CTGTAC Meeting # 47
Animal Models for Porcine
Xenotransplantation Products Intended to
Treat Type 1 Diabetes or Acute Liver
Failure
May 14, 2009
BRIEFING DOCUMENT
INTRODUCTION
..1
BACKGROUND
2
PRODUCTS
...3
The
FDA/CBER Biological Response Modifiers Advisory Committee (BRMAC)
Xenotransplantation
Subcommittee
...3
The
Secretarys Advisory Committee on Xenotransplantation (SACX)
....3
CLINICAL
CONSIDERATIONS
..4
PRECLINICAL
CONSIDERATIONS
...5
Overall
Concepts
5
Animal Models of Acute Liver Failure (ALF)
.5
Selected
Published Examples of the Use of BAL Devices in Animal Models of ALF
..9
Animal Models of Type 1 Diabetes (T1D)
..10
Small Animal
Models of T1D
...11
Large Animal
Models of T1D
...12
Concluding
Comment
13
ADVISORY
COMMITTEE DISCUSSION QUESTIONS
..14
1.
Animal
Models of Acute Liver Failure
..14
2.
Animals
Models of Type 1 Diabetes
..14
REFERENCES
......15
INTRODUCTION
This meeting is convened to
provide the Food and Drug Administration (FDA) with insights and perspectives
regarding the use of available animal models to support the administration of
porcine xenotransplantation products in patients with acute liver failure (ALF)
or Type 1 Diabetes (T1D). No specific
products will be presented for regulatory review at this meeting. The source
origin of these products will be restricted to porcine. Animal husbandry and
product manufacturing issues will not be discussed in the meeting. At this
meeting, invited experts who are developing animal models to evaluate
xenotransplantation products to treat these two diseases will present their
viewpoints on the major issues confronting the development of these
models. Members of the committee will be
requested to consider this information, as well as the information presented in
this document, and to discuss and provide advice regarding FDA questions
provided in this briefing package. Discussions
of specific issues related to risks of xenogeneic infections are beyond the
scope of this meeting. These issues are addressed in multiple guidances and
guidelines and contribute to the basis of FDA assessment of risks to subjects
in clinical trials [1-5]. Finally,
discussions of genetically engineered animals are beyond the scope of this
meeting and are addressed in FDA guidance [6].
BACKGROUND
FDA is responsible for assessing the safety and efficacy of new therapies, as well as ensuring that subjects enrolled in clinical trials of such products are not exposed to unreasonable risk. These responsibilities include regulation of clinical trials of xenotransplantation products. According to both the PHS guideline released in January 2001 entitled PHS Guideline on Infectious Disease Issues in Xenotransplantation, and the FDA guidance document released in April 2003 entitled, Guidance for Industry: Source Animal, Product, Preclinical, and Clinical Issues Concerning the Use of Xenotransplantation Products in Humans, xenotransplantation is defined as any procedure that involves the transplantation, implantation, or infusion into humans of: 1) live cells, tissues, or organs from a nonhuman animal source or 2) human body fluids, cells, tissues, or organs that have had ex vivo contact with live nonhuman animal cells, tissues, or organs [1-5]. Xenotransplantation products may be able to offer novel treatments for patients suffering from a variety of diseases and injuries. Ongoing scientific research is directed towards improving the potential of new xenotransplantation products to provide clinical efficacy, while continuing to investigate issues related to safety.
The potential risks of xenotransplantation in the targeted patient population must be considered in the overall regulatory decision-making process for permitting the initiation of a first-in-human clinical trial. These risks include, but are not necessarily limited to: 1) the transmission of known pathogens, 2) the potential for introducing a new infectious disease into the general population, 3) the potential for adverse inflammatory and immunological responses of the host to the product or its secreted proteins, 4) the potential for rejection of the source animal cells/tissues/organs and any associated adverse effects, 5) the risks of using immunosuppressive agents in an attempt to prolong the transplanted graft, and 6) potential zoonotic risks to personal contacts and healthcare professionals. Although preclinical animal studies are conducted in order to characterize these risks, current investigational modalities cannot provide absolute assurance that all risks will be fully identified and evaluated. In developing clinical trial protocols, prospective sponsors should take the preclinical data into account when considering what the possible benefit to subjects from participating in the trial would be and/or how the resultant data from the first-in-human trial would contribute substantially to scientific knowledge and prudent product development. Demonstration of pharmacodynamic action of the investigational product will be helpful in these considerations. Given the potential risks due to the particular product, administration procedure, and concomitant medications and treatments, studies of animal models of the targeted disease are important to estimate the potential for positive pharmacodynamic action of the xenogeneic product, as well as the potential for harm. The goal of this meeting is to obtain expert advice regarding the availability and use of relevant animal models to evaluate and define the safety and clinical activity of porcine xenotransplantation products to treat patients with ALF or T1D.
