Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

EXPERT REPORT by sofiaie


Pr Pierre Déchelotte
University Hospital of Rouen

                   Gastrointestinal and nutritional
                adverse effects of Zavesca° (miglustat)

        This expert report has been established for registration purposes at the request of
Actelion to summarize the data on gastrointestinal and nutritional side effects reported in the
preclinical and clinical of the product in Gaucher's disease and other diseases.
        This report is based on the available literature of clinical studies and of the
confidential investigator's brochure provided by Health Experts on behalf of Actelion. It
reflects the expert's independent analysis of the subject and proposals for clinical counselling
of patients treated with miglustat.

Introduction :
        Miglustat (OGT 918) inhibits glucosylceramide synthase, the initial enzyme in the
generation of glycosphingolipids. Thus, it offers a new approach for the treatment of the
Gaucher's disease and related pathologies, so called substrate reduction therapy, in
replacement of or in association to the available substitutive enzyme therapy aiming at
restoring the hydrolysis of accumulated glycosphingolipids.

Description of available data:
        Abundant preclinical and clinical safety data are available, documenting
pharmacokinetics and safety data in various species including humans, as well as acute or
chronic toxicity data in animals, and in vitro pharmacodynamics. Clinical efficacy and safety
data are limited to 3 completed studies in type1 Gaucher's disease, with ongoing extension,
and two additional ongoing studies. Additional clinical safety data are provided by one pilot
study in Fabri's disease. Recent studies in other related pathologies (Niemann-Pick disease
type C, GM2 gangliosidosis, type 3 Gaucher's disease) are not yet contributive.
        Previous studies of miglustat in HIV provide kinetics and safety in several groups of
patients that contribute to the understanding of side effects. However, they are not fully
relevant to the intended present registration, since dosage in HIV patients was much higher
(up to 3000 mg/d) than that intended in Gaucher's disease (100 to 200 tid).


      The main adverse event observed with miglustat overall is diarrhoea. Associated
symptoms such as flatulence, abdominal cramps, nausea are also frequent.

Miglustat and diarrhoea: preclinical data :

       Pharmacokinetics indicate that miglustat is readily absorbed, both in animals and
humans. The pharmacokinetics of the drug is linear over a wide range of doses. The main
routes of elimination are the urine and the feces, both accounting for about half of the dose.
Miglustat is poorly metabolised in all species studied. Autoradiography studies have been
performed with the perbutyrated prodrug OGT 924 in the rat. Both studies (MRC-92B-0345

and MRC-92-0008) indicate that the drug is highly concentrated in the gastrointestinal tract,
mainly in the stomach and caecum, and in the small bowel to a lower extent. In vitro
incubation with mucosal homogenates indicate that the drug is poorly metabolised by
gastrointestinal tissues (MRC-93S-0028).

        In accordance with these kinetics data, toxicology data confirm that the
gastrointestinal tract is a major target for the drug. In acute toxicity studies, rats given the top
dose of 4200 mg/kg/d exhibited stress-related changes with mucus depletion and atrophy of
the small bowel. The dominant feature of the toxicity was profound watery diarrhoea,
mimicking osmotic diarrhoea. Similar although less severe observations were gained with
1680 mg/kg.d (PSA-89S-3378). Yet, lower daily doses were not associated with digestive side
effects. Diarrhoea was also observed in dogs and monkeys receiving the highest doses in
escalation toxicology studies (PSA-89C-3394, 3467, and 3346).
        Some toxic effects on the gut were also observed during subacute and chronic toxicity
studies in rats (PSA-90C-3476), including diarrhoea and red stools. In monkey, vacuolisation
of pancreatic acinar cells was also observed. Again, it must be emphasized that only the top
doses induced digestive side effects in these experiments.

       Thus, preclinical data indicate that gastrointestinal mucosa is a key target for
miglustat, but that toxic effects are observed only for very high doses.

