
Diabetic gastroparesis: a review

by Joseph Tripodi, D.O. and Robert Burakoff, M.D.

From the Voice Editor: Although the following article is highly technical, it
provides valuable information for laypersons as well as health professionals.
For people with diabetes who are experiencing problems with digestion, this
information may be beneficial. Moreover, many physicians will be able to
better serve patients after reviewing this article.

If you are a diabetic, I urge you to share this report with your doctor. He or
she may be enlightened and better able to serve patients.

Gastroparesis is a motility disorder that results in delayed emptying of
gastric contents without evidence of outlet obstruction. Diabetes mellitus
remains one of the most common conditions associated with delayed gastric
emptying, especially in those patients with known autonomic neuropathies.
Providing relief of clinical symptoms while maintaining adequate nutrition and
good blood glucose control is a challenging problem. Advances in the
understanding of gastric motility and improvements in diagnostic techniques
have aided the clinician in making the diagnosis. The recently developed drug
therapies have dramatically altered the treatment strategy.

Pathophysiology

The stomach can be considered in two physiologic parts: proximal (fundus and
body) and distal (lower body and antrum).
The proximal stomach acts as a reservoir for food and adaptively dilates in
response to gastric distention. This adaptation occurs by inhibition of the
basal gastric contraction and is abolished after vagotomy is performed.
The distal stomach acts to mix and grind solid food to a chymelike consistency
(<2 mm in size). Distal gastric contractions are regulated by a pacemaker
found on the greater curve of the upper body. The pacemaker generates
depolarizing events that are propagated to the pylorus cyclically at a rate of
3-4 per minute. Action potentials are stimulated on top of these depolarizing
events by the presence of food and neurotransmitters. Action potentials cause
muscular contractions that increase in amplitude and velocity as they move
distally. As the peristaltic wave approaches the antrum, the pylorus closes,
causing mixing and grinding of solid food. Solids are finally emptied after
being altered to a chymelike consistency. Liquids are emptied down a pressure
gradient created by the basal tonic pressure in the stomach. Indigestible
solid material is emptied by a fasting electro-mechanical activity known as
the migrating motor complex, which occurs every two hours. Factors which are
important in the rate of gastric emptying are summarized in Figure 1.
In diabetes, various abnormalities can account for abnormal gastric emptying.
These abnormalities include increased basal pyloric tone, the absence of the
interdigestive migrating motor complexes, and the presence of antroduodenal
incoordination. Some studies in people with diabetes have shown that
morphological damage to the vagus nerve and elevated blood glucose levels may
be responsible for delayed gastric emptying.

