Katholieke Universiteit Leuven Group Biomedical Sciences Faculty of Medicine Department of Experimental Medicine Endocrinology
β-CELL TRANSPLANTATION IN TYPE 1 DIABETIC PATIENTS: A WORK IN PROGRESS TO CURE
Pieter GILLARD
GLYCEMIA
β -CELL GRAFT
400 300 200 100 0 -6
0
6
12
MONTHS Thesis submitted in fulfillment of the requirements for the degree of Doctor of Philosophy in Medcal Sciences-Katholieke Universiteit Leuven-2008
“All we have to decide is what to do with the time that is given to us.” J.R.R Tolkien
To my family
2
Katholieke Universiteit Leuven Group Biomedical Sciences Faculty of Medicine Department of Experimental Medicine Endocrinology
β-CELL TRANSPLANTATION IN TYPE 1 DIABETIC PATIENTS: A WORK IN PROGRESS TO CURE
Pieter GILLARD
Thesis submitted in fulfillment of the requirements for the degree of Doctor of Philosophy in Medical Sciences-Katholieke Universiteit Leuven-2008
3
β-CELL TRANSPLANTATION IN TYPE 1 DIABETIC PATIENTS: A WORK IN PROGRESS TO CURE
Pieter GILLARD
Thesis submitted in fulfillment of the requirements for the degree of Doctor of Philosophy in Medical Sciences-Katholieke Universiteit Leuven-2008
Promotor: Prof. Dr. C. Mathieu (University Hospitals, Catholic University Leuven-KUL, Belgium) Copromotor: Prof. Dr. B. Keymeulen (University Hospital, Brussels Free University-VUB, Belgium) Jury Chair: Prof. Dr. J. Ceuppens (University Hospitals, Catholic University Leuven-KUL, Belgium) Secretary: Prof. Dr. B. Decallonne (University Hospitals, Catholic University Leuven-KUL, Belgium) Prof. Dr. B. Hering (Diabetes Institute for Immunology and Transplantation, University of Minnesota, Minneapolis, USA) Prof. Dr. D. Ysebaert (University Hospitals, University of Antwerp-UZA, Belgium) Prof. Dr. A. Scheen (Centre Hospitalier Universitaire, University of Liège, ULG, Belgium) Prof. Dr. J. Pirenne (University Hospitals, Catholic University Leuven-KUL, Belgium)
Leuven, 26-11-2008 Doctoral thesis in Medical Sciences Katholieke Universiteit Leuven Group Biomedical Sciences Faculty of Medicine Department of Experimental Medicine Endocrinology
4
Table of Contents List of abbreviations
7
Part 1: Background
9
Part 2: Situation in the field I Type 1 diabetes I.1 Definition
11 11 11
I.2 Epidemiology
11
I.3 Pathogenesis
12
I.4 Pathophysiology and clinical presentation
14
I.5 Complications
16
I.5.1 Microvascular complications
16
I.5.1.a Diabetic nephropathy
16
I.5.1.b Diabetic retinopathy
17
I.5.1.c Diabetic neuropathy
17
I.5.2 Macrovascular complications
18
I.6 Therapy
19
I.6.1 Intensive insulin therapy
19
I.6.2 Insulin analogues
22
I.6.3 Continuous subcutaneous insulin infusion
22
I.7 Prevention of diabetes
24
II β-cell replacing therapies
27
II.1 Pancreas transplantation
27
II.1.1 Rationale
27
II.1.2 History
27
II.1.3 Implantation techniques
28
II.1.4 Recent clinical results
30
II.1.4.a Patient selection and timing of transplantation 30 II.1.4.b Benefits of pancreas transplantation
32
II.1.4.c Morbidity and mortality of the procedure
35
II.1.4.d Side effects of immune suppressive therapy
36
II.1.5 Long-term surviving β-cell mass
37
II.1.6 Long-term perspectives
39
II.2 Islet transplantation
43
II.2.1 Rationale
43
II.2.2 History
43
5
II.2.2.a Pre-Edmonton
43
II.2.2.b Edmonton
46
II.2.2.c Post-Edmonton
46
II.2.3 Isolation and implantation techniques
48
II.2.4 Recent clinical results
52
II.2.4.a Patient selection and timing
52
II.2.4.b Benefits of islet transplantation
57
II.2.4.c Morbidity and mortality of the procedure
59
II.2.4.d Side effects of immune suppression
61
II.2.5 Determinants of islet graft function
63
II.2.5.a β-cell mass
64
II.2.5.b Instant inflammatory reaction: IBMIR
71
II.2.5.c The role of autoimmunity
71
II.2.5.d The role of alloimmunity
73
II.2.5.e Toxicity of immune suppression
74
II.2.6 Long-term perspectives
76
II.2.6.a Alternative immune suppression
76
II.2.6.b Escaping the immune suppression
78
II.2.6.c Alternative sites for islet transplantation
80
II.2.6.d Imaging of the β-cell graft
81
II.2.6.e Optimizing islet isolation
82
II.2.6.f Alternative sources of β-cells
83
Part 3: Objective and aims of the thesis
87
Part 4: Own contributions
89
Part 5: Summary of the work
131
Part 6: Conclusions and perspectives
133
Part 7: Samenvatting
139
Publications and presentations
143
References
147
Dankwoord
175
Curriculum vitae
181
6
