EXENATIDE, injection, 5 microgram per dose, 10 microgram per dose, pre-filled pen, 60 doses, Byetta, July 2007
Public summary document for EXENATIDE, injection, 5 microgram per dose, 10 microgram per dose, pre-filled pen, 60 doses, Byetta, July 2007
Page last updated: 16 November 2007
PDF version of this page (PDF 42KB)
Public Summary Document
Product: EXENATIDE, injection, 5 microgram per dose, 10 microgram per dose, pre-filled pen,
60 doses, Byetta
Sponsor: Eli Lilly Australia Pty Ltd
Date of PBAC Consideration: July 2007
1. Purpose of Application
The submission sought an authority required listing as adjunctive therapy in patients with type 2 diabetes no longer achieving glycaemic control despite optimal therapy with metformin and/or a sulfonylurea.
2. Background
This drug had not previously been considered by the PBAC.
3. Registration Status
Exenatide was registered by the TGA on 28 June 2007 as an adjunctive therapy to improve glycaemic control in patients with type 2 diabetes mellitus who are taking metformin, a sulfonylurea, or a combination of metformin and a sulfonylurea but are not achieving adequate glycaemic control.
4. Listing Requested and PBAC’s View
Authority required
Triple oral combination therapy with metformin and a sulfonylurea
Initiation of therapy, in combination with metformin and a sulfonylurea, in type 2
diabetes mellitus patients who have an HbA1c greater than 7% despite maximally tolerated
doses of metformin and a sulfonylurea.
The date of the HbA1c measurement, which must be no greater than 4 months old at the
time of application, must be provided.
Continuation of therapy, in combination with metformin and a sulfonylurea, in type
2 diabetes mellitus patients where the patient has previously been issued with an
authority prescription for exenatide.
Authority required
Dual oral combination therapy with metformin or a sulfonylurea
Initiation of therapy, in combination with either metformin or a sulfonylurea, in
type 2 diabetes mellitus patients who have an HbA1c greater than 7% and in whom a
combination of metformin and a sulfonylurea is contraindicated or not tolerated.
The date of the HbA1c measurement, which must be no greater than 4 months old at the
time of application, must be provided
Continuation of therapy, in combination with either metformin or a sulfonylurea, in
type 2 diabetes mellitus patients where the patient has previously been issued with
an authority prescription for rosiglitazone maleate or pioglitazone hydrochloride.
NOTE:
Exenatide is not PBS-subsidised as monotherapy or in combination with insulins or
a thiazolidinedine.
a. Blood glucose monitoring as an alternative assessment to HbA1c levels will be accepted
in the following circumstances:
clinical conditions with reduced red blood cell survival, including haemolytic anaemias
and haemoglobinopathies
and/or
b. red cell transfusion within the previous 3 months.
Patients in these circumstances will be eligible for treatment where blood glucose
monitoring over a 2 week period shows blood glucose levels greater than 10 mmol per
L in more than 20% of tests. The date of measurement of the most recent blood glucose
level, which must be no greater than 4 months old at the time of application, must
be provided.
See Recommendation and Reasons for PBAC’s view
5. Clinical Place for the Proposed Therapy
Exenatide provides clinicians with an alternative class of drug for adjunctive third
line therapy in adults with type 2 diabetes who are taking metformin, a sulfonylurea,
or a combination of metformin and a sulfonylurea but are not achieving adequate glycaemic
control.
6. Comparator
The submission nominated rosiglitazone as the comparator in its primary analysis and
insulin glargine as the comparator in a secondary analysis.
The PBAC agreed that rosiglitazone and insulin glargine were both appropriate comparators.
7. Clinical Trials
The submission presented two analyses of the clinical trial data:
- Primary analysis: The relative clinical efficacy of exenatide and rosiglitazone was evaluated by an indirect comparison of 8 randomised trials using either placebo (6 trials) or insulin glargine (2 trials) as common comparator.
- Secondary analysis: The relative clinical efficacy of exenatide and insulin glargine was evaluated by a single randomised cross-over trial (GWAO). The results of a second comparative study of exenatide and insulin glargine (GWAA) were used in the economic evaluation. However, the submission did not use Study GWAA in a comparison of clinical effectiveness of exenatide and insulin glargine.
The trials forming the basis of the submission and which have been published are tabulated
below. Only studies used in the indirect comparison have been published.