Xenotransplantation products are subject to FDA regulation
under Section 351 of the Public Health Service (PHS) Act (42.U.S.C.262) and the
Federal Food, Drug and Cosmetic Act (21 U.S.C. 321). In addition to various published guidances
intended to assist investigators in this field [1-5], several public advisory
committee meetings have been held to discuss scientific, medical, ethical,
social, and regulatory issues on xenotransplantation products.
The FDA/CBER
Biological Response Modifiers Advisory Committee (BRMAC) Xenotransplantation
Subcommittee [7]
The first BRMAC Xenotransplantation Subcommittee meeting was held in December 1997 in response to FDAs need to address regulatory issues associated with these products. This subcommittee consisted of 15 voting members, including the Chair, who was also a member of the BRMAC. The members had expertise in xenotransplantation, epidemiology, virology, microbiology, infectious diseases, molecular biology, veterinary medicine, immunology, transplantation surgery, public health, applicable law, bioethics, social sciences, patient advocacy, and/or animal welfare. Nonvoting participants also included experts from the Centers for Disease Control and Prevention (CDC), the National Institutes of Health (NIH) and FDA/CBER. This subcommittee was tasked with addressing risk assessment and management for xenotransplantation products and to highlight and discuss new scientific data generated in this rapidly evolving field. The last subcommittee meeting was held in 2000.
The Secretarys Advisory Committee on Xenotransplantation (SACX) [8]
The Secretarys Advisory Committee on Xenotransplantation (SACX) was chartered in 1999 to advise the Secretary of the Department of Health and Human Services (DHHS) on all aspects of the scientific development and clinical application of xenotransplantation. The committee considered the scientific, medical, social, and ethical issues in the context of patient and public health concerns raised by xenotransplantation. These discussions extended to ongoing and proposed clinical trial protocols. Based on these public meetings, the SACX made recommendations to the Secretary of DHHS on policy and procedures associated with this pioneering area of research. The goal of the SACX was to facilitate DHHS efforts to develop an integrated approach to addressing emerging public health issues in xenotransplantation. The SACX consisted of 18 voting members that were appointed by the Secretary or a designee, with expertise as described for the FDA BRMAC Xenotransplantation Subcommittee. At least one SACX member was also a current member of the FDA BRMAC and at least one member was a current member of the CDC Healthcare Infection Control Practices Advisory Committee. Additional non-voting members from DHHS agencies including the Office of the Secretary, CDC, FDA, Health Resources and Services Administration, NIH and others as deemed appropriate by the Secretary or a designee, participated. The first SACX meeting was held in February 2001; the last meeting was held in February 2004.
In addition, FDA actively participates in international
activities in the field of xenotransplantation because of the potential global
impact of communicable disease outbreaks.
Agency efforts in the global community include participating in the WHO
efforts to develop international recommendations regarding xenotransplantation
products.
CLINICAL
CONSIDERATIONS
Xenotransplantation products offer the
possibility of replacing, for varying periods of time, failed organ, glandular,
or tissue function in conditions for which human-derived cells and tissues are
in limited supply. Acute liver failure (ALF) and type 1 diabetes (T1D)
represent particularly important examples of such diseases, because the need
for cellular treatment options far exceeds the current or foreseeable supply of
human-derived differentiated cells and there are no other currently-available
or approved sources of such differentiated cells.
