" methods.

b cells are one of four endocrine cell types within an islet and comprise 60-80% of the islet. The development of imaging techniques that could distinguish b cells from both the other endocrine cells and the surrounding exocrine acini and ductal tissue should be encouraged. b cells have both a specialized metabolism and many neuroendocrine properties that may allow them to be specifically visualized. Development of novel imaging techniques may also ultimately improve detection of islet cell tumors. Several suggested areas of exploration are:

• Evaluate known receptors that are highly or specifically expressed in b cells. Some potentially useful receptors are GLP-1, somatostatin, and the high-affinity sulphonylurea receptor SUR-1. It is critical to determine the receptor number per cell (or per mg protein) for candidate ligands that could be used to label the b cells. This information is needed as a first screen because the target to non target ratio (T/NT) is equal to the receptor concentration divided by the K_d for its ligand at a very high specific activity. This will be a maximum ratio and is necessary, but not sufficient information, because of the complicating factors of protein binding, metabolism, non-optimal transport, etc. This ratio assumes that cell surface receptors are not internalized during the imaging time.
• Evaluate beta cell surface molecules that may be specifically associated with diabetes or a high risk for diabetes. For instance, Class I MHC appears to be hyperexpressed in all cell types of islets containing residual insulin in the pancreas of type 1 diabetics.
• Set up an islet cell surface marker discovery program. At present there are a number of known monoclonal antibodies that specifically (within the pancreas) bind to b cells, e.g. A2B5, and CD98. These should be examined as potential molecules for tagging with imageable markers.
• Develop and evaluate non-radioactive labeling of ligands for receptors or other b cell surface markers.
• Develop in vivo models and test those promising compounds that exhibit reasonable T/NT ratios. Although the present spatial resolution for PET is 5 mm, a high T/NT ratio can lead to a high "biochemical" resolution. Stress tests using glucose, sulfonylureas, or peptides that are known to stimulate or inhibit insulin secretion may differentiate between b cells and acinar cells.
• Assess the delivery of contrast agents intracellularly. It may be necessary to develop carrier systems. For instance, could phogrin, a protein that spans the membrane of the insulin granule and is recycled with the membrane of the exocytosed granule, act as a carrier?

The b cell has many unique metabolic characteristics that function to allow it to monitor the blood glucose concentration and regulate insulin secretion. Many of these characteristics offer potential tools for imaging the b cells and possibly to monitor their function. These can be tested on isolated islets in vitro or in islets transplanted under the kidney capsule; those protocols that show promise should then be taken to in vivo animal models and human populations.

• Application of current in vivo NMR spectroscopic techniques (¹H-³¹P-¹³C-¹⁹F-²³Na etc.) to the pancreas. Are there characteristic spectral profiles that can give estimates of b cell mass even without their actual visualization? Isolated islets could be used to identify potential molecules for in vivo spectroscopy. Exocrine tissue should also be evaluated in order to identify differences between these two tissues. For example, the phosphate contained within the insulin granules may be detectable; n-acetyl aspartate or myoinositol can be detected spectrally in other neuroendocrine tissues, may be present in b cells, and may uniquely identify them in the pancreas and/or islet.
• Identification of new chemical markers of b cells which are detectable using NMR spectroscopy. The unique profile of metabolic enzymes (high glucokinase, low hexokinase, low lactate dehydrogenase) that characterize the high aerobic metabolism of the b cell may provide tools.
• Explore the use of new methods for monitoring intracellular calcium ion concentration and flux in response to inhibitors and stimulators of insulin secretion.
• Identify, characterize and develop transgenic models with b cells enhanced for imaging using transgenes such as insulin promoter with isotope binding domains, modified green fluorescent proteins, luciferase, or b-galactosidase. These mice might be useful in visualizing the b cells within the pancreas for mass assessment as well as to determine b cell lineages through development.

Type 1 diabetes has a clear etiology of immune destruction of the pancreatic b cells but it is not known whether the actual loss of b cells is gradual or sudden. There may be differences in the dynamics of b cell loss in children, who appear to have a very quick onset of diabetes, versus adults who may experience a slower course. Even when a child is diagnosed with type 1 diabetes in ketoacidotic crisis, he may experience a "honeymoon" phase of unknown origin, in which he has a lowered requirement for therapeutic insulin. This may be evidence that these children experience a gradual loss of beta cell reserve. It is thought that insulitis, the accumulation of lymphocytes in the islet, can occur for many years before the clinical onset of type 1 diabetes. Certainly, auto-antibodies are found in the serum of high-risk patients long before they become diabetic. Insulitis has been studied most in the NOD mouse, where it is much more pronounced than in humans. In these animals, insulitis can be destructive or non-destructive, accompanied by different patterns of cytokine production. Imaging of b cell mass and insulitis over time may therefore provide important information about the progression of type 1 diabetes. Imaging would also aid the search for and evaluation of therapies to prevent or delay the onset of type 1 diabetes in high-risk patients who have circulating auto-antibodies. These measurements may eventually constitute new surrogate end points for clinical trials designed to assess intervention strategies.