Trial ID/First author |
Protocol title/ Publication title |
Publication citation |
---|---|---|
Exenatide versus placebo |
||
T112/ DeFronzo RA et al (2005) |
Effects of exenatide (exendin-4) on glycemic control and weight over 30 weeks in metformin-treated patients with type 2 diabetes. |
Diabetes Care (2005) 28(5): 1092-1100. |
T113/ Buse JB et al (2004) |
Exenatide-113 C. Effects of exenatide (exendin-4) on glycemic control over 30 weeks in sulfonylurea-treated patients with type 2 diabetes. |
Diabetes Care (2004) 28(5): 1092-1100. |
T115/ Kendall DM et al (2005) |
Effects of exenatide (exendin-4) on glycemic control over 30 weeks in patients with type 2 diabetes treated with metformin and a sulfonylurea. |
Diabetes Care (2005) 28 (5): 1083-1091 |
Exenatide versus insulin glargine |
||
GWAA/ Heine RJ et al (2005) |
GWAA S. Exenatide versus insulin glargine in patients with suboptimally controlled type 2 diabetes: a randomized trial. |
Annals of Internal Medicine 2005; 143 (8): 559-569. |
Rosiglitazone versus placebo in addition to metformin therapy |
||
49653/094 or Fonseca et al (2000) |
Effect of metformin and rosiglitazone combination therapy in patients with type 2 diabetes mellitus: a randomized controlled trial. |
JAMA (2000) 283: 1695-1702 Erratum JAMA (2000) 284 (11):1384 |
Rosiglitazone versus placebo in addition to sulfonylurea therapy |
||
49653/135 or Rosenstock J et al (2006a) |
For the RESULT Study Group. Effect of early addition of rosiglitazone to sulfonylurea in older type 2 diabetes mellitus patients (>60 years): the Rosiglitazone Early vs SULfonylurea Titration (RESULT) study. |
Diabetes, Obesity and Metabolism (2006) 8: 49-57 |
Rosiglitazone versus placebo in addition to metformin & sulfonylurea therapy |
||
Dailey GE et al (2004) |
Glycemic control with glyburide/metformin tablets in combination with rosiglitazone in patients with type 2 diabetes: a randomized, double-blind trial. |
American Journal of Medicine (2004) 116: 223-229 |
Rosiglitazone versus insulin glargine |
||
Rosenstock J et al (2006b) |
Triple therapy in type 2 diabetes: insulin glargine or rosiglitazone added to combination therapy of sulfonylurea plus metformin in insulin-naive patients. |
Diabetes Care (2006) 29 (3):554-559 |
8. Results of Trials
Exenatide versus rosiglitazone (primary analysis):
The indirect comparisons presented by the submission are summarised in the table below.
Summary of indirect comparisons presented by the submission for the primary analysis
of exenatide versus rosiglitazone
Measures of efficacy |
Combin-ation therapy |
Trials compared in indirect comparison |
Difference in treatment effect |
(Exenatide ? Rosiglitazone) (95% CI) |
---|---|---|---|---|
|
|
|
Bucher method |
Bayesian method |
Change in HbA1c (%) |
MET |
- |
Not performed |
Not performed |
|
SU |
- |
Not performed |
Not performed |
|
MET & SU |
- |
Not performed |
Not performed |
Proportion achieving target HbA1c ?7% (%) |
MET |
- |
Not performed |
Not performed |
|
SU |
- |
Not performed |
Not performed |
|
MET & SU |
- |
Not performed |
Not performed |
Change in body weight (kg) |
MET |
- |
Not performed |
Not performed |
|
SU |
- |
Not performed |
Not performed |
|
MET & SU |
115 vs Dailey |
?2.15 (?2.49 to ?1.81)* |
?3.77 (?4.79 to ?2.76) |
|
GWAA vs Rosenstock b |
?3.47 (?3.99 to ?2.96) |
?5.32 (?6.36 to ?4.31) |
|
|
Pooled |
?2.15 (?2.49 to ?1.81)* |
?4.45 (?5.32 to ?3.65) |
|
Change in fasting total cholesterol (mg/dL) |
MET |
112 vs Fonseca |
?25.9 (?36.0 to ?15.8) |
Not performed |
|
SU |
- |
Not performed |
Not performed |
|
MET & SU |
115 vs Dailey |
?17.8 (?26.8 to ?8.8) |
Not performed |
|
Pooled |
?21.6 (?28.2 to ?14.9) |
Not performed |
|
Change in fasting LDL (mg/dL) |
MET |
112 vs Fonseca |
?13.6 (?22.4 to ?4.8) |
Not performed |
|
SU |
- |
Not performed |
Not performed |
|
MET & SU |
115 vs Dailey |
?10.7 (?18.0 to ?3.4) |
Not performed |
|
Pooled |
?11.6 (?17.0 to ?6.2) |
Not performed |
|
Change in fasting HDL (mg/dL) |
MET |
112 vs Fonseca |
?4.5 (?7.0 to ?2.0) |
Not performed |
|
SU |
- |
Not performed |
Not performed |
|
MET & SU |
115 vs Dailey |
?3.8 (?5.8 to ?1.8) |
Not performed |
|
Pooled |
?3.6 (?5.1 to ?2.1) |
Not performed |
|
Change in triglycerides (mg/dL) |
MET |
- |
Not performed |
Not performed |
|
SU |
- |
Not performed |
Not performed |
|
MET & SU |
- |
Not performed |
Not performed |
HbA1c=glycosylated haemoglobin; HDL=high-density lipoprotein; MET=metformin; LDL=low-density
lipoprotein; SU=sulfonylurea; vs=versus
* These are the results reported in submission but were considered flawed.