As reported in the published literature,
investigational cellular products to treat ALF consist of hepatocytes or liver
tissue placed in an extracorporeal blood circulation device, termed a Bioartificial
Liver [BAL] assist system [9-11]. The intent is to use the system
intermittently or continuously, under intensive care monitoring, until a liver
is available for orthotopic transplantation or the patient spontaneously
recovers. The acute clinical situation, which is life-threatening, generally
reaches resolution (recovery, transplantation, or death) within weeks. To date,
there are no approved human or xenogeneic cellular products or BAL systems
containing human or xenogeneic cells or tissues for this indication.
For the treatment of Type 1 diabetes (T1D),
the medical community has had considerable experience with allogeneic islet
transplantation under the regulatory auspices of Investigational New Drug (IND)
applications, as several hundred patients have received human islets via
intraportal infusion, with immunosuppression.
The major benefits, which have been demonstrated in a substantial
proportion of subjects in clinical trials, are insulin independence,
attenuation or disappearance of severe hypoglycemia, and improved metabolic
control without increasing the frequency of hypoglycemia, even in recipients
who still require some exogenous insulin. The principal limitations of this
approach are the transient nature of clinical benefits (generally of the order
of a few years), the need for continuous immunosuppression, and the limited
supply of human islets. Current supplies
of cadaveric islets could satisfy less than 1% of potential demand, even
assuming very stringent clinical and metabolic criteria for treatment. Porcine-derived islets may offer the
potential to provide cellular therapy for a large number of diabetic patients.
Porcine and human beta cells have similar metabolic responses, and porcine
insulin has been used therapeutically for decades. In addition, the use of
islet encapsulation may reduce or eliminate the need for immunosuppression and
may extend xenogeneic islet survival post-transplant. Accordingly, porcine
islets, which are in plentiful supply, offer the possibility of treating large
numbers of metabolically unstable diabetic patients.
PRECLINICAL
CONSIDERATIONS
Overall Concepts
According to Title 21 of the
Code of Federal Regulations (CFR) Part 312.23 (a)(8), the sponsor of a clinical
trial should provide
adequate information about the pharmacological
and toxicological studies
on the basis
of which the sponsor has concluded that it is reasonably safe to conduct the
proposed clinical investigations. The
kind, duration, and scope of animal and other tests required vary with the
duration and nature of the proposed clinical investigations. The design of
preclinical studies is a critical determinant of their ability to provide
appropriate and sufficient data to support clinical development of a
product. For example, as previously
stated in this document, the administration of xenotransplantation products in
humans poses many risks, thus comprehensive preclinical studies can provide
data to be considered in planning clinical trials to identify and characterize
these concerns. Selection of the most
appropriate animal model(s) to evaluate a xenotransplantation product intended
for a specific clinical disease will serve to provide insight regarding
dose/activity and dose/toxicity relationships.
This information will provide the most reliable determination of the
therapeutic potential of the product, as well as assess the likely duration of
clinical effect. Duration of effect is
important, given the inherent risks of xenotransplantation products. An understanding of the biological actions of
the xenotransplantation product following administration in animals that model
the intended clinical disease to the extent possible (pathophysiology,
metabolically, immunologically) will potentially help to mitigate some of the
risks to humans enrolled in clinical trials.
Animal Models of Acute Liver Failure (ALF)
ALF presents as a multi-system syndrome, occurring as a consequence of acute or acute-on-chronic failure of hepatic function. Major elements of the syndrome include hepatic encephalopathy (HE), jaundice, coagulopathy, severe metabolic abnormalities, renal insufficiency and hemodynamic instability [12]. The pathophysiologic basis of HE is incompletely understood and considered to be multifactorial [13]. Elevations of blood levels of ammonia and other toxic substances, together with numerous ALF-associated metabolic and electrolyte derangements, may contribute to the syndrome. These associated disorders include hyponatremia; hypokalemia; hypophosphatemia; lactic acidosis; hypoglycemia; acute pancreatitis; and renal, respiratory, and circulatory failure.