Miglustat and diarrhoea: clinical data

        In a phase I/II open-label study, 28 adult patients with Gaucher disease received oral
miglustat over 12 months (Cox et al 2000). Dosage was initiated at 100 mg TID, and
individualised to a maximum of 300 mg TID. Efficacy was demonstrated on the reduction of
liver and spleen volume and evolution of haematological variables. The most frequent adverse
effect was diarrhoea, which occurred in 79% of patients, shortly after initiation of OGT 918.
Diarrhoea was usually mild and responded well to treatment with loperamide or codeine
phosphate. In addition, diarrhoea tended to improve spontaneously over time. In two of the
three patients whose dose was increased after 6 months, diarrhoea recurred transiently.
Gastrointestinal complaints caused treatment withdrawal in two patients.
        A similar study was performed by Heitner et al in 18 patients with type 1 Gaucher
disease over 6 months. Starting dose was 50 mg TID and could be reduced to 50 mg BID in
case of unacceptable toxicity or if the trough concentration exceeded 2 µg/ml. Diarrhoea
occurred in 94% of patients which was generally mild, excepted in one patient who withdrew
because of severe diarrhoea and flatulence. In most cases, diarrhoea occurred during the first
month of therapy and responded well to treatment with loperamide or to lactose-free diet.
Flatulence and abdominal pain were also reported in 50% of patients. During the extended
period of this study (OGT 918-003X), one subject withdrew because of severe digestive
symptoms. Despite these effects, clinically significant improvements were seen with the
Quality of Life survey from baseline to month 6.
        Since miglustat and substitutive enzyme therapy (Cerezyme) have different
mechanisms of action, a study was performed in 36 patients with type 1 Gaucher disease to
compare each treatment alone to the combination of both (OGT 918-004). The dose of
miglustat was started at 100 mg TID and could be reduced if necessary. All subjects in the
miglustat group and 11 out of 12 in the combination group reported at least one treatment-
related adverse event in the gastrointestinal system, as compared to only 50% in the Cerezyme
group. Diarrhoea occurred in all patients in the miglustat group. Intensity of adverse events

was mild to moderate, and the incidence of diarrhoea tended to decrease from weeks 0-4 to
after week 13 in the main study, and decreased further after week 39 in the extended study.

       Some additional safety data on miglustat in type 1 Gaucher disease have been also
obtained from the ongoing study by Pastores et al (OGT 918-005) and Zimran et al (OGT
918-016). In addition, some former studies in HIV patients with much higher doses of
miglustat (up to 3000 mg/d) reported high incidence of gastrointestinal side effects, mostly
diarrhoea and flatulence, of moderate intensity, that responded well to treatment with
loperamide or carbohydrate reduction diet (Tierney et al, Fischl et al). Diarrhoea occurred also
in 94% of patients with Fabri’s disease treated with miglustat (OGT 918-002).

       In summary, diarrhoea appears as a very frequent adverse effect during miglustat
treatment with doses ranging 50-300 mg TID; other associated symptoms are also frequent.
Some observations suggest a dose-relationship for diarrhoea. Overall, the severity of
diarrhoea was mild, and it responded well to usual treatment with loperamide or diet


       A frequent adverse effect reported in both preclinical and clinical studies with
miglustat is a moderate weight loss.

Miglustat and weight loss : preclinical data.

       During pharmacokinetics study in rats, a slowering of weight gain was observed in the
two top dose groups (PSA-90S-3493). Reduction of weight gain was also a common finding
of acute toxicity studies, again with top-dose rats and to a lesser extent in the mid-dose
animals (PSA-89S-3341). In dogs receiving an intermediate dose, loss of appetite and weight
loss was observed (PSA-89C-3394). Monkeys given even low dose exhibited loss of appetite
and some weight loss (PSA-89C-3346).
       During subacute toxicity studies, weight gain was reduced in top dose rats (PSA-90C-
3476). In a one-year chronic toxicity study in rats, weight gain reduction appeared dose-
dependently (PSA-91C-3490).

        In summary, preclinical data exhibit weight gain slowering of weight loss, depending
on species, to be associated to miglustat use, with a clear dose-dependency, the most severe
effects being seen at very high toxic doses, while mid or low doses, more relevant to the
normal pharmacological dose-range are associated with limited influence on weight.