The evaluation noted that for Dailey (2004) only the mean change in body weight for
the treatment groups was reported in the publication but not the standard error or
95% confidence intervals. To compute the standard error for the between-group difference
in treatment effect, the standard errors for the treatment effect of the comparing
groups are required. For the standard error for the treatment effect of placebo group,
the submission used “the standard error from the comparable placebo arm of trial 115
(SE 0.17)”. For the standard error for the treatment effect of rosiglitazone, the
submission used the standard error from the comparable rosiglitazone arm of Rosenstock
2006b (SE 0.4)”. The validity of indirect comparison results is therefore also dependent
on the assumptions that (i) the placebo arm of trial 115 was comparable with Dailey
2004 and (ii) the rosiglitazone arm in Dailey (2004) was comparable to the rosiglitazone
arm in Rosenstock (2006b). The placebo groups for Study 115 and Dailey (2004) appeared
to be fairly comparable except the placebo group in Study 115 had slightly higher
mean BMI (33.8 versus 32 kg/m2), longer history of diabetes (9.4 versus 9 years) and
higher mean HbA1c value (8.5 versus 8.1%) at baseline. The rosiglitazone groups for
Dailey (2004) and Rosenstock (2006b) appeared to be fairly comparable except the rosiglitazone
group in Dailey (2004) had longer history of diabetes (9 versus 8.1 years) and lower
mean HbA1c value (8.1 versus 8.7%) at baseline.
The table below shows the proportion of subjects achieving target HbA1c of ?7% at
the end of study. For the three placebo-controlled exenatide studies, the proportion
of subjects achieving target HbA1c of ?7% was statistically significantly greater
in both exenatide treatment arms of 5 ?g and 10 ?g than in placebo. For the GWAA study
which compared exenatide with insulin glargine, the between-group difference was not
statistically significant. The submission reported that only one rosiglitazone study
(Dailey 2004) reported sufficient data to allow the proportion of subjects achieving
HbA1c levels ?7% to be calculated using the full ITT population. In this particular
study, the proportion of subjects achieving target HbA1c of ?7% was statistically
significantly greater in rosiglitazone group than in placebo group. The submission
did not provide a formal indirect comparison of this outcome.
Although the details of concomitant use with statins were provided for exenatide,
such details are not available from the rosiglitazone trials; it is therefore unknown
whether the use of concomitant medication was different between the two drugs and
this may have affected the therapeutic outcome observed in the trials. The economic
model assumes the same proportion of statin use in both treatment arms.