The majority of extracorporeal BAL devices that are currently under development for clinical use in the ALF setting consist of liver tissue or hepatocytes obtained from porcine source animals. This approach is attractive because: 1) porcine tissue/hepatocytes are readily available, 2) important metabolic activities of porcine hepatocytes (i.e. clearance of ammonia and other metabolites, urea synthesis, p450 activity) are similar to those in humans, and 3) many of the porcine hepatic tissue/cell preparations can tolerate a wide range of handling and storage conditions and still have the potential to remain functional in a BAL device [9].
Numerous etiologies and complications of ALF in humans render a
notable challenge in the development of an animal model that accurately
reflects this condition. Many
different procedures have been used to induce ALF in small and large animal
species. Among these approaches are: 1) total or partial liver resection
(hepatectomy model); 2) complete or transient hepatic devascularization by
ligation or temporary clamping of the portal vein and hepatic artery (ischemia
model); 3) administration of hepatotoxic drugs and chemicals (such as CCl4)
(hepatotoxic drug/chemical model); and 4) exposure to viral agents (infectious
model). We refer the reader to Belanger
and Butterworth 2005 for a comprehensive review of many of these models, and
the table
provided in this document is adapted from this reference [14].
The infectious models are probably the least favored approach due to the potential for human exposure to the hazardous infectious agents used to create the model [14]. Therefore we will not review them here.
The hepatotoxic drug/chemical models are generated via the use of agents such as acetaminophen, galactosamine, and thioacetamine (TAA). These are compromised due to the lack of inter-animal reproducibility, including the degree of toxicity relative to chemical dose, the clinical manifestations and laboratory abnormalities of ALF, and the time to death. In addition, extrahepatic toxicities occur commonly. For example, administration of acetaminophen in pigs and dogs has been reported to cause toxicity to the kidneys, heart, and lungs due to methemoblobinemia caused by the oxidation of hemoglobin by acetaminophen. TAA has been reported to cause neurotoxicity in animals, independent of liver toxicity, limiting the ability to assess the potential for a BAL to relieve hepatic encephalopathy in this model.
Such direct extrahepatic toxicities can generally be avoided in the hepatectomy and ischemia models, but the latter rely heavily on the surgical technique and expertise of laboratory personnel to ensure that reproducibility of the model is achieved. Models that depend on a total hepatectomy or a complete devascularization are too severe, as death inevitably results, with no evidence of liver regeneration/spontaneous recovery. Animals that are subject to a total hepatectomy are more artificial in nature, as the patho-physiology of an injured liver to link to the clinical setting does not exist. For example, encephalopathy is detected only a few hours before death in this model, whereas it is observed early on in patients with ALF, and progressively worsens with time. Minimal hematologic and hemodynamic changes and brain edema have also been reported with this model, which does not reflect the severity of these symptoms when humans present to the hospital. Animals that are subjected to a partial hepatectomy do not usually progress to hepatic coma, and this model has been historically poor in recapitulating the clinical manifestations of ALF in patients.
Progressive encephalopathy and hepatic coma, accompanied by brain edema and intracranial hypertension, which are observed in patients with ALF, have been documented in animals exposed to complete devascularization (without hepatectomy). While these abnormalities are also manifested in animals exposed to transient devascularization, the ability to reproduce these laboratory and clinical changes with acceptable consistency is relatively poor.
The generation of an animal model of ALF through a combined surgical and hepatoxic drug/chemical-induced approach has also been reported. An example is a rat ALF model created via a partial hepatectomy (70% liver resection), followed by intravenous injection of endotoxin [15]. However, induction of an ALF model via this combined strategy is infrequently reported in the scientific literature.