List of abbreviations ADA ATG APC BDR BW BMI BM CAD CSII CTLA4 CVD CVfg DCCT ESC ESRD ET GADA HDL HSC HLA I(A)A IA2-A IBMIR ICA IEQ IL-2 ITN IVGTT JDRF LDL MHC MRI NHBD NOD OGTT PAK PERV PET PTA SC SPK TAT TNF
American diabetes association antithymocyte globulin antigen presenting cell Belgian diabetes registry bodyweight body mass index bone marrow coronary artery disease continuous subcutaneous insulin infusion cytotoxic T-lymphocyte-associated protein 4 cardiovascular disease coefficient of variation of fasting glycemia diabetes control and complications trial embryonic stem cell end stage renal disease Eurotranplant glutamic acid decarboxylase antibody high-density lipoprotein hematopoietic stem cell human leucocyte antigen insulin (auto)antibody protein tyrosine phosphatase IA2 antibody instant blood-mediated inflammatory reaction islet cell antibody islet equivalents interleukin 2 immune tolerance network intravenous glucose tolerance test juvenile diabetes research foundation low-density lipoprotein major histocompatibilty complex magnetic resonance imaging non-heart beating donor non-obese diabetic oral glucose tolerance test pancreas after kidney transplant porcine endogenous retrovirus positron emission tomography pancreas transplant alone subcutaneous simultaneous pancreas and kidney transplant thrombin antithrombin tumor necrosis factor
7
8
Part 1: Background of the study Type 1 diabetes is characterized by a selective destruction of the insulinsecreting islet β-cells leading to insulin deficiency. At the time of diagnosis, lifelong need for subcutaneous insulin injections is unavoidable. These daily injections represent a heavy social burden for the patient and his/her family, since even with intensive insulin therapy strong fluctuations in glycemia are inevitable, leading to suboptimal glycemic control and the development of chronic complications of diabetes1. These complications (i.e. retinopathy, nephropathy, diabetic foot and amputations, cerebrovascular accidents and myocardial infarct) are responsible for most diabetes related mortality and morbidity. New and experimental therapies aim to prevent or delay these detrimental complications. Transplantation of the insulin-secreting islet β-cells is such an experimental therapy. By implantation of a metabolic adequate number of β-cells, insulin secretion is again adjusted to the metabolic needs and glucose homeostasis is restored. As a consequence, the incidence of life threatening hypoglycemia decreases and chronic complications might be avoided. Animal studies have shown that the technique of β-cell implantation is simple, carries a low risk of procedure related complications and results in long-term correction of glycemic control. For years these results could however not be repeated in type 1 diabetic patients, but in the last 10 years major progress has been made. First, human islet grafts were shown to correct glycemia in type 1 diabetic patients who had received a donor kidney prior to, or simultaneously with, the islet graft
2;3
,
although long-term success remained rare. In 2000 Shapiro et al. showed in their landmark study insulin-independence at 1 year after transplantation in islet-transplanted type 1 diabetic patients in 7 out of 7 cases, giving rise to the ‘Edmonton protocol’. These encouraging results have boosted trials in non-
9
uremic patients by meeting the first requirement for further considering β-cell transplantation as a potential cure for diabetes, namely the ability to prepare islet β-cell grafts that correct glycemia in type 1 diabetic patients. In Belgium the JDRF center for beta cell therapy in diabetes, a collaborative program, was founded in 2001, aiming at improving the risk-to-benefit of islet β-cell transplantation and understanding the major determinants of its success (www.betacelltherapy.org). In the clinical part of that program, pancreases are procured by hospitals affiliated with the Eurotransplant Foundation (Leiden) and send to the central unit in Brussels were the Beta Cell Bank prepares cultured β-cell grafts with defined composition. Implantation is done in the two university hospitals of Brussel (UZ-Brussel) and Leuven (UZ-Leuven). Standardised clinical follow-up is currently performed in the university hospitals of Brussel (UZ-Brussel), Leuven (UZ-Leuven), Bruxelles (Hôpital Erasme) and Antwerp (UZ-Antwerpen). This doctoral project is part of that program.
