Efficacy of Amlexanox vs. Placebo in Type 2 Diabetic Patients

This proposal focuses on a clinical proof of concept study to evaluate a novel hypothesis. We
propose that overnutrition, leading to obesity, represents a state of excessive energy
storage promoted by insulin. The stress of excessive energy storage initiates a set of
homeostatic responses that includes the generation of adipose tissue and liver inflammation,
which involves the secretion of proinflammatory cytokines that block energy storage in fat
and liver by inducing insulin resistance, thus relieving stress on the adipocyte and
hepatocyte. However, our recent data suggest that inflammation itself is countered by a
secondary feedback loop, which we have termed "counter-inflammation". We propose that this
process keeps inflammation in check, ensuring that it is maintained at low levels. At the
same time counter-inflammation also promotes energy conservation via increased energy storage
and decreased utilization in a manner that is insulin-independent. We hypothesize that the
noncanonical IkB kinases Ikke and TBK1 are "counterinflammatory" kinases that feedback
inhibit inflammatory pathways while helping to support a positive energy balance.

How can this hypothesis be addressed in human subjects? A number of clinical trials have
already been initiated with the intention of improving insulin resistance by attenuating
inflammation. Many of these have focused on the blockade of single cytokines or their
receptors such as TNFa, IL1 and MCP1 (CCR2 antagonist). While the record on these trials is
mixed, the complexity of inflammatory pathways elicited during obesity suggests that
inhibition of only a single cytokine is not likely to be sufficient. Additionally, because
numerous studies have implicated the NFkB pathway in obesity-induced inflammation, NFkB
inhibitors have been contemplated. Salsalate was originally resuscitated due to its perceived
activity to block NFkB and hence to block inflammation. However, although clinical studies on
this drug have been promising, precisely how salsalate works, what its target is, and whether
it actually functions an NFkB inhibitor remains controversial (recent data indicate that it
may act via activation of AMPK).

We describe here a new potential therapeutic agent that is the result of an academic
collaborative drug discovery effort carried out at the University of Michigan. Our studies on
elucidating the role of NFkB in insulin resistance and macrophage activation led to the
identification of Ikke and TBK1 as potential targets for therapeutic intervention in obesity
and type 2 diabetes. We performed a high throughput screen searching for inhibitors of these
kinases, evaluating a group of 175,000 compounds that included a collection of approved
drugs. Interestingly, the most potent and specific inhibitor discovered was amlexanox; an
older drug originally developed in Japan for the treatment of asthma and allergic rhinitis,
and later in the US for aphthous ulcers, but without a well understood mechanism of action.
It has been in use since 1987, and has a favorable side effect profile.

This protocol is the first proof of concept study to evaluate the possible repurposing of
Amlexanox for the treatment of type 2 diabetes and obesity. At the same time, we hope to test
the hypothesis that IKKe and TBK1 may be important players in maintaining a positive energy
balance, and also in supporting insulin resistance, during prolonged over-nutrition. Finally,
these studies will also provide valuable insight into the validity of these two protein
kinases as drug targets for the future development of new chemical entities useful in the
treatment of type 2 diabetes. Therefore our specific aims are

1. Does amlexanox treatment improve measures of type 2 diabetes, insulin resistance and
obesity? To address this question, we intend to treat patients with clinical Type 2
diabetes and Grade 1/2 obesity (BMI 27 to 45 kg/m2) for a period of 3 months in a
double-blinded placebo controlled proof-of concept study. We will enroll 40 patients
without evidence of severe end-organ damage from chronic complications of Type 2
diabetes who also are on metformin and/or DPPIV inhibitors or no medications, who also
have clear-cut evidence of hyperglycemia (HbA1c>6.5% and <10.0%). These patients will be
treated with either placebo or Amlexanox oral tablets at a starting dose of 25 mg PO tid
for 2 weeks and then titrated up to a dose of 50 mg PO tid (doses used to treat asthma
in humans and leading to serum concentrations comparable to metabolically effective
doses in rodent studies) for the remainder of the therapy period. HbA1c is targeted as
primary end-point in this study. Fasting lipid profiles, plasma insulin, HOMA-IR,
clinical markers of inflammation and adipocytokine profiles will be measured as other
parameters of interest at baseline, 6 weeks and 12 weeks. Insulin secretion as well as
insulin sensitivity will be probed using a mixed meal test. Weight is also an important
but secondary end-point in this study. We will determine resting energy expenditure with
indirect calorimetry and total body composition using DEXA at baseline and 12 weeks.
Safety monitoring with CBC, comprehensive metabolic biochemistry, CPK and urine analyses
will be undertaken at 1, 2, 4, 6, and 12 weeks, and urinary albumin at baseline and 12
weeks. Measures of safety and efficacy will be repeated 4 weeks after the drug has been
discontinued as well.