Various animal
models of ALF reported in the published literature
|
Model |
Species Used |
HE |
Intracranial Hypertension |
Reproducibility |
Reversibility/ Spontaneous Recovery |
Hazard to Study
Personnel |
||||
|
Pig |
Dog |
Rabbit |
Rat |
Mouse
|
||||||
|
Acetaminophen |
Φ |
Φ |
Φ |
|
|
Yes |
NA |
+/- |
Potential |
No |
|
Galactosamine |
|
Φ |
Φ |
Φ |
|
Yes |
Yes |
+/- |
Potential |
Yes |
|
Thioacetamide |
|
|
Φ |
|
Φ |
Yes |
Yes |
Yes |
Potential |
Yes |
|
Viral hepatitis |
|
|
Φ |
|
|
NA |
Yes |
+/- |
Potential |
No |
|
Total hepatectomy |
Φ |
Φ |
|
Φ |
|
Yes |
Yes |
Yes |
No |
No |
|
Partial hepatectomy |
Φ |
|
|
Φ |
|
No |
NA |
Yes |
Yes |
No |
|
Complete hepatic devascularization |
Φ |
Φ |
Φ |
Φ |
|
Yes |
Yes |
Yes |
No |
No |
|
Transient hepatic devascularization |
Φ |
|
|
|
|
Yes |
NA |
+/- |
Potential |
No |
HE = Hepatic encephalopathy;
NA = data not available
Reversibility = possible liver regeneration
[Belanger M and
Butterworth RF (2005). Acute liver failure: a critical appraisal of available
animal models. Metab Brain Dis
20:409-23]
Selected Published Examples of the Use of BAL Devices in Animal Models
of ALF
Flendrig and colleagues used a
BAL device containing porcine hepatocytes in a rat ALF model created by an
end-to-side portacaval shunt placed three days before hepatic artery and bile
duct ligation [16]. The rats were
treated with the device, starting 30 minutes after ligating the hepatic artery
and bile duct until death. The study endpoints included HE assessment
(neurological symptoms of normal to deep coma) and determination of arterial
pressure, blood ammonia levels, lactate levels, and hepatic enzyme levels. Per the investigators, a treatment effect was
demonstrated by reduction of blood ammonia and lactate levels, and by
improvement of HE scores (p<0.05). All the non-treated control and empty
device (lack of porcine hepatocytes)-treated control animals died between 3 to
8 hours after inducing liver ischemia, while the treated animals died between 8
to 14 hours after inducing liver ischemia.
Abouna and colleagues studied the
use of a BAL device containing dog liver or calf liver in dogs (6/group) that
had end-to-side portacaval shunts put in place the day prior to induction of
ALF by occlusion of the hepatic artery for 2 hours [17]. At the onset of ALF,
which was defined as the appearance of encephalopathy, hyperammonemia,
hyperbilirubinemia, and elevated prothrombin time and hepatic enzymes
(occurring about 10-12 hours after clamping the hepatic artery); the dogs were
treated using the BAL device for 6-8 hours.
The study endpoints included: serum bilirubin levels (excretion of toxic
metabolites), hepatic enzyme levels, blood ammonia levels (detoxification),
prothrombin time, various physiological parameters (i.e. portal vein pressure,
hepatic artery pressure, body temperature, bile flow, blood pressure, pulse
rates), overall survival, and gross and microscopic examination of the animals
liver. According to the authors, a
treatment effect was demonstrated by a reduction in bilirubin levels and
decreased hepatic necrosis in the dogs.
The control animals that received medical support only, died between 32
to 38 hours post shunt placement, while the treated animals died between 60
hours to 7 days post shunt placement.
Fruhauf et al, conducted a study
in a pig ALF model generated by a total hepatectomy [18]. One hour after the
surgery the pigs (n=6/group) were treated continuously using a BAL device
containing porcine hepatocytes until death.
Note that the plasmapheresis cartouches were replaced every 8
hours.
The study endpoints included albumin and ammonia levels and assessment
of intracranial pressure (ICP). The
investigators reported treatment effects demonstrated by increased albumin
levels, decreased ammonia levels, a lower ICP, and extension of survival (24.8
±4.3 hours) compared to untreated control animals (16.4 ±4.7 hours).
The studies using animal models
of ALF described in the current scientific literature were largely conducted
using a single cycle of exposure to the BAL device (i.e., a single treatment). However, due to the wide variation in
individual device design, such as: 1) the xenogeneic components
(hepatocytes vs. liver tissue), 2) the quantity/weight/volume of the xenogeneic
components, 3) the rate of extracorporeal blood flow through the device, and 4)
product viability (pre-and post-treatment), as well as the possibility of a
host immune response to the source transplantation product, a single cycle of
treatment may not be sufficient to achieve a clinically significant biological
effect. Thus the duration of a single treatment and the ability to repeat the
treatment, as would quite likely be required in the clinical setting, is an
important consideration when developing animal models of ALF. In this regard, it is also useful to consider
the theoretical assessments of Iwata and Ueda, in which the importance of
hepatocyte mass and blood flow are discussed [19].