Proportion of subjects achieving target HbA1c levels of ? 7% at end of study
Trial analysis |
Report analysis: full ITT population |
|||||||
---|---|---|---|---|---|---|---|---|
n/N |
% |
RR |
P |
n/N |
% |
RR |
P |
|
112 |
||||||||
Placebo |
10/77 |
13 |
- |
- |
10/113 |
8.8 |
- |
- |
Exenatide 5 ?g |
25/79 |
31.6 |
2.4 |
0.008 |
25/110 |
22.7 |
2.6 |
0.007 |
Exenatide 10 ?g |
39/84 |
46.4 |
3.6 |
<0.0001 |
39/113 |
34.5 |
3.9 |
<0.0001 |
1132 |
||||||||
Placebo |
6/68 |
8.8 |
6/123 |
4.9 |
- |
- |
||
Exenatide 5 ?g |
28/86 |
32.6 |
3.7 |
0.002 |
28/125 |
22.4 |
4.6 |
0.0004 |
Exenatide 10 ?g |
33/80 |
41.3 |
4.7 |
0.0002 |
33/129 |
25.6 |
5.2 |
<0.0001 |
1152 |
||||||||
Placebo |
16/174 |
9.2 |
- |
- |
16/247 |
6.5 |
- |
- |
Exenatide 5 ?g |
54/197 |
27.4 |
3.0 |
<0.0001 |
54/245 |
22.0 |
3.4 |
<0.00001 |
Exenatide 10 ?g |
60/179 |
33.5 |
3.7 |
<0.00001 |
60/242 |
24.8 |
3.8 |
<0.0001 |
GWAA3 |
||||||||
Glargine |
110/267 |
41.2 |
- |
- |
110/268 |
41.0 |
- |
- |
Exenatide 10 ?g |
117/282 |
41.5 |
1.0 |
0.94 |
117/283 |
41.3 |
1.0 |
0.94 |
Fonesca (2000)2 |
||||||||
Placebo |
NR |
7.6 |
- |
- |
||||
Rosiglitazone 4 mg |
NR |
NR |
NR |
NR |
NR |
NR |
- |
- |
Rosiglitazone 8 mg |
25/89 |
28.1 |
3.7 |
NR |
25/113 |
22.1 |
NR |
NR |
Rosenstock (2006a)2 |
||||||||
Placebo |
NR |
NR |
- |
- |
NR |
NR |
- |
- |
Rosiglitazone 4 mg |
NR |
NR |
NR |
NR |
NR |
NR |
NR |
NR |
Dailey4 |
||||||||
Placebo |
24/178 |
14 |
- |
24/184 |
13.0 |
- |
- |
|
Rosiglitazone 4 mg |
75/177 |
42 |
3.1 |
<0.00001 |
75/181 |
41.4 |
3.2 |
<0.00001 |
Rosenstock (2006b) |
||||||||
Glargine |
NR |
48 |
- |
- |
NR |
NR |
- |
- |
Rosiglitazone 4 mg |
NR |
49 |
- |
- |
NR |
NR |
NR |
NR |
EP, evaluable population; HbA1c, Glycosylated haemoglobin specific A1c fraction; ITT,
intent-to-treat; PP, per-protocol population; NR, not reported
1P-value: placebo or glargine is the reference group;
2Study analysis performed on evaluable population;
3Study analysis performed on per-protocol population;
4Study analysis performed on the ITT-population
Exenatide versus insulin glargine (secondary analysis):
In Study GWAO, the primary objective of demonstrating the non-inferiority of exenatide
to insulin glargine in change in HbA1c was supported by the location of the upper
limit of the CI below the pre-specified non-inferiority margin of 0.4%.
Results from Study GWAA were not used in the submission for direct comparison of clinical
effectiveness of exenatide and insulin glargine. The least squares (LS) mean change
in HbA1c from baseline in the study was ?1.00% (SE=0.06) in exenatide group and ?1.05%
(SE=0.06) in glargine group. The proportion of subjects achieving target HbA1c ?7%
at endpoint was 46.4% (=117/282) in exenatide group and 48.03% (=110/267) in glargine
group.
For PBAC’s view of these results, see Recommendations and Reasons.
9. Clinical Claim
The submission described exenatide as being similar to rosiglitazone and insulin glargine in its effectiveness in lowering HbA1c and in safety but is superior to rosiglitazone and insulin glargine in body weight management. The PBAC partially accepted this claim, see Recommendations and Reasons.
10. Economic Analysis
The submission presented a modelled economic evaluation. The choice of the cost-utility
approach was considered valid. Total cost included drug costs, patient management
costs and costs of management of diabetes-related complications.
The estimated base case incremental cost per extra discounted quality adjusted life
year versus either rosiglitazone or insulin glargine fell in the range $15,000 to
$45,000. The estimated base case incremental cost per extra discounted life year gained
versus either rosiglitazone or insulin glargine was > $200,000.
The PBAC had a number of concerns with the modelled economic evaluation presented,
which could result in the incremental cost effectiveness ratio being much higher and
less certain than estimated by the submission, see Recommendations and Reasons..
11. Estimated PBS Usage and Financial Implications
The financial cost/year to the PBS (excluding co-payments) minus any savings in use of other drugs was estimated by the submission to be between $10 - $30 million in Year 5. This is likely to be an underestimate.
