The role of intestinal integrity and inflammation in the pathogenesis of insulin resistance and non-alcoholic steatohepatitis in morbid obesity.

The incidence of obesity, non-alcoholic fatty liver disease (NAFLD) and insulin
resistance (IR)/diabetes mellitus type II (DMII) is strongly increasing. There
is an important overlap between the development of obesity, NAFLD and IR/DMII.
Obese patients have an increased risk of developing DMII (5-40x) and 20fold
increased risk of developing NAFLD, correlated to increasing BMI. Furthermore,
IR is in itself an important risk factor to develop NAFLD.
NAFLD can lead to liver inflammation, also known as non-alcoholic
steatohepatitis (NASH). Eventually this process can progress to liver fibrosis
and cirrhosis, with a high mortality. Currently there are no non-invasive
methods to determine whether or not patients suffer from NAFLD, and there is no
treatment yet, other than prevention of weight gain. After weight reduction due
to bariatric surgery, liver inflammation diminishes.

In the last decade, researchers showed that activation of the immune system is
an important causal factor in the development of IR/DMII and NASH.{zie
referenties in het onderzoeksprotocol: Shoelson, 2006 #10}
In mice studies, mice deficient for genes encoding the production of
inflammatory cytokines are resistent to the development of obesity-induced
DMII. This is illustrated by the fact that tumor necrosis factor (TNF)-α
knock-out mice do not develop obesity induced IR [3].
Unfortunately, the source of inflammation in the obese population remains
unclear. However, exposure to bacterial endotoxins seems one of the major
players in the development of both NASH and IR/DMII. From previous research we
know that obese mice and rats develop NASH relatively fast in respone to low
dose exposure to bacterial endotoxins {Yang, 1997 #43}.

There is more evidence for the suggestion that living bacteria play a role in
the pathogenesis of NASH, since administration of probiotics reduce liver
inflammation{Li, 2003 #41}. The most likely source of bacterial endotoxins is
the intestine. Strikingly, it is indeed true that mice on high fat diet have
relatively high levels of endotoxin in their portal vein {Yoshimatsu, 2004
#44}. More recently, obese mice have a higher permeable intestine leading to
higher portal endotoxin levels and as a consequence activation of immune cells
in the liver{Brun, 2006 #8}. Another supporting finding is the fact that the
luminar bacterial flora in humans influences bodyweight.{Ley, 2006 #22}
Furthermore, the microflora is and intestinal integrity varies in a mice model
of obese mice with DMII versus controls, and their insulin resistance improves
when they are treated with antibiotics (Cani Diabetes 2007).
The first human results show that, just like in mice, obesity coincides with a
relative overgrowth of bacteria of the Firmicutes class, which is reduced after
weight reduction. {Ley, 2006 #22},{Turnbaugh, 2008 #106}
However, human data are still scarce. In a study of Creely, endotxin plasma
levels in type 2 diabetics was 76% higher than in matched controls{Creely, 2006
#38}. Even though it was unknown whether these patients suffered NASH, it seems
presumable since the majority of the type 2 diabetics has NASH. However, the
official diagnosis NASH can only be diagnosed with histological liver biopsies.
More recently, in patients with NASH, after aspirin intake, there is a higher
intestinal permeability, and a higher level of endotoxins, compared to controls
(Farhadi A, Gundlapalli S, Shaikh M, et al Liver Int. Susceptibility to gut
leakiness: a possible mechanism for endotoxaemia in non-alcoholic
steatohepatitis.2008 Aug;28(7):1026-33).
Moreover patients with NAFLD (preliminary NASH), have higher plasma levels of
both endotoxins and inflammatory markers versus controls (Thuy J Nutr 2008).
Overall these studies are performed with small numbers of patients.

Conclusively, obesity coincides with a disturbed balance in intestinal
microflora, a higher permeability and elevated plasma endotoxin levels. These
endotoxins induce inflammation and thereby contribute to the development of
NASH and insulin resistance.
The treatment of morbid obese individuals with bariatric surgery (laparoscopic
gastric banding, gastric bypass, etc.) leads to a significant reduction of
incidence and prevalence of DMII and NAFLD/NASH. Interestingly enough, parallel
to the improvement of DMII and NASH, there is a reduction of initially elevated
inflammatory mediators involved in the pathogenesis of IR/DMII/NASH.
It is therefore expected that also the intesinal permeability and the endotoxin
plasma levels will decrease after bariatric surgery.

The most important objectives of this study:

1) examine the bacterial translocation as a consequence of elevated intestinal
permeability in the morbid obese compared to controls, and to proove this is
causes an inflammatory status predisposing IR/DMII en NAFLD

2) Investigate the improvement of intestinal barrier integrity, expected after
the weight reduction due to baratric surgery, and subsequent reduction of
inflammation IR and NAFLD in the morbid obese.

We herefor use markers such as intestinal damage, microflora, plasma
inflammatory markers, inflammation in fat and insulin target organs (liver,
muscle) and metabolic parameters.

The hypotheses we wat to further investigate are:

1.Is the intestinal permeability higher in morbid obese with NAFLD and/or DMII
compared to morbid obese without NAFLD and/or DMII?

2.Is the intestinal permeability correlated with plasma endotoxin levels in the
morbid obese?

3.Do plasma endotoxin levels correlate with inflammation, insulin sensitivity
and NAFLD in the morbide obese?

4.Is the intraluminal microflora different in the morbid obese with NAFLD and /
or DMII compared to morbid obese without NAFLD and / or DMII?

5.Does the expected weight reduction after bariatric surgery lead to a
diminishment of the previously elevated intestinal permeability, endotoxin
levels, inflammation, insulin resistence and NAFLD in the morbid obese?