10
Part 2: Situation in the field I. Type 1 diabetes I.1 Definition Type 1 diabetes is characterized by a selective destruction of the insulinsecreting islet β-cells due to T-cell mediated autoimmunity, ultimately leading to insulin deficiency 4-6.
I.2 Epidemiology Type 1 diabetes accounts for about 10% of all cases of diabetes, affecting about 2 million people in Europe and North America. It occurs most commonly in individuals of European descent and has a marked geographic variation in incidence, even within the same country. In countries such as China the incidence rates are about 0.5 cases per 100 000 per year, while the rate is roughly 20 times higher in Belgium and almost 100-fold higher (about 48–49 per 100 000 per year) in Finland and Sardinia 7. Although some studies show that migrating populations can take on the incidence rates of their new countries 8, others stress the importance of the parental country of birth in determination of the disease incidence 9. In Belgium the incidence of insulinrequiring type 1 diabetes diagnosed before age 40 years is 9.9 cases per 100 000 individuals per year. In the Belgian Diabetes Registry (BDR), there was a significant tendency from 1989 to 2000 toward increasing incidence under age 15 years (1.8%) and even more under age 5 years (5.0%), at the expense of a decreasing incidence between ages 15 and 40 years
10
. The
incidence was similar in both sexes under age 15 years, but marked male excess was noted for adult-onset disease, in particular after age 20 years (male-to-female ratio of 0.9 under age 15 years vs. 1.6 thereafter).
11
These epidemiological features point to an important contribution of both environmental triggers and susceptibility genes to the cause of type 1 diabetes.
I.3 Pathogenesis An essential step forward in understanding the pathogenesis of type 1 diabetes was Gepts’ description of pancreatic islet infiltration by lymphocytes and macrophages (insulitis) in children and young adults who died soon after clinical onset
11;12
. Many years later there is still no all-
explaining theory about the cause of this insulitis, although recent data have expanded our knowledge significantly. It is now widely accepted that the lymphocytic infiltration and consequently β-cell loss are the consequence of a T-cell-mediated autoimmune attack to one or more β-cell autoantigens 5. Additional supportive evidence for the autoimmune pathogenesis of type 1 diabetes comes from the susceptibility of type 1 diabetic patients to other autoimmune conditions including autoimmune thyroid disease (overt dysthyroidism up to 5%, prevalence of thyroid peroxidase-antibody positivity up to 30 %
13
), coeliac disease (in about 3 – 10%) and to a lesser
extent Addison's disease, myasthenia gravis and vitiligo (reviewed in 14). The autoimmune attack occurs in genetically susceptible individuals, probably initiated by some environmental factor(s). Support for a genetic background comes from the observation that the risk for type 1 diabetes ranges from 0.4% when no family history is present, 2% if the mother has diabetes, 6% if the father is affected, over 8 % in dizygotic twins to 50 % in monozygotic twins.
The strongest genetic associations are found with
Human Leucocyte Antigen (HLA) DR and DQ molecules. For instance, DQB1*0602 alleles are associated with protection, and DR3-DQ2 molecules (DQB1*0201) and DR4-DQ8 (DQB1*0302) are associated with susceptibility (reviewed in 12
15
). The HLA locus accounts for about 50% of the genetic
susceptibility, while multiple additional genes are implicated as contributing to diabetes susceptibility i.e. CTLA4 (cytotoxic T-lymphocyte-associated protein 4), IFIH1 (interferon induced with helicase C domain 1), ITPR3 (inositol 1,4,5-triphosphate receptor 3), IL-2 receptor (interleukin 2 receptor) and PTPN22 (protein tyrosine phosphatase, nonreceptor type 22)15. Some other genes are associated with rare syndromes including diabetes (e.g. AIRE and Foxp3). On the background of different genetic susceptibility, one or more environmental factors probably trigger β-cell destruction. The increasing importance of these environmental triggers is illustrated by the observation of a reduced contribution from high-risk HLA haplotypes in the incidence of type 1 diabetes
. Possible triggers that have been linked with type 1
16
diabetes, but until now not clearly identified as the smoking gun trigger, include viruses (e.g. enteroviruses
17
, coxsackie
environmental toxins (e.g. nitrosamines
20
18
, congenital rubella
19
),
) or foods (e.g. early exposure to
bovine milk proteins 21;22, cereals 23, gluten 24) or deficiency of vitamin D 25.