From the foregoing, the
development of appropriate animal models for assessment of products designed to
treat ALF has been fraught with numerous complexities and difficulties. The ability of existing animal models to
reflect the immunologic, metabolic, and physiologic aspects of patients with
ALF and the application of these models to inform first-in-human trial design
are important considerations.
Animal
Models of Type 1 Diabetes (t1d)
Rapid progress in several scientific areas
may enable use of relevant animal models of diabetes in preclinical evaluations
of xenotransplantation products prior to introduction of clinical trials in
patients. Of particular importance are: 1) the increased understanding of the
molecular and cellular basis of immune suppressive/regulatory mechanisms and
the subsequent development of animal models that are genetically or
immunologically modified to tolerate xenotransplantation products; 2) advances
in the ability to introduce genetic modifications in the pigs that will be the
source animals; and 3) progress in techniques for the encapsulation and
immunoisolation of xenogeneic islets.
Considerations for the use of an animal
model(s) to generate safety and efficacy data using porcine islets include: 1)
the method of induction of a diabetic state (i.e., spontaneous, chemical,
surgical, immunological), 2) the final porcine islet formulation (i.e.,
encapsulated vs. unencapsulated), 3) the
transplant procedure, 4) the glucose metabolic set points of the recipient
animal versus porcine islet set points, 5) the immunological profiles of the
porcine source animal and of the recipient animal, 6) the duration of a clinically meaningful
effect post-transplant and the need for re-transplantation to maintain this
effect, and 7) identification of relevant parameters that indicate substantial
clinical benefit (i.e. insulin requirements, fasting and glucose challenged animal
source and recipient C-peptide levels, HbA1c, etc...). For each animal model that has been used in
an attempt to satisfy these factors, advantages and limitations exist.
T1D
occurs as a result of the complete or nearly complete destruction of the
insulin-producing beta cells in the pancreas. Thus, for products designed to
treat T1D, the induction of a diabetic
state in recipient animals has generally been accomplished via one of three
modalities that adversely affect beta cell function. Spontaneously diabetic animals, commonly rodents (e.g. Non-Obese
Diabetic [NOD] mice and Bio Breeder [BB] rats), permit the investigation of
xenotransplant islet function in an autoimmune setting that more closely
approximates the immunologic dysfunction associated with the human disease. In particular, as with humans, the MHC plays
a role in the susceptibility of NOD mice and BB rats to diabetes. Transplanted
islets in these two models are subjected to immune attack by T cells, B cells,
NK cells, and macrophages, accompanied by the production of a variety of
autoantibodies, such as anti-GAD and anti-insulin antibodies, similar to the
human scenario. Insulitis occurs in NOD
mice at around 4-5 weeks of age, with frank diabetes presenting at
approximately 12-30 weeks of age. The BB
rats exhibit weight loss, polyuria, polydipsia, hyperglycemia, and insulopenia
at about 12 weeks of age, displaying a severe ketoacidosis that is fatal
without exogenous insulin administration. However, differences between the human and animal autoimmune models do
exist. For example, ketoacidosis in NOD mice is mild and animals can survive
for weeks without exogenous insulin. There are also differences in genetic
susceptibility markers that could potentially confound study outcome. Even so, spontaneous models of diabetes
permit the study of a highly genetically controlled population from which
valuable information can be interpreted from xenotransplantation studies [20].
The second modality used to induce a diabetic
state in large or small animal species is via chemical induction. The
two most common agents employed are the toxic glucose analogues streptozotocin
and alloxan. Streptozotocin, a
nitrosurea derivative from Streptomyces
achromogenes, induces a diabetic state when injected intraperitoneally in
multiple small doses typically over five consecutive days (in order to avoid
direct toxicity to the pancreas and other organs) [21]. Alloxan, an oxygenated pyrimidine derivative,
is a toxic glucose analogue that accumulates in the beta cells via the GLUT2
glucose transporter and generates reactive oxygen species that ultimately
trigger the death of the cells. As with
streptozotocin, alloxan also inhibits glucose-induced insulin secretion [22].