Clinical Presentation

Gastroparesis is usually a complication of long-standing insulin-dependent
diabetes. It reportedly effects 20% to 30% of patients with diabetes mellitus,
with the average duration of diabetes prior to the onset of symptomatic
gastroparesis being slightly more than 18 years. However, using modern
diagnostic techniques, abnormal gastric motility can be detected in a greater
number of patients and often before clinical symptoms are apparent.
Liquids usually empty from the stomach quickly in the first 15 minutes in the
majority of diabetic patients. This is due to the lack of the normal adaptive
relaxation of the proximal stomach. It is the ingestion of solid foods that
usually produces gastrointestinal symptoms in most diabetics with
gastroparesis. Indigestible food matter retained in the stomach can lead to
the formation of bezoars due to the lack of migrating motor complexes.
Symptoms are often nonspecific and do not correlate well with the degree of
delay in gastric emptying. Epigastric pain, nausea, vomiting, early satiety,
belching, postprandial fullness, and weight loss are among the symptoms that
suggest the diagnosis. Gastroparesis is usually associated with other evidence
of autonomic dysfunction such as postural hypotension, abnormal response to
the Valsalva maneuver, bladder dysfunction, and impotence. Constipation and
diarrhea may also be found. The presence of a succession splash over the left
upper quadrant on physical exam may indicate a problem with gastric emptying,
but it is neither sensitive nor specific for the diagnosis of diabetic
gastroparesis.
A careful search should be conducted to determine the etiology of
gastroparesis (Figure 2). Upper-gastrointestinal endoscopy is recommended as
the primary diagnostic procedure for this purpose. Endoscopy defines anatomic
gastric outlet obstruction and can exclude morphological lesions, including
ulcers or tumors. Endoscopy may also be therapeutic in the presence of bezoars
or retained food particles. Endoscopy is not useful, however, for detecting
motility disorders.
Gastric emptying can be measured by intubative techniques, radiological
techniques, or radioisotopic techniques. The saline load test is an intubative
technique that requires the insertion of a nasogastric tube and instillation
of 750 cc of saline into the stomach. Recovery of more than 200 cc of fluid 30
minutes later is indicative of gastric retention. Though easy to perform, the
test can only assess liquid emptying and therefore is of limited value in
patients with diabetic gastroparesis as liquid emptying is usually normal in
these patients. Radiological techniques assess the emptying of liquid barium
sulfate or radiopaque solid meals. Unfortunately, only the time of complete
emptying of the stomach can be measured, as residual barium cannot be
accurately quantified. Furthermore, barium-impregnated food may not reflect
solid emptying, as the barium granules rapidly disassociate into a liquid
phase.
The radioisotope method using external scanning has proven to be a reliable,
non-invasive technique for quantitatively measuring gastric emptying. The most
widely used radioisotope is technetium 99m (99m Tc). Technetium 99m bound to
sulfur colloid is injected into a live chicken. Ninety percent of the colloid
is taken up by the liver. The liver is cooked and then ingested by the patient
on an empty stomach as a test meal. In our hospital, the meal consists of 7.5
oz. of untagged beef stew and 20 g of tagged chicken liver, delivering a total
dose of 200eCi. The radiation exposure is considered well within the tolerable
range. External scanning is begun at 1 minute and recorded continuously for 2
hours for solids and 20 minutes for liquids. A gastric-emptying profile is
computer-generated, and abnormal emptying of a solid meal is defined as
greater than 50% retention after 120 minutes. It is also possible to record
both liquid-phase and solid-phase gastric emptying simultaneously using dual
radioisotopes. This diagnostic technique also allows assessment of response to
prokinetic drugs. Most authorities now agree that the radioisotope
gastric-emptying scan is essential to secure the diagnosis of gastroparesis.
Measurements of gastric impedance using cutaneous electrodes or the use of
real-time ultrasound have yet to achieve a role in clinical practice.