Miglustat and weight loss : clinical data.

        In the study by Cox et al in adult patients with Gaucher disease, weight decrease
affected 50% of patients, with a mean weight loss of 6-7%, which recovered progressively
during the extension phase, during(OGT 918-002) the second or third year of treatment.
        Similarly, 12 out of 18 patients exhibited some degree of weight loss (> 5%) in the
first 6 months of treatment in the study by Heitner et al in 18 patients. Overall, most patients
remained in the same category of body mass index. Only one patient moved to the overweight
category, while three over-weighted patients moved to normal BMI, and only two moved
from the normal category to slight underweight. Only one patient had a severe weight loss

during the extended period of this study (OGT 918-003X), which was not drug-related but
occurred due to concurrent severe illness.
        In the study by Zimran et al (OGT 918-004) comparing miglustat alone to Cerezyme
and combination therapy, change in body weight from baseline to month 6 (-2.7 kg) was
significant in the miglustat as compared to Cerezyme groups; moderate weight loss also
occurred in the combination group. Of the three treatment groups, the miglustat alone group
showed the greatest decrease in BMI classification, with three subjects moving from normal
to underweight, and two moving from overweight to normal. During the extended study,
weight loss between month 6 and 12 did not increase in miglustat groups, and even tended to
be reduced, although not significantly (-4.6 to -3.5 kg, respectively). In contrast, in the
Cerezyme group, mean weight reduction from baseline was greater at month 12 than at month
6 (4.4 kg versus 1.0 kg, respectively).
        The additional data in type 1 Gaucher disease obtained from the ongoing study by
Pastores et al (OGT 918-005) and Zimran et al (OGT 918-016) have been included in the
overall assessment of safety in 80 subjects exposed to miglustat. Overall, weight decrease is
associated with miglustat therapy in 56 patients (70%), while 5 (6%) patients gained weight.
Weight loss tended to be much less frequent after one year of treatment; this finding however
is only qualitative since only one third of patients remained treated after one year in the
available studies.
        In patients with Fabri’s disease treated with miglustat (OGT 918-002), weight loss was
also common and observed overall in 11/16 patients (60%); in contrast to Gaucher disease
patients, weight loss tended to be more frequent after 6 months of treatment, but remained
        The incidence of weight loss in HIV patients (Tierney et al, Fischl et al) was notably
lower (overall 10%) than that observed in Gaucher disease, despite a much higher dosing.

       In conclusion, weight loss is common finding in patients with Gaucher and Fabry’s
disease treated by miglustat. Overall, weight loss is moderate, with significant clinical
consequences, and tend to ameliorate with time.


        In all preclinical and clinical studies, there seems to be a good time-correlation
between the occurrence of diarrhoea and associated symptoms and weight loss. Both adverse
effects remain generally moderate in intensity, and tend to ameliorate with time.

       Several mechanisms could be discussed to explain diarrhoea, among which some may
also contribute explain weight loss:

1/ accelerated transit due a primary motor effect : although data specifically addressing this
issue are limited, miglustat appears to have no effect on gut smooth muscle motility, at least
in vitro (BRD-92D-1625), and do not influence gastrointestinal transit (BRD-88D-1381).

2/ volumogenic diarrhoea : this mechanisms is observed in some diseases such as Zollinger-
Ellison syndrome, in response to massive gastric acid secretion; it seems unlikely, since
miglustat was observed rather to decrease acid secretion (BRD-88D-1381).

3/ lymphoid blockade and subsequent lipid malabsorption : this was not found in a rat study
examining mesenteric lymph nodes, excepted in top dose animals (K00/024).

4/ pancreatic insufficiency and subsequent global malabsorption: although vacuolisation of
pancreatic acinar cells was observed in some toxicity studies, a clinical relevant pancreatic
insufficiency is unlikely because of the high functional reserve of pancreas, and the tendency
to recovery with time. However, it could be worthy to address this issue by a pragmatic
approach of pancreatic enzyme therapy in the first months of treatment.