When either diabetogenic agent is used in an animal species, adequate numbers
of animals and utilization of appropriate controls are needed to ensure that
baseline parameters (e.g. glucose, C-peptide, etc...) are established for a
sufficient period of time both before and after a diabetogenic state has been
reached. Based on the existing
scientific literature, chemical induction is a primary means of generating a
large animal diabetic model.
The third modality used to create an animal
model of diabetes is via the surgical removal of the pancreas. Although pancreatectomies have been primarily
conducted on large animal species, such as dogs and nonhuman primates (NHPs),
generation of this model has also used mice and rats. Use of pancreatectomized animals can be
compromised by complications of the surgery itself and by the complete removal
of both endocrine and exocrine functions of the organ [23]. Pancreatic regeneration can also occur in the
case of incomplete resection [24].
Small
Animal Models of T1D
Small animal models of T1D can provide a
wealth of information on disease initiation, progression, and potential
therapeutic strategies. These models provide many advantages, including
availability, cost, the potential for large sample sizes, and the existence of
a wide array of diverse rodent strains that have been extensively studied. In
addition, genetic manipulation of rodents, utilizing transgenic and knockout
models, provides a relatively rapid tool for dissecting the molecular and
cellular mechanisms of the disease itself, as well as a means to provide
initial assessment of xenotransplant safety and potential benefit [25-26].
Porcine islet xenotransplantation studies in
small animal models have helped to establish preliminary rationale and safety
for novel immunosuppressive regimens, enabled rigorous characterization of
molecular and cellular mechanisms of immune reactions and immunomodulation, and
allowed for evaluation of the biocompatibility/biostability and potential
effectiveness of encapsulation techniques [27-31]. However, translation of the data generated
with these models to patients with T1D has limitations, such as: 1) significant
immunological differences between rodents and humans, including MHC class II
expression and the existence of preformed antibodies to pig antigens in humans;
2) metabolic differences (e.g., glucose set points, animal activity, and
feeding cycles); and 3) dissimilarities in general physiology and other
parameters due to large differences in physical size. These impediments provide
a substantial challenge to the use of small animal models for the preclinical
assessment of safety and pharmacodynamic function of islet products prior to
testing in humans [25, 32].
Recent advances utilizing an array of
transgenic and knockout mice may now provide more useful models to investigate
pig-to-human xenotransplantation. For instance, studies of discordant
xenotransplantation in pig-to-αGal-KO mice or the use of mice
reconstituted with human immune cells (humanized mice) may provide powerful
tools linking to a more clinically relevant scenario [33-34].
Large
Animal Models of T1D
Large animal models, mainly dogs, pigs, and
NHPs, provide genetic, metabolic, and physical profiles that more closely
resemble the human situation [33]. Some of the major factors supporting the use
of large animal models of T1D are immunological responses to the porcine islets
and to any encapsulation components: metabolic set-points: product placement
via the intended clinical procedure into the planned clinical anatomical site:
and the ability to re-transplant the porcine islets (as will be the likely
clinical requirement for an efficacious product). Based on the published
literature, the diabetic state is primarily generated by chemical or surgical
means. These animal models typically do
not recapitulate the autoimmune condition that exists with the human disease,
nor do they manifest the microvascular complications or hypoglycemic
unawareness that often accompanies patients with established T1D. Nonetheless,
these models can yield useful information to define the safety and possible
benefits of xenogeneic islet transplantation, and thus potentially have the
ability to provide some degree of assurance of clinical efficacy prior to the
first-in-human trial.
NHPs, like humans, possess xenoreactive
natural antibodies including the well-studied Galα1-3Galβ1-4GlcNac
(αGal) specific antibodies, which can lead to HyperAcute Rejection (HAR)
and Acute Humoral Xenograft Rejection (AHXR) in vascularized tissue/organ
xenotransplants. Islets typically express
little αGal, but the use of NHPs could provide added confidence when
assessing the success or failure of porcine islet preparation methods,
immunosuppressive regimens, or encapsulation techniques. In addition, genetically engineered pigs that
lack αGal or express complement inhibitory factors may further eliminate
concerns over HAR and AHXR [32].