Treatment

The main goal of therapy for gastroparesis is to promote gastric emptying of
solid food. First, treatment should be directed toward correction of fluid and
electrolyte imbalance. Large meals should be replaced by small meals, which
are better tolerated in gastroparesis. Indigestible foods, such as green
vegetables or other high-fiber foods, may induce postprandial symptoms and
should be avoided. Associated factors contributing to delayed gastric emptying
should be treated; these include concomitant peptic ulcer disease and reflux
esophagitis. The administration of anticholinergic and narcotic drugs should
be discontinued if possible. Glucose levels above 200 mg/dl have been shown to
delay gastric emptying, so better glucose control may be beneficial. Drug
therapy remains the mainstay of treatment for this disorder, and several
prokinetic drugs have recently been approved to expand the pharmacologic
armamentarium.
Bethanechol, a cholinergic agent, was used initially for the treatment of
gastroparesis because earlier studies showed that it increased the frequency
and amplitude of gastric contractions, but radionuclide studies failed to
confirm accelerated transit times. The inconsistent data and an unfavorable
side-effect profile has led many physicians to abandon its use as a prokinetic
agent.
Metoclopramide, a procainamide derivative, has strong cholinergic and
antidopaminergic effects. The drug significantly accelerates gastric emptying
by inhibiting fundic receptive relaxation and coordinating gastric, duodenal,
and pyloric motility. In diabetic gastroparesis, the drug also activates
peristalsis by initiating previously absent interdigestive migrating motor
complexes. Metoclopramide exhibits an antiemetic effect by passing through the
blood-brain barrier and directly influencing the chemotactic trigger zone.
Therapy with metoclopramide has been shown to evoke symptomatic relief of
diabetic gastroparesis even without simultaneous improvement in gastric
emptying. This improvement may be explained by its central effects on the
chemotactic trigger zone. The usual dosage is 10 mg orally four times a day,
given approximately 15-30 minutes before each meal and at bedtime. Improved
gastric emptying of solids and liquids in diabetic gastroparesis has also been
demonstrated with intravenous metoclopramide at similar dosages. Severe
gastroparesis unresponsive to oral metoclopramide has been reported to respond
to rectal administration. Although metoclopramide is effective initially,
there are two uncontrolled studies on a small number of patients that indicate
the therapeutic effect may partly disappear with long-term therapy. Further
investigation needs to be done before a final recommendation can be made
regarding long-term treatment. Side effects (Figure 3) occur in 10%-20% of
patients and are severe in 1%. Metoclopramide should not be used in patients
with intestinal obstruction, perforation or hemorrhage, Parkinsonism, or
epilepsy.
Domperidone, a benzimidazole derivative, is a powerful prokinetic agent that
acts by blocking dopamine inhibition of acetylcholine release at dopamine D2
receptors located on postganglionic cholinergic neurons. Short-term therapy
with intravenous or oral domperidone improves gastric emptying of liquids and
solids, but long-term administration enhances liquid emptying only. While
gastric emptying of solids appears unchanged with long-term therapy, patients
do have some symptomatic relief. The most commonly studied dose is 20 mg taken
orally four times per day. However, dosages as low as 10 mg four times a day
were successful in improving symptoms in patients with diabetic gastroparesis
that was unresponsive to metoclopramide. Intravenous administration has been
associated with cardiac arrhythmias and is not recommended. A major advantage
of the drug is that it does not cross the blood-brain barrier and therefore
has fewer CNS side effects. However 10% to 15% of patients develop side
effects related to drug-induced elevation of prolactin levels. Other side
effects include dry mouth, transient skin rash, headache, diarrhea, and
nervousness. Although domperidone has been used successfully in Canada, at
this time it is not currently approved by the FDA for use in the United
States.
Cisapride, a substituted piperidinyl benzamide, is a new class of drug that
does not have antidopaminergic properties but is believed to increase gastric
motility by enhancing the release of acetylcholine from the myenteric plexus
in the gut. In vitro studies demonstrate that cisapride as well as
metoclopramide have 5-hydroxytryptamine (5-HT4) receptor agonist activity, and
this may in part explain their prokinetic activity. In healthy volunteers,
cisapride has been found to increase antral and duodenal contractions and
improve antroduodenal coordination. Studies in people with diabetes illustrate
that the effects of cisapride (both oral and IV) are dose related and
correlate well with plasma drug levels. The improvement of solid-phase gastric
emptying with cisapride was greater than that with metoclopramide at similar
doses. In contrast to metoclopramide and domperidone, long-term administration
of cisapride has been associated with improved gastric emptying and symptom
control for greater than one year. Two placebo-controlled double-blind trials
conducted exclusively on patients with diabetic gastroparesis have
demonstrated accelerated gastric emptying and improved symptoms. However, in
two other double-blind crossover placebo-controlled studies, there was no
significant difference in gastric emptying or symptom improvement, but there
was a trend toward decreased epigastric fullness and less diarrhea. These
differences may reflect the limited number of patients studied or the presence
of underlying medical problems that may have masked the therapeutic benefits.
Perhaps higher dosages may be necessary to achieve sufficient drug levels for
a desired clinical effect. In the above-mentioned studies, cisapride did not
produce a significant change in glycemic control, and adverse effects were
rarely reported (Figure 3). Despite the appearances of effectiveness and the
narrow side-effect profile, cisapride has not been approved by the FDA for the
treatment of diabetic gastroparesis. In fact, its only approved use is for
symptomatic treatment of nocturnal heartburn due to gastro-esophageal reflux
disease.
Erythromycin, a macrolide antibiotic, has been found to stimulate gastric
motility by binding to motilin receptors, which are found mainly in the
gastric antrum and proximal duodenum. This effect is related to its molecular
structure and not its antibiotic activity, a fact that may be helpful in
developing newer prokinetic agents without antimicrobial activity. One study
showed that 200 mg of erythromycin ethylsuccinate given intravenously
postprandially improved gastric emptying of radiolabeled liquids and solids in
diabetic gastroparesis. This effect persisted after four weeks of oral therapy
with 250 mg erythromycin ethylsuccinate taken three times a day 30 minutes
before meals. A more recent study confirmed these findings using an oral dose
of 500 mg four times a day one-half hour before meals and before bedtime.
However, a significant improvement in clinical symptoms was not demonstrated,
perhaps due to the small sample size studied. Thirty percent of subjects could
not tolerate this dose due to side effects but tolerated a lower dose of 250
mg or 125 mg four times daily, thus supporting the need for individual
titration of dosage and frequency. An oral suspension may be the preferred
form of administration because it allows the greatest dosing flexibility. The
most common side effects are abdominal pain, nausea, vomiting, and diarrhea.
The potential for a reversible chemical hepatitis with the estolate form would
favor the use of the ethylsuccinate form of erythromycin. Serious liver injury
was not seen in the studies mentioned above, but careful observation for
interactions with drugs metabolized by the P450 system should be performed to
ensure safety. The role for long-term treatment with erythromycin has not been
proven. In fact, in vitro work with rabbits showed that motilin receptors may
be down-regulated by continuous stimulation with erythromycin. There have been
no published trials comparing erythromycin to other prokinetic agents.
Surgical therapy employing pyloroplasty, gastric resection,and
gastrojejunostomy have been tried. These are usually reserved for severe
refractory gastroparesis. Roux-en-Y gastrectomy has been reported to have some
success when conservative surgical procedures have failed. An endoscopically
placed double-lumen gastrostomy tube (one lumen in the duodenum for feeding
and other lumen in the gastric fundus for decompression) has also been
successful in selected cases.