A major goal of modern imaging techniques is to track the distribution and movement of specific types of cells non-invasively in animals and humans. For example, when organ rejection occurs, immune cells, such as T-cells and macrophages, accumulate at the rejecting graft. Superparamagnetic iron-oxide (SPIO) particles can be incorporated into rat T-cells and macrophages. By high resolution MRI, single SPIO-labeled T-cells in 3% gelatin can be visualized. These SPIO-labeled T-cells, or SPIO particles, can be introduced into live rats and tracked as they migrate to the site of tissue rejection. SPIO labeled T-cells or macrophages can also be detected in vivo by MRI in allogeneic transplanted kidneys and lungs. Similarly, monitoring immune cell trafficking in animal models may be possible using immune cells tagged by photoprotein expression where the light emission can be monitored externally to follow these cells in vivo. Such techniques could provide markers for the progression of islet inflammation.

Specific recommendations for exploring and developing techniques applicable to the endocrine pancreas:

• Use magnetically labeled T cells in adoptive transfer in NOD mice to follow their path of homing to the islets by high resolution MR imaging;.
• Develop new methods to image trafficking of immune cells.
• Develop transgenic mice with labeled immune cells, e.g. photoprotein expression that can be passively transferred into NOD mice and their temporal and spatial distribution monitored.
• Use specific immune cells and mixtures of specific cells in these trafficking experiments to determine the key interactions between immune cells that mediate disease; test molecules that may specifically inhibit these interactions using imaging to assess trafficking and localization to pancreas tissue.
• Correlate b cell mass with the presence of inflammation; this will require a combination of techniques.
• Assess the inflammation (T cells) in humans during autoimmunity and combine with new techniques for monitoring b cell mass to provide important data on the progression of the disease.
• Target macrophages, using nuclear, MR or bioluminescence imaging techniques.

One of the characteristics of type 2 diabetes is the deposition of amyloid in the pericapillary spaces of many, but not all, islets. This occurs in humans, monkeys and cats, but not in rodents except for humanized transgenic mice. Islet amyloid is composed of fibrillar IAPP (islet amyloid polypeptide) that is secreted by the b cells along with insulin. It can come to occupy 50-70% of an islet in advanced type 2 diabetes. IAPP fibrils (particularly intermediate sized ones) are toxic to b cells in vitro.

• Evaluate if the b pleated sheets of amyloid can be visualized. The use of I¹²¹ small amyloid protein (SAP) has been used to image amyloid in vivo. Likewise it may be useful in labeling islet amyloid.
• Islet amyloid deposition precedes the overt development of diabetes in monkeys and humans as well as in transgenic rodent models. Detection of the amyloid prior to onset of diabetes in animal models is therefore an important validation of any imaging technique, and would determine whether islet amyloid could be a useful marker to identify humans at high risk of developing type 2 diabetes.
• Imaging could be used to understand the temporal pattern of b cell loss in type 2 diabetes; currently it is not understood whether b cell mass expands prior to this loss or not.

A promising therapy for type 1 diabetes is the transplantation of isolated pancreatic islets. Primary islet non-function after islet transplantation is a common problem, but it is unclear if it is due to local inflammation around engrafted islets, to the engraftment of an inadequate islet mass, or some other reason. Often a recipient only achieves insulin independence six months after transplantation, so the question of whether this arises only after an expansion of b cells should be addressed.

• Evaluate the mass of islets and/or b cells that are engrafted after one week, one month, six months.
• Assess the presence of local inflammation around the engrafted islets.
• Localize the engrafted islets before biopsy; the engraftment of islets scattered throughout the liver has created a veritable "needle in the haystack" situation when biopsies are needed.
• Determine the origin of the blood perfusion of engrafted islets placed into the liver (hepatic artery vs. portal vein); would different transplant procedure allow better oxygenation of transplanted islets?
• Develop new non-invasive methods for monitoring transplant function. For instance, quantitative measurement of renal cortical perfusion by MRI with arterial spin labeling of water can provide a non-invasive diagnostic method for monitoring the status of kidney transplants.

• Issue RFAs to support exploratory research. The R21 mechanism could be used to fund the collection of feasibility data for new approaches and the development of new tools.
• Encourage interdisciplinary research (b cell and imaging community; basic science and clinical community working together).
• Allocate money to jump-start new research groups (i.e. $3,000,000/year which would support about 20 R21s).
• Encourage new investigators and encourage imaging centers to invest in new tenure track faculty.
• Have key members of working group make sure that these objectives/ recommendations are achieved.
• Sponsor "Gordon Conference" like meetings that take place over several days to brain storm and foster collaborations.
• Sponsor (cross) training of investigators.
• Establish a network for equipment availability to pursue above studies.
• Speed track this RFA.

Co-Chairs: Susan Bonner-Weir, Ph.D.
Ralph Weissleder, M.D., Ph.D.
NIH: Maren R. Laughlin, Ph.D.
JDF: Elaine Young, Ph.D.

MAREN R. LAUGHLIN, Ph.D.
Director of Metabolism Program
Division of Diabetes, Endocrinology and Metabolic Diseases
NIDDK, NIH
Building 45, Room 5AN-24J
45 Center Dr, MSC 6600
Bethesda, MD 20892-6600
P: (301) 594-8802
F: (301) 480-3503
Maren.laughlin@nih.gov

ELAINE YOUNG, Ph.D.
Associate National Scientific Program Manager
Juvenile Diabetes Foundation International
Research Department
1400 I Street, NW, Suite 530
Washington, DC 20005
P: (202) 371-9746, ext. 15
F: (202) 371-2760
eyoung@jdfcure.org

LYDIA AGUILAR-BRYAN, Ph.D.
Baylor College Of Medicine
Dept/Medicine/Endocrinology
One Baylor Plaza 537E
Houston, TX 77030
P: (713) 798-3462
F: (713) 790-4585
lbryan@bcm.tmc.edu

SUSAN BONNER-WEIR, Ph.D.
Workshop Co-Chairman
Joslin Diabetes Center
Islet Transplant/Cell Bio Section
One Joslin Place
Boston, MA 02215
P: (617) 732-2581
F: (617) 732-2650
Susan.Bonner-Weir@joslin.harvard.edu

PETER C. BUTLER, M.D.
Professor and Chief
Division of Endocrine
USC School of Medicine
2025 Zonal Avenue, BMT-B11
Los Angeles, CA 90089-9326
P: (323)226-2804
F: (323)226-2809
pbutler@hsc.usc.edu

BARBARA E. CORKEY, Ph.D.
Boston Medical Center
Dept Of Medicine
88 East Newton Street
Boston, MA 02118
P: (617) 638-7091
F: (617) 638-7124
bcorkey@med-med1.bu.edu

GEORGE S EISENBARTH, M.D.
Univ. of Colorado Hlth Scis Ctr
Dept. of Pediatrics Med Immune
B Davis Center Childhood Diabetes
4200 E 9th Ave Box B140
Denver, CO 80262
P: (303) 315-4891
F: (303) 315-4892
george.eisenbarth@uchsc.edu

ALAN FOULIS, M.D.
Pathology Dept.
Royal Infirmary
Glasgow G4 OSF
United Kingdom
P: 44-141-211-4224
F: 44-141-211-4884
gcl083@clinmed.gla.ac.uk

STEVEN D FREEDMAN, M.D., Ph.D.
Beth Israel Deaconess Med. Center
Dana 501
330 Brookline Ave
Boston, MA 02215-02215
P: (617) 667-5576
F: (617) 667-2767
Pager (617) 667-5555 #1258
sfreedma@caregroup.harvard.edu

DAVID M. HARLAN, M.D., MC, USN
Dir. Immune Cell Biology Program
Naval Medical Res. Inst. (Code 61)
8901 Wisconsin Ave, Bldg.18, Rm:232B
Bethesda, MD 20889-5607
P: (301) 295-3548
F: (301) 295-0376
Harland@nmripo.nmri.nnmc.navy.mil

JOHN C. HUTTON, Ph.D.
Barbara Davis Center Child Diabetes
Univ. of Colorado Health Science Center
Dept. of Pediatrics
4200 E 9th Ave, Box B140
Denver, CO 80262
P: (303) 315-4128
F: (303) 315-4892
john.hutton@uchsc.edu

LEIF JANSSON, M.D.
Dept of Medical Cell Biology
Uppsala University
Box 571, S-751-23
Uppsala, Sweden
Leif.Jansson@medcellbiol.uu.se

JOHN L. LEAHY, M.D.
University of Vermont College of Medicine
Given Building C-332
Burlington, VT 05405-0068
P: (802) 656-0835
F: (802) 656-8031
Jleahy@zoo.uvm.edu

ZHIDONG LING, PhD
Diabetes Research Center
Vrije Universiteit Brussel
Laarbeeklaan 103
B-1090 Brussels, Belgium
P.: 32 2 4774544
F: 32 2 4774545
zling@mebo.vub.ac.be

MICHAEL J. MACDONALD, M.D.
University of Wisconsin-Madison
Dept of Pediatrics
1300 University Ave Rm 3459
Medical Sciences Center
Madison, WI 53706 53706
P: (608) 262-1195
FAX: (608) 262-9300
mjmacdon@facstaff.wisc.edu

JERRY P. PALMER, M.D.
University Of Washington
Dept. of Medicine / MS-111
1660 South Columbian Way
Seattle, WA 98108
P: (206) 764-2495 or 1438
F: (206) 764-2693
jpp@u.washington.edu

KENNETH S. POLONSKY, M.D.
University Of Chicago
Dept. of Medicine, Section of Endocrinology
5841 S. Maryland Ave, MC 1027
Chicago, IL 60637
P: (773) 702-6217
F: (773) 834-0486
polonsky@medicine.bsd.uchicago.edu

CAROLYN J. ANDERSON, Ph.D.
Asst. Prof. of Radiology, Molecular Biology and Pharmacology
Washington Univ. School of Medicine
510 S. Kingshighway, Box 8225
St. Louis, MO 63110
P: (314) 362-1053
F: (314) 362-6971
andersoncj@mirlink.wustl.edu

K. TY BAE, M.D., Ph.D.
Assistant Professor of Radiology
Washington University School of Medicine
510 S. Kingshighway Blvd. Box 8131
St. Louis, MO 63110
P: (314) 362-1053
F: (314) 362-6971
Bae@mirlink.wustl.edu

ALEXEI A. BOGDANOV, Ph.D.
Center for Molecular Imaging Research
Massachusetts General Hospital
149 13^th Street
Charlestown, MA 02129
P: (617) 726-8226
F: (617) 726-5708
abogdanov@helix.mgh.harvard.edu

CHRISTOPHER H. CONTAG, Ph.D.
Stanford University School Of Medicine
Dept Of Pediatrics
300 Pasteur Drive, S257
Stanford, CA 94305-5208
P: (650) 723-0707
F: (650) 725-7723
ccontag@cmgm.stanford.edu

WILLIAM C. ECKELMAN
Chief, Positron Emission Tomography Department
Clinical Center, NIH
Building 10, Room 1C40
Bethesda, MD 20892-1180
P: (301) 496-6455
F: (301) 402-3521
Bill.Eckelman@nih.gov

CHIEN HO, Ph.D.
Dept. Biological Sciences
Carnegie Mellon University
4400 Fifth Ave
Pittsburgh, PA 15213-2683
P: (412) 268-3395
F: (412) 268-7083
Chienho@andrew.cmu.edu

ALAN KORETSKY, Ph.D.
Dept. Biological Sciences
Carnegie Mellon University
4400 Fifth Ave
Pittsburgh, PA 15213-2683
P: (412) 268-5104
Ak5d+@andrew.cmu.edu

THOMAS MEADE, Ph.D.
California Technical Institute Building 139-74
Passadena, CA 91125
P: (626) 395-2776
Tmeade@gg.caltech.edu

RAVI MENON, Ph.D.
John P Robarts Research Institute
P O Box 5015
100 Perth Drive
London, Ontario Canada N6A 5K8
P: (519) 663-5777
or 663-4148
F: (519) 663-3403
Rmenon@irus.rri.on.ca

RONALD D. NEUMANN, M.D.
Chief, Nuclear Medicine Department
Clinical Center, NIH
Building 10, Room 1C401
Bethesda, MD 20892-1180
P: (301) 496-6455
Ron.Neumann@nih.gov

BRIAN D. ROSS, Ph.D.
Huntington Medical Research Institute
Dept of Magnetic Resonance
660 South Fair Oaks Ave
Pasadena, CA 91105
P: (818) 397-3271
F: (818) 397-3332
Soccss@hmri.org

RICHARD L WAHL, M.D.
University Of Michigan Medical Center
Div Of Nuclear Medicine
1500 E Medical Center Drive
Ann Arbor, MI 48109-0028
P: (313) 936-5384
F: (313) 936-8182
rwahl@umich.edu

RALPH WEISSLEDER, M.D., Ph.D.
Workshop Co-Chairman
Associate Professor of Radiology
Director, Center for Molecular Imaging Research
Massachusetts General Hospital
149 13^th Street
Charlestown, MA 02129
P: (617) 726-8226
F: (617) 726-5708
Weissleder@helix.mgh.harvard.edu

KURT ZINN, D.V.M., Ph.D.
University of Alabama
Dept. of Radiology
1808 7th Ave Boshell, Room 821, Box 11,
Birmingham, AL 25344
P: (205) 975-6414
F: (205) 975-6522
kurtzinn@uab.edu