5/ impairment of intestinal absorption and subsequent osmotic diarrhoea :
        This mechanism seems the most likely.
        Indeed, miglustat is non-specific enzymatic inhibitor. Besides its desired effect on
glucosyltransferase, miglustat is a weak inhibitor of several other enzymes including -
glucosidase and disaccharidases. Thus, miglustat may inhibit to some extent the intraluminal
digestion of starch and the final hydrolysis and absorption of oligosaccharides; it may also
inhibit lactase activity. Consequently, intraluminal unhydrolysed oligosaccharides or lactose
drive a significant osmotic gradient with subsequent intestinal secretion of water and
electrolytes in the small bowel, and entry of a high volume of fluid in the colon. Colonic
bacterial fermentation is a compensatory mechanism that tends that provides some metabolic
salvage of these glucidic caloric supply by means of short chain fatty acid production. Despite
this, carbohydrate malabsorption in the gut usually causes some degree of symptomatic
osmotic diarrhoea, and enhanced colonic fermentation results in several associated symptoms
such as flatulence, abdominal distension and discomfort, nausea. In turn, these symptoms
frequently lead to loss of appetite and fear for eating to reduce post-prandial diarrhoea.
        The fact that diarrhoea and weight loss alleviated with time is consistent with the fact
that adaptation mechanisms may occur: induction of glucosidase activity in the small bowel;
enhanced fermentation capacity in the caecum, due to the adaptation of the microflora to
enhanced substrate availability for fermentation processes. Similar adaptative processes have
been well described in short bowel patients, and during long term treatment with the osmotic
laxative lactulose.

        Thus, diarrhoea appears likely due mainly to osmotic mechanisms secondary to partial
inhibition of carbohydrates hydrolysis and absorption in the small bowel. Weight loss is likely
due mostly to spontaneous reduction of food intake as a consequence of post-prandial
digestive side-effects, although carbohydrate malabsorption may contribute to some extent to
the negative caloric balance.


        Thus, diarrhoea and weight should be considered as an intrinsic effect of miglustat and
explained to the patient, as being moderate and time-evolving side effects. It should be
anticipated that increasing doses would result in increased frequency and intensity of these
side effects.
        Patients should be advised to undergo loperamide therapy routinely, in order to
prolong the stay of luminale contents in the small bowel and optimise carbohydrate digestion
and absorption. This should limit the abdominal discomfort and diarrhoea. In addition,
patient’s tailored therapy with trimebutine may be useful to alleviate abdominal distension.
        Patients should avoid very high intake of carbohydrates (bread, potatoes, rice, paste,
sodas, pastries…) on one meal, and rather turn to additional small snacks between meals if
they tend to loose weight. Increase intake of fructose (as fruit or fructose-enriched beverages)
may also be helpful since absorption of single ose is unlikely to be affected. In case of severe
weight loss, increasing caloric intake in the form of lipids is easy to achieve, by addition of
standard long chain triglycerides in the diet, or by using medium-chain triglycerides oil or

margarine (Liprocil°, Ceres°) in case of poor tolerance or contraindication to long chain
triglycerides supplementation. In case of suspected pancreatic insufficiency (fatty stools), a
dosage of faecal elastase-1 should be prescribed. A treatment with pancreatic enzymes
(Eurobiol°, Alipase°, Créon°) could be considered.
        Since colonic fermentation is an adaptative process limiting the severity of diarrhoea,
attention should be paid to the risk that concurrent antibiotic therapy would increase the
intensity of diarrhoea. In these cases, probiotics sources such as yoghurt or saccharomyces
boulardii (Ultralevure°) should be considered.


        Diarrhoea and weight loss are common with miglustat therapy but should be managed
with simple precautions to reduce their consequences on the quality of life of patients on long
term treatment. The mechanisms of diarrhoea seems well identified, and adaptation with time
occurs. No serious long term consequences of these adverse effects should be anticipated.

Prof Pierre Déchelotte
Rouen, April 18th 2003.


To top