Conclusion

Gastroparesis is a common complication of diabetes mellitus. The clinical
presentation is often nonspecific and does not correlate with pathology. The
radionuclide gastric-emptying scan is the best test to confirm the diagnosis.
Although metoclopramide is the only drug approved for the treatment of
diabetic gastroparesis, cisapride, domperidone, and erythromycin may be
logical alternatives when efficacy is limited or when side effects prevent the
use of metoclopramide. The combination of agents with different mechanisms of
action may also play a future role in the treatment strategy. Further studies
are needed to determine the role of these drugs in the long-term management of
diabetic gastroparesis.

Dr. Tripodi is a Gastroenterology Fellow with the Department of
Gastroenterology and Nutrition, Winthrop-University Hospital, Mineola, New
York. Dr. Burakoff is Chief of the Division of Gastroenterology at the same
institution.

(Note: This article appeared in Practical Diabetology. Reprinted with
permission.)   

((( above story with 3 charts )))

Figure 1.

Factors that affect gastric emptying

Factors that increase gastric emptying
    Volume (liquids)
    Food composition:
          Neutral pH
          Iso-osmolar
          Calorically inert
Factors that decrease gastric emptying
    Acidity
    Hyperosmolarity
    Fat
    Amino acids

Figure 2.

Differential diagnosis of gastroparesis

Mechanical
Pyloric stenosis due to ulcers
Obstructing gastric cancer
Prolapsing gastric polyp
Infantile and adult hypertrophic pyloric stenosis

Functional
    Metabolic
          Diabetic gastroparesis
          Hypothyroidism
          Uremia
          Pregnancy
    Medications
          Anticholinergic drugs
          Narcotics
          l-dopa
          Tricyclics
    Infiltrative diseases
          Amyloid
          Gastric malignancy
    Neuromuscular
          Dermatomyositis
          Muscular dystrophy
          Myotonic dystrophy
          Scleroderma
    Psychiatric
          Anorexia nervosa
          Bulimia
    Surgery related
          Vagotomy
          Gastric resection
    Idiopathic
          Gastric dysrhythmias

Figure 3.

Side effects of gastroparesis medications

Metoclopramide
     Drowsiness
     Restlessness
     Anxiety
     Constipation
     Diarrhea
     Dystonic reactions (1%)
     Tremor
     Rigidity
     Akinesia
     Raises prolactin
Domperidone
     Raises prolactin
     Nervousness
     Headache
     Dry mouth
     Diarrhea
Erythromycin
     GI upset
     Hepatotoxicity
     Rash
     Reversible hearing loss
Cisapride
     Abdominal cramps
     Diarrhea
     Headache
     Dizziness