2. Does amlexanox treatment improve measures of NAFLD in obese patients? All participants
in the clinical proof-of-concept study outlined in the aim above will be studied for the
degree of steatosis in the liver with the use of MRI and MR spectroscopy at baseline and
at 3 months of the study.

3. Development of biomarkers for treatment of patients with type 2 diabetes by amlexanox.
We will evaluate the phosphorylation of proteins known to be targets of IKKe and TBK1 in
blood, and adipose tissue samples. We will use phospho-specific antibodies to evaluate
the regulation of targets that play a role in energy metabolism, and also in feedback
inhibition of NFkB. These studies will pave the way for future evaluations probing the
metabolic relevance of IKKe and TBK1 in humans using the same agent or more potent ones
to be discovered or designed.

Inclusion Criteria:

- ≥ 18 years old at baseline.

- Is male, female not of childbearing potential, or meets all the following criteria if
female of childbearing potential (including perimenopausal women who have had a
menstrual period within one year):

- Not breastfeeding.

- Negative pregnancy test result (human chorionic gonadotropin, beta subunit
[βhCG]) at baseline (not applicable to hysterectomized females).

- Must practice and be willing to continue to practice appropriate birth control
(defined as a method which results in a low failure rate when use consistently
and correctly, such as implants, injectables, oral contraceptives, some
intrauterine contraceptive devices, sexual abstinence, tubal ligation, or a
vasectomized partner) during the entire duration of study period.

- Has physician-confirmed diabetes mellitus with a clear diagnosis or per ADA criteria
with fasting glucose>126 mg/dL or HbA1c >6.4% or 2 hour GTT >200 mg/dL.

- BMI ≥27 and <45 kg/m2.

- On no medications or only on first line oral medications (such as Metformin and/or DPP
IV inhibitors) for treatment of Type 2 diabetes mellitus with a stable regimen for >12
weeks.

- Alcohol consumption of less than 40 grams/week.

- A liver US confirming presence of fatty infiltration of the liver.

- Is able to read, understand and sign the U of M IRBMED approved informed consent form
(ICF), communicate with study physician and study team, understand and comply with
protocol requirements.

Exclusion Criteria:

- On insulin, or other injectables for treatment of Type 2 diabetes.

- Unable to conduct home based glucose monitoring.

- HbA1c <6.5% and >10.0% (set to achieve uniformity in the study population).

- Presence of advanced liver disease (as evidenced by abnormal synthetic function,
abnormal PT or albumin).

- Evidence of other etiologies of viral hepatitis.

- Presence of hematologic, bone marrow and/or other abnormalities.

- Presence of hemoglobinopathy or other hematological abnormalities that will interfere
with accurate measurement of HbA1c.

- Presence of HIV infection.

- Inability to give informed consent.

- Presence of ESRD, any type of active cancer, or >class 2 congestive heart failure
((New York Heart Association Functional Classification System), based on medical
history and physical examination.

- Active chronic infection such as known chronic osteomyelitis or tb, etc. (may be
transient).

- Creatinine >1.5 mg/dL.

- Proliferative diabetic retinopathy, nonproliferative retinopathy is allowed.

- Unable to ambulate.

- Clinically relevant CAD: history of stent, CABG or cardiologist confirmed angina.

- Any other condition in the opinion of the investigators that may impede successful
data collection.