12. Recommendation and Reasons
The PBAC agreed that insulin glargine and rosiglitazone were both appropriate comparators
and considered that the submission reasonably established that exenatide is non-inferior
to insulin glargine in terms of its effects on HbA1c, which remains the most important
clinical outcome in the management of diabetes, but that it appears to be associated
with a higher incidence of adverse effects.
The PBAC had some residual concerns about whether the submission had adequately demonstrated
that exenatide is non-inferior to rosiglitazone in terms of its effect on HbA1c, although
acknowledging the difficulties associated with conducting a formal indirect comparison
of the two drugs in terms of this parameter. The Committee was somewhat reassured
by the results of the indirect comparison of the proportion of subjects achieving
target HbA1c levels ≤ 7%. However it was noted that a number of relevant rosiglitazone
trials were excluded from the analyses presented. The PBAC also noted that there is
an ongoing Phase III randomised trial comparing exenatide, rosiglitazone, and exenatide
plus rosiglitazone in metformin treated Type 2 diabetics, and considered that the
results of this trial would help inform this issue. It was further noted that no formal
indirect comparison of the relative safety and toxicity of exenatide and rosiglitazone
had been presented, and the Committee was therefore unable to form a view on this
issue.
The Committee did not accept the submission’s main therapeutic claim of greater weight
loss with exenatide compared with rosiglitazone, noting it is based on data from indirect
comparisons of randomised trials of short duration where the apparent advantage of
exenatide with respect to weight was an observational outcome. This benefit has not
been verified in a properly designed weight loss study or quality of life studies,
nor has it been shown to be durable over the long term. Additionally the claimed advantages
for exenatide in weight gain have not been shown to result in improved morbidity and
mortality outcomes. The Committee did not accept the secondary therapeutic claim of
lowered LDL cholesterol levels for similar reasons.
A final area of clinical concern for the Committee was the question of whether exenatide
treatment increases the risk of malignancies. However, it is noted in the European
Public Assessment Report that a causal relationship between exenatide treatment and
cancer is unlikely, although the relatively small number of subjects and short duration
of follow-up precluded a definitive conclusion.
The PBAC had a number of concerns with the modelled economic evaluation presented.
The model is based on the UKPDS diabetes model which has been validated in terms of
the effect of changes in HbA1c on cardiovascular outcomes, but not in terms of incremental
weight change as a surrogate outcome which quantitatively predicts a subsequent incremental
treatment effect on major cardiovascular events. Additionally the benefit of exenatide
in regard to weight gain is unsupported by any long term studies of clinical outcomes.
The model is extremely sensitive to the disutility associated with weight gain. If
no disutility values for body mass index (BMI) are applied to the primary analysis,
the incremental cost effectiveness ratio (ICER) increases from within a range of $15,000
- $45,000 to > $200,000. The disutility associated with long-term twice daily injection
therapy is not included. The model assumes a constant linear relationship between
BMI over 25 and disutility which may not be reasonable. Given that a BMI > 25 is relatively
common in the population, this assumption imposes a decrement in QOL (and therefore
benefit for exenatide by off-setting this decrement) to people who may have an average
BMI.
There is uncertainty regarding the incremental cost effectiveness of using exenatide
compared to insulin glargine because special price arrangements apply to insulin glargine.
Therefore the PBAC rejected the submission on the grounds of high and uncertain cost-effectiveness
against the comparators in the absence of any evidence of clinical benefit other than
the observational finding of weight loss which has not been shown to be durable or
to translate into morbidity or mortality benefits, and because of unresolved safety
concerns.
Recommendation
Reject
13. Context for Decision
The PBAC helps decide whether and, if so, how medicines should be subsidised in Australia. It considers submissions in this context. A PBAC decision not to recommend listing or not to recommend changing a listing does not represent a final PBAC view about the merits of the medicine. A company can resubmit to the PBAC or seek independent review of the PBAC decision.
14. Sponsor’s Comment
The sponsor has since consulted with the PBAC and Department of Health and Ageing
and clinical experts to clarify the positioning and the valuation of the claimed benefits
and will address these topics in a resubmission. The sponsor disagrees with the statement
that study GWAA was not included in the evaluation of outcomes but will address this
concern in any resubmission. The sponsor also notes regulatory agencies have generally
accepted the product's safety and some international reimbursement agencies have accepted
the value of weight loss in a population of patients with Type 2 diabetes. The sponsor
refers to its own website (www.lilly.com.au) for more details.