As a consequence of the inflammatory response within the islets, β-cell destruction is also characterised by a humoral response (autoantibody production) to β-cell autoantigens. Different antibodies have been described: islet cell antibodies (ICA), insulin antibodies (IAA), glutamic acid decarboxylase antibodies (GADA), and the protein tyrosine phosphatase IA2 antibodies (IA-2A). More recently, autoantibodies against zinc transporter (ZnT8) were described
26
. Positivity for one or more autoantibodies can
precede the clinical onset of type 1 diabetes, making them important diagnostic markers of preclinical type 1 diabetes
27-30
. There is however no
proof in humans that any of these antibodies has an active role in the pathogenesis of the disease 31.
13
The combination of genetic and immunologic parameters can be used to predict the development of type 1 diabetes among relatives of patients with type 1 diabetes. This prediction based on genetic and immunologic parameters can be quite accurate in the very high and very low risk populations. To discriminate within the largest group of median risk relatives, more extensive metabolic parameters or additional genetic or serologic markers will be needed. At this moment, major research efforts are being put into refining the ways of identifying individuals at risk of type 1 diabetes, in order to make it possible to intervene in the chronic process at an early stage (prevention and intervention trials).
I.4 Pathophysiology and clinical presentation After initiation of the β-cell damage, there is continuing progressive decline of β-cell mass and function evolving over a period of years 32. In general, β-cells are destroyed more rapidly when onset of clinical diabetes takes place at a young age 33, in male
34
and in ICA-positive subjects at diagnosis 35. The exact
surviving β-cell mass at diagnosis remains poorly defined since there are only a few studies of insulitis before onset of diabetes, data were mainly obtained in children
11;12
and methods of β-cell quantification were not standardised nor
compared with normal controls. In a recent study investigating treatment of type 1 diabetic patients at clinical onset with a CD3-antibody, residual β-cell function, measured using a hyperglycemic clamp, averaged 25 percent of normal controls
36
which is higher than the 10 percent that is classically
mentioned.
The continuing decline in β-cell mass and function is first of all evidenced by loss of first-phase insulin response to an intravenous glucose challenge 37;38 and later, when the insulin secretion falls below a critical capacity, usually by impaired glucose tolerance assessed during an oral glucose tolerance test and 14
sporadically by impaired fasting glucose
39
. Finally, when most β-cells have
been destroyed, a state of absolute insulin deficiency and frank hyperglycemia develops for which exogenous insulin has to be started. After this clinical onset, β-cell destruction continues with a marked heterogeneity
36
. Most
subjects progress over months to years to undetectable plasma C-peptide levels (i.e. plasma C-peptide negativity). Initiation of insulin therapy is frequently followed by a honeymoon period during which residual β-cell function tends to recover from exhausted but not yet destroyed β-cells 40. Despite this transient amelioration, lifelong need for subcutaneous insulin injections remains unavoidable.
These daily injections are a heavy social
burden for the patient and his family, since even with intensive insulin therapy strong fluctuations in glycemia are unavoidable leading to suboptimal glycemic control and subsequent chronic complications of diabetes1;41, accounting for most diabetes-related mortality and morbidity. An important finding in the Diabetes Control and Complications Trial (DCCT) trial was that any C-peptide secretion, but especially at higher and sustained levels of stimulated C-peptide, was associated with reduced incidences of retinopathy, nephropathy and hypoglycemia (the major complication of intensive diabetic therapy) 42. Recent data of the DCCT trial showed that longer term fluctuations in glycemia, defined as variation of HbA1c, independently relate to the development of retinopathy and nephropathy 43. Other risk factors for developing chronic complications are early onset and long duration of type 1 diabetes, genetic predisposition (e.g. family history of diabetes-related complications or hypertension), smoking, obesity, sedentary lifestyle, hypertension and hyperlipidemia.