Additional immunological concerns that have
been well characterized in NHPs include Instant Blood-Mediated Inflammatory
Reactions (IBMIR) and the adaptive immune responses. IBMIR, responsible for
early loss (within days) of intraportally transplanted porcine islets, involves
activation of platelets, the coagulation process, complement cascades and the
infiltration of neutrophils. This reaction has also been studied in athymic
mice after intraportal injection of adult porcine islets [35]. IBMIR concerns
can potentially be minimized through the use of immune modulating transgenic or
knockout source pigs, immunosuppression, encapsulation, and/or the transplantation
of porcine islets into a different anatomical location. However, delayed
adaptive immune responses that appear to be mediated primarily by T cells, B
cells, and macrophages through mechanisms that are still not fully understood,
remain major obstacles to successful long-term implantation in NHPs, as well as
in other large and small animal models of diabetes. Even so, recent published studies in NHPs
have reported porcine islet survival and function for
up to 6 months post-transplantation utilizing a multi-agent immunosuppressant
regimen or encapsulated islets [36].
In addition to the above-stated issues, other
considerations when investigating the use of large animal models of T1D must
also include cost, animal husbandry issues, availability of highly trained
personnel, animal availability, and group sample size [37-38]. An understanding
of the limitations and capabilities inherent in each animal model of T1D can
provide insight into selection of the most appropriate species/model. When considering the preclinical data that
will provide support regarding the safety of the xenogeneic islet cells and any
associated agents/delivery devices that are used, it may be of benefit to
consider a tiered approach using multiple disease models in order to address specific
questions/unknowns. The limitations and
advantages of existing animal models and what can be learned from these models
to guide first-in-human trial design are important considerations.
Concluding Comment
As mentioned above, the use of animal models
is essential to support the development and regulatory evaluation of
xenotransplantation products for the treatment of ALF and T1D. Further as
summarized above, the details of these complex models may place significant
limitations on the conclusions that can be drawn from any specific study. Therefore the FDA is seeking scientific
advice in this area to guide informed decisions regarding potential models in
future applications that may be submitted for review.
ADVISORY COMMITTEE DISCUSSION QUESTIONS
FDAs decision to allow the initiation of a first-in-human clinical trial using a xenotransplantation product is based on careful examination of all available preclinical data, as well as the clinical indication and review of the proposed clinical trial protocol. This overall appraisal is greatly dependent on selection of appropriate animal models that can be employed to evaluate safety and demonstrate substantial pharmacodynamic activity of the xenogeneic product.
Animal Models of Acute Liver Failure (ALF)
1. Please discuss the limitations and capabilities of available animal models of ALF to assess the safety and clinical activity of bioartificial liver assist devices containing porcine cells or tissues as a bridge to spontaneous recovery or liver transplantation. Please consider the following in the discussion:
a. The ability of the animals to model the clinical manifestations and laboratory abnormalities of ALF in humans.
b. Treatment duration and the ability to repeat the treatment, as would likely be required by the clinical condition of ALF patients.
c. Study
endpoints the changes in laboratory values and clinical responses in test
animals that would be considered clinically meaningful and predictive of
potential clinical benefit in patients
Animal Models of Type 1 Diabetes (T1D)
2. Porcine islet products are currently under development to treat Type 1 diabetics who are chronically metabolically unstable. Please discuss the limitations and capabilities of available animal models of Type 1 diabetes that can be used to assess the safety and clinical activity of porcine islet cell transplantation. Please consider the following in the discussion:
a. The ability of the animals to model the immunological and metabolic manifestations of Type 1 diabetic patients.
b. Treatment duration and the ability to re-transplant, as would likely be required by the chronic clinical condition of Type 1 diabetic patients.
c. Study endpoints the changes in laboratory values and clinical responses in animals that would be considered clinically meaningful in diabetic patients.
d. The intended clinical immunosuppression regimen, as applicable.
REFERENCES