6.Does weight reduction following bariatric surgery change the intestinal
microflora in the morbid obese?

When a patient is eligable to undergo bariatric surgery, he or she is
approached to participate in this research project. When the patient decides to
participate and meets inclusion criteria, an appointment for further
explanations and informed consent is made in outpatient clinic. Prior to
surgery an appointment for the intestinal permeability test is made, where
participants are asked to drink a glass of water with dissolved sugars, and
five hours urine collection takes place.
Before this permeability test, the evening before and the day of the test,
patients are asked to take a tablet of 400mg ibuprofen

The patient will be operated following general guidelines. Patients will be
asked to collect faeces in advance to evaluate the intestinal microflora. At
the day of surgery 20ml blood samples are taken routinely, when a patient
participates in the research project, 6ml extra is taken to determine relevant
standard measurements. The blood sampling takes place in the surgical ward
where the patient awaits surgery, simultaneously with routine blood sampling.

During surgery, biopsies of visceral and subcutaneous fat, muscle and liver are
taken. When the operation involves the intestine (in case of a gastric bypass
or biliopancreatic diversion) a biopsy of the intestine is taken.

During the standard postoperative protocol for follow-up, related to the
bariatric surgery, the patient will return to the outpatient clinic after 6, 12
and 24 weeks, and after one and two years. When patients participate in the
research project, compared to otherwise, we will sample blood samples with an
extra 6 ml.
After one year postoperatively, a subcutaneous fat biopsy is taken and there
will be a second measurement of intestinal permeability (sugars in water, blood
and urine sample after intake of ibuprofen).
Furthermore patients will be asked again to collect faecal material at 12
months post surgery.

A schematic design is illustrated in the research protocol (figure 1).

Risk of testing sugar substances
The saccharide solution consists of nutritional substances approved by the
European Union. These substances have been used previously for research
purposes in these amounts and no (side) effects were reported. In this study
the amounts proposed are very small and therefore the risk of use is expected
to be very small.

The risk of the use of NSAIDs
The non-steroid anti-inflammatory drugs (NSAIDs) administered the evening
before intestinal permeability testing and the morning of the test (in a dose
of 400mg each time) are commonly used drugs freely available in drugstores.
Importantly, the research population, both morbid obese subjects as well as
controls, are often familiar with these drugs.
The NSAID are for example prescribed for pain, artritis, headache, migraine,
tooth extractions, non-bacterial infections, reumatic exacerbations, oedema and
dysmenorroea.

The side effects in these doses are generally mild. Side effects are upset
stomach, most prevalent are nausea, vomiting, heartburn, diarrhea,
constipation, ulcus in case of chronic use. Central effect such as headache,
drowsiness, unusual fatigue and ringing in the ears (tinnitus) may occur.
Very seldom allergies can occur, such as skin reactions (rash, itching,
swelling). However in our population and in the prescribed dosage, these side
effects are most unlikely.
The expected risk is therefore minimal. To further reduce the risk evidently
each patient will be thoroughly questioned about previous drug use and previous
experiences.

Risk associated with interventions
Blood sampling will take place in outpatient clinic, as a service to the
patients, by an experienced physician. Usually, when patients do not
participate in this research projects, they have to turn to the general
hospital blood sampling laboratory. However, we will make sure patients are
sampled for trivial standard follow-up and additionally take 6ml extra blood
for research purposes.
The blood is withdrawn from the lower arm, the procedure causes little
discomfort. The risk is neglicable, it is a standard procedure in this
hospital. The only important risk is when the vein is not correctly found,
there can occur a blue (hematoma). This will however vanish within days.

The subcutaneous fat biopties are sampled in outpatient clinic as well,
simulaneously with a blood sample, in outpatient clinic when the patient would
otherwise also come to the hospital for standard postoperative follow-up.
If necessary, supervision can be present (dr. Bouvy, surgeon in the MUMC+, and
dr. Greve, surgeon in the Atrium Medical Centre).
The procedure is as follows: primarily a standard needle (also used for blood
sampling) is introduced at the previous operation site to prevent scarring and
in order to give a local anaesthetic. Hereafter the procedure is no longer
painfull for the patient (and of course this is also tested).
Maximal 100 mg subcutaneous fat tissue is excised. This is small enough to
leave no cosmetic damage. Possible risks are the occurence of an hematoma or a
superficial infection at the site of intervention. In advance, participants are
explained what they can and cannot expect and in case of doubts or troubles are
requested to contact the researcher.
If necessary, the previously mentioned surgeons can be contacted to reassure
further treatment in case of infection.
However this procedure is performed approximately 80 times in the morbid obese
in the MEC 04-141 and there were no infections. 2 (!?) patients suffered an
hematoma.

The biopsies taken peroperatively (liver- fat- muscle- and intestinal biopsies
if possible) don't cause discomfort since they are taken by the surgeon
performing the operation, and the patient is under general anaesthesia at the
time of collection.
One possible complication is bleeding at the site of the biopsy after
collection (only in the case of fat, muscle and liver).
However in the current study **Role of fatty tissue resident macrofages in the
pathogenesis of insulin resistence in the morbid obese* (MEC 04-141) over 130
participants have been treated comparably. None of the biopsies caused bleeding.

In a similar study of 515 morbid obese where similar biopsies were taken during
gastric banding procedures, there was no complication reported (35).
We do not expect any complications, but in case they might occur, biopsies are
always taken under clear view and possible bleeding can be treated immedeately
by the surgeon.
Given the small risk of complications and the unique oppertunity to study this
material and get valuable information about the inflammatory status of the
patients liver, insulin target organs and the functional integrity of the
intestine, we consider the procedure of taking these biopsies justified.