15
I.5 Complications I.5.1 Microvascular complications I.5.1.a Diabetic nephropathy Diabetic nephropathy is the most common cause of renal failure in the developed world. Approximately 20 to 30 percent of type 1 diabetic patients will have microalbuminuria after a mean duration of diabetes of 15 years 44;45. Diabetic nephropathy develops in only a subgroup of patients affected by type 1 diabetes, probably trough genetic susceptibility. Once it is present, fifty to eighty percent of patients with microalbuminuria show progression to other
stages
of
renal
failure
;
46-48
from
subclinical
disease
over
microalbuminuria (defined as a urinary albumin excretion rate >30 300 mg per day) and finally up to end-stage renal failure. More recent studies however seem to indicate that progression is less frequent (in 19% of patients with microalbuminuria) while a considerable number of patients (60%) even regress to normoalbuminuria study of 2005, Finne et al.
49
48
. In a recent
reported in their cohort of 20 005 individuals
between 1965 – 1999 a cumulative prevalence of end-stage renal disease of 2.2% at 20 years and 7.7% at 30 years, which are much lower than previously estimated. This is probably due to the increased attention paid to glycemic control, more aggressive blood pressure reduction and the use of angiotensin converting enzyme inhibitors. Prevention is currently considered the first goal to achieve. Tight glucose control from the time of diagnosis is probably the most important factor to prevent nephropathy in type 1 diabetes 1, since hyperglycemia has immediate effects on kidney function and structure. Annual screening of individuals with type 1 diabetes for microalbuminuria should be instituted early after diagnosis. After a positive screening, additional renoprotection should be started, including tight control of hypertension and hyperlipidemia
50
. The
current drugs of choice in this context are angiotensin-converting enzyme 16
inhibitors or angiotensin receptor-blocking agents, since they are both highly effective in slowing progression of diabetic nephropathy
50,51
. If however
evolution to end-stage renal disease can not be prevented, it is important to know that these patients have a particularly poor prognosis when on dialysis, and effort should be directed toward timely pancreas-kidney transplantation.
I.5.1.b Diabetic retinopathy Among patients, diabetic retinopathy is probably the most feared complication of diabetes. After > 20 years of diabetes duration, it has a prevalence of almost 100% for non-proliferative retinopathy and 20-25% for proliferative retinopathy 52. It progresses from early non-proliferative changes (background retinopathy: microaneurysms, exsudates and haemorrhages) over preproliferative retinopathy to proliferative retinopathy (with risk of retinal detachment and vitreous haemorrhage) and macular oedema. The later stages make diabetic retinopathy the most common cause of acquired blindness in the western world. Screening for diabetic retinopathy should also begin early after diagnosis and be performed annually. There is also clear evidence that strict control of glycemia, lipids and blood pressure can lower the risk of developing diabetic retinopathy and reduce disease progression 1. If present, diabetic retinopathy often requires laser therapy in case of sightthreatening pathology.
I.5.1.c Diabetic neuropathy Diabetic neuropathy is the most common neuropathy in the Western world 53. Approximately 50 percent of patients with diabetes will eventually develop neuropathy 54. Diabetic neuropathy can be focal (e.g. carpal tunnel syndrome, peroneal nerve and third cranial nerve palsies, and diabetic amyotrophy) or generalised
(e.g.
sensorimotor
polyneuropathy).
Sensorimotor
polyneuropathy often also affects the autonomic system with cardiac 17
dysfunction, gastroparesis and erectile dysfunction. It is a risk factor for diabetic foot and amputations. Screening for neuropathy should also start early using specific methods (e.g. monofilament test to detect loss of distal sensitivity) and clinical assessment for the other manifestations. Attention to preventive foot care is an essential component of diabetes management. Neuropathy is still difficult to treat and remains symptomatic despite the introduction of several newer medications 55.
I.5.2 Macrovascular complications The relative risk of cardiovascular disease in type 1 diabetes is up to 10-fold higher than that in non-diabetic individuals 56. An incidence of coronary artery disease (CAD) of 9% over 7 years 57 of followup in type 1 diabetic patients has been reported. In patients ≥ 35 years total incidence of CAD has been estimated at >2% per year diabetic
nephropathy,
autonomic
neuropathy,
hypertension are the most important risk factors.
44
. The presence of
hyperlipidemia
and
Patients with type 1
diabetes for instance have more severe progressive coronary artery atherosclerosis for any level of low-density lipoprotein cholesterol
41;58
. Risk
reduction therefore includes attention to healthy lifestyle (weight control and physical activity), smoking avoidance, optimising glycemic control